Proceedings of the Third U.S.-Japan Workshop Water System Seismic Practices

Transcription

1 Proceedings of the Third U.S.-Japan Workshop on Water System Seismic Practices August 6-8, 2003 This workshop was coordinated by and held at the Los Angeles Department of Water & Power, Los Angeles, California, and primarily supported by the America Water Works Association Research Foundation under Project 2964.

2

3 Proceedings of the Third U.S.-Japan Workshop on Water System Seismic Practices Held at the Los Angeles Department of Water and Power John Ferraro Building Los Angeles, California August 6-8, 2003 Editor & Technical Producer Craig A. Davis, Los Angeles Department of Water and Power Associate Editors Volumes 1, 2 & 4 Jianping Hu, Los Angeles Department of Water and Power Marjar Magee, Los Angeles Department of Water and Power Associate Producers Volume 3 Reggie Brewer, Los Angeles Department of Water and Power T. C. Richard, Los Angeles Department of Water and Power Publication Date: December 29, 2003 Co-sponsored by Los Angeles Department of Water and Power (LADWP) American Water Works Association Research Foundation (AwwaRF) Japan Water Works Association (JWWA) Taiwan Water Works Association (TWWA) Multidisciplinary Center for Earthquake Engineering Research (MCEER)

4 First row (kneeling) left to right: Kenei Ishii, Farid Niknam, Michael Conner, Nobuhisa Suzuki, Yoshihisa Iwasaki, Yoshiharu Sorakuma, Hiroshi Yamada, Katsutoshi Fukuda, Takashi Furuya, Keiichi Murakami. Second row (standing) left to right: William Heubach, Norio Iijima, Glenn Singley, David Lee, Marilyn Miller, Elizabeth Kawczynski, Toshio Toshima, Kenetsu Kojima, Kiyoshi Naito, Tetsuo Tobita, Endi Zhai, Martin Adams. Third row left to right: Ping-Hsin Wang, Le Val Lund, Craig Davis, Seishi Nonaka, Koichi Murata, Masakatsu Miyajima, Hitoshi Hasegawa, Wolfgang Roth. Back row left to right: Jim Woodhams, John Eidinger, Makoto Matsushita, Bruce Maison, Thomas O Rourke, Charles Pickel, Frank Collins, Donald Ballantyne. Participants and observers not shown in photograph: Jean-Pierre Bardet, Surinderjeet (Jeet) Bajwa, Gerald A. Gewe, Donald J. Goralski, Marjar E. Magee, Scott J. Munson, Susan R. Rowghani, Clark Sandberg, Paul Somerville, Mitsuo Takasue, Kevin Wattier, Steve Welch, Cheryl S. F. Chi, Jiin-Song Tsai. Front cover photos: 1) Pipe failure during the 1995 Kobe, Japan Earthquake 2) Damages to pump, operation, and monitor instruments at the Fongyuan Water Treatment Plant from the 1999 Ji-Ji earthquake in Taiwan 3) Lower San Fernando Dam upstream slope failure cause by the 1971 San Fernando California earthquake, USA.

5 Preface The Third US Japan Workshop on Water System Seismic Practices capitalized and expanded upon the successes of the previous two workshops by maintaining the basic principles on which the workshop series was established, providing a forum for water system practitioners engaged in seismic issues to discuss and share their knowledge on the subject, and inviting additional organizations to enhance the knowledge transfer opportunities. In this workshop the Taiwan Water Works Association (TWWA) and the Multidisciplinary Center for Earthquake Engineering Research (MCEER) participated along with the two founding organizations, the Japan Water Works Association (JWWA) and the American Water Works Association Research Foundation (AwwaRF), and all four organizations played an important role in organizing and ensuring a successful workshop. The Los Angeles Department of Water and Power hosted the workshop at the John Ferarro Building in downtown Los Angeles and provided coordination for organizing the workshop. The workshop had a primary focus of encouraging discussions between all participants and documenting the information discussed and provided the opportunity to acquire information from an international gathering of expert practitioners and researchers concerning important aspects to include in a seismic improvement program and future directions for improving water system seismic practices. Highlights of the workshop results include: 1. The first 4-volume proceedings incorporating multi-media interaction that allows the viewing of video recordings of all workshop presentations on a DVD along with the original presentation slides and the final written published papers. 2. An agreement to formally include the TWWA in future workshops and identification of the next two workshop locations at Kobe, Japan, and Taiwan. 3. A summary of discussion results and written recordings identifying future directions for improvements in water system seismic practices and topics for future discussion and development. 4. A summary of goals and tasks important for inclusion in a water system seismic improvement program. As the primary workshop coordinator, I made every effort to make this an enjoyable event and to create a comfortable atmosphere for everyone to meet, talk, and learn about water system seismic practices. In addition, great effort was put forth by numerous individuals to document the information and present it in an understandable and useful manner. It is my hope that the information shared during the workshop and provided in the proceedings will help make advances in the water utility industries understanding of how to approach and implement seismic improvements and mitigations and that the workshop serves as a catalyst for continued communication for the benefit of everyone. Craig A. Davis, Ph.D. Waterworks Engineer, Los Angeles Department of Water and Power i

6 ii

7 US-Japan Workshops on Water System Seismic Practices In August 2000 the East Bay Municipal Water District (EBMUD), in cooperation with the American Water Works Association (AwwaRF) and the Japan Water Works Association (JWWA), organized and hosted a planning meeting between U.S. and Japan water utility organizations for the purpose of: 1. Sharing information and technologies between the two countries specifically relating to water system seismic practices, which transformed into the First US- Japan Workshop on Water System Seismic Practices, and 2. Exploring the concept of developing and holding future workshops in Japan and the United States to foster the interchange of information important for improving water system seismic performances. The planning meeting and first workshop, which was initially conceived by EBMUD in 1997, was successful in developing a cooperative agreement between AwwaRF and JWWA for holding future workshops and also developed general meeting guidelines. To date, three workshop meetings have been held: 1. Oakland, CA, August 8-9, Tokyo, Japan, August 6-8, Los Angeles, CA, August 6-8, Results of each workshop are documented in proceedings available from AwwaRF and JWWA. The workshops emphasize the practical understanding and implementation of water system seismic practices through active participation of the water organizations with some cooperation and participation of academic and consulting experts in the field. The workshop objectives are to: Create a forum for the active discussion and exchange of information between countries. Identify and document the best available current technologies in water system seismic mitigation practices based on current waterworks organization practices. Provide a practical information source on seismic mitigation efforts for water utilities. Identify, where practicable, technology improvements needed to practically and efficiently advance water system seismic practices. It is important for the workshop to focus primarily around the practicing water utility organizations to foster the transfer of information between organizations and countries in the development of water system seismic practices. It is equally important to include participants from academia to help foster the learning and discussion of water system practitioners and to also provide an opportunity for academic researchers to learn from practitioners and identify where industry research is needed. Selection of academic researchers who are familiar with the waterworks industry in the US, Japan, and Taiwan is critical. Academic professors have provided a significant contribution to the iii

8 establishment of this series of workshops and improvement of the American, Japanese, and Taiwanese water system seismic practices. The series of workshops has shown expanding interests from the international community. The 2001 workshop gained interest and participation from organizations in Great Britain, Taiwan, and Central America. The 2003 workshop incorporated active inclusion of the Taiwan Water Works Association (TWWA). Following the 2003 workshop there was a unanimous agreement to formally include the TWWA in future workshops and the workshop titles will reflect their active participation. The expanding interest in these workshops is only one of many indicators of their great success. During the 2003 workshop, as in the previous two workshops, there was unanimous agreement among attendees to hold future workshops. As a result, the next two workshop locations were identified; in addition, procedures for identifying the workshop titles and guidelines for timing between each workshop were developed. Discussion of these topics reflects the maturing and expansion of the workshop program, which is a direct result of the obvious importance to the waterworks industry. The next two workshops are planned for: 4. Kobe, Japan, sometime between September 2004 and April Taiwan, approximately 18 to 24 months following the fourth workshop. Future workshops will be held approximately 18 to 24 months following the previous workshop. The first country named in the official workshop title will be the host country and the host country will determine the order of the following names. For example the next workshop title will be the 4 th Japan-US-Taiwan Workshop on Water System Seismic Practices and the following will be the 5 th Taiwan-Japan-US Workshop on Water System Seismic Practices. Information exchanged in the workshops has made significant advances for the water utility industries understanding of how to approach and implement seismic improvements and mitigations. Water system seismic practices encompass emerging and developing fields that are not yet fully understood. As a result, it is important that organizations learn from each earthquake and from the knowledge gained within each water organization in each participating country. The intent of these workshops is to foster such knowledge and document the information for the benefit and interest of every water organization. iv

9 Third U.S. Japan Workshop on Water System Seismic Practices Overview The Third US-Japan Workshop on Water System Seismic Practices was held at the Los Angeles Department of Water and Power (LADWP) John Ferraro Building main headquarters in Los Angeles, California from August 6 to 8, There were 16 attendees representing the Japan Water Works Association (JWWA), 3 attendees representing the Taiwan Water Works Association (TWWA), 15 attendees representing the American Water Works Association Research Foundation (AwwaRF), and 3 attendees representing the Multidisciplinary Center for Earthquake Engineering Research (MCEER). In addition, there were several LADWP managers who attended and participated in the workshop, eight guests from the Southern California area, and numerous LADWP employees who observed the workshop discussions and presentations. A listing of United States, Japan, and Taiwan participants are provided with contact information in Appendix VI. The participants included a very good mix of practitioners from water utilities, consultants, and researchers. In total, over 100 people gained direct benefit of the workshop through participation and observation as it transpired. The entire workshop was video recorded, except for the tour, with the intent of distributing the recordings to interested people and organizations to provide further learning benefits from this workshop. The workshop had two days of presentations and discussion and one day for a tour, with a welcome reception held on the evening of August 5, 2003 at the New Otani Hotel in Los Angeles, California. There were a total 26 presentations during seven meeting sessions, covering the following general topics: Number of papers Topic AwwaRF JWWA TWWA Total Risk Assessment Seismic Performances Seismic Preparedness and Readiness Risk Management Post Earthquake Recovery Seismic Resistant Design Total The Table of Contents on page xi and the Agenda in Appendix I show the presentation order and titles. The workshop was organized with a focus of providing time for the participants to discuss the presentation topics. Time was allotted at the end of each session for open discussion and several breaks and events were incorporated in the agenda to allow further informal discussions. Written recordings were made for each session, which are summarized in Appendix II. Transcriptions of the opening and closing remarks were made from the video recording and are included as a part of the proceedings. There was a key note presentation describing the History of Los Angeles Water System Seismic Improvements used as an introduction to a day long tour of Los Angeles Water System facilities on August 7, 2003, and a key note lecture discussing system performance and management, titled Lessons v

10 Learned From The World Trade Center Disaster for Water Supply Management given at the banquet the evening of August 7, A summary of the technical tour is provided in Appendix V. The panel discussion, having the main theme of Important Aspects to Include in a Water System Seismic Improvement Program, was held on August 8, 2003 during the final Session VIII. The panel of 5 experts, one practitioner and one researcher each from the United States and Japan and one researcher from Taiwan, created an open discussion with all attendees with the goal of identifying what they and the audience feel are important aspects of water system seismic practices and what are the primary topics and issues in need of further discussion and development. The discussion was monitored and recorded by experts in the field. The panel identified and presented information from their knowledge and experiences and used information presented in the workshop to stimulate the audience to participate in an open discussion and share their knowledge, ideas, and experiences. The panel discussion was very successful in achieving its goals and a transcribed record of it is provided in these proceedings. Appendix II summarizes the discussion written recordings and results and identified the following future directions for improvements in water system seismic practices, which may serve as future workshop themes and topics: Regional redundancy and back up in water supply systems. US/Taiwan design guidelines for pipe joint seismic design, including licensing Japan ductile iron pipe S-joint in other countries. Pipeline replacement cycles for seismic improvement. Overall water system seismic performance assessments. Needs and use of near real-time ground motion assessments. Seismic performance criteria and objectives. US/Japan/Taiwan water system seismic design standards/guidelines. Considerations of national vs. local standards. Comprehensive seismic readiness/recovery strategy. Multi-benefit aspects of seismic improvements. How to establish localized seismic performance and post-earthquake recovery goals. Financial considerations for seismic preparations and post-earthquake recovery. The workshop discussions and recordings indicate how the US, Japan, and Taiwan have similar seismic improvement interests. In some cases the seismic problems are approached in similar ways while in other cases the problems are approached very differently. The greatest value attained from these workshops may lie in working to understand the differences. A survey was sent out prior to the workshop requesting input from all participating organizations on 5 aspects they consider important to include in a seismic improvement program. Appendix IV presents the survey and summarizes results of the 64 different aspects from the 13 responses received and provides some evaluation and conclusions based on the survey responses. vi

11 Appendix III describes and tabulates important aspects to include in a water system seismic improvement program. The survey results, combined with information provided throughout the workshop, were processed to develop a listing of 6 goals and 32 primary tasks that are important to include in a water system seismic improvement program. The information summary identified the purpose of a water system seismic improvement program is to ensure the safe provision of water following an earthquake and the following six seismic improvement goals: Provide adequate post-earthquake water supply throughout service area. Reduce earthquake damage to facilities. Ensure minimum level system functionality and rapid system recovery. Achieve a rapid emergency response. Accomplish a well planned, cost-effective, and publicly responsible seismic improvement program to ensure public safety. Continually develop and improve earthquake disaster prevention capabilities. The workshop was concluded with the placement of eyes on a Daruma doll as a symbol of good success and achievement for an accomplishment that has enlightened our vision and goals for the future through persistence and determination (Nana Korobi Yaoki). The Daruma doll was presented during the Panel Discussion, and Mr. Matsushita from the Kobe Waterworks Bureau explained the Japanese custom of placing eyes on the Daruma. A Daruma is an eyeless Japanese folk art tumbler doll (okiagari koboshi) shaped to persistently return to an upright position. The eye painting ceremony was performed by Craig Davis presenting the Daruma to Elizabeth Kawczynski and Kenei Ishii to hold on behalf of AwwaRF and JWWA, with the eyes being painted by Norio Iijma and Glenn Singley, mangers from the Tokyo Waterworks Bureau and LADWP; symbolizing the cooperative success of this third workshop and commitment of the United States and Japanese waterworks organizations toward furthering improvements in Water System Seismic Practices through these series of workshops. Following the Daruma eye painting ceremony, the workshop was officially closed. Preliminary proceedings were provided as a part of the workshop and the final workshop proceedings are divided into 4 volumes as described below: Volume Item and description I Workshop proceedings including agenda, written papers, opening and closing statements, discussion results, survey results, contact information, seismic improvement goals and tasks, tour information. Initially provided on CD and print copy. II Workshop presentation slides, CD version only. III Recorded DVDs, 4 disk set. IV Workshop photographs, CD version only. All proceeding volumes are organized consistent with the agenda outline presented in Appendix I. The proceedings of Volume I are initially provided in a hard print copy or as Compact Disk (CD) with an acrobat reader file. Volume II provides a copy of the vii

12 slides used by each presenter during the workshop, which provide significant and important information that is not available in the printed papers. The slides also provide a useful reference in conjunction with the video recordings for better clarification and transfer of information. Volume II is only provided on a CD with an acrobat reader file. As apart of this workshop a video recording was made and incorporated as Volume III of the proceedings in the form of a Digital Video Disk (DVD). Volume III comes in a 4 DVD disk set that is organized consistent with information provided in the agenda. Each presentation can be accessed individually from the DVD menu as summarized in the following table. Disk DVD Content Date 1 Opening Remarks, Session 1, Session 2. August 6 2 Session 3, Session 4, Session 5. August 6 and 7 3 Session 6, Session 7. August 8 4 Session 8, Closing Remarks. August 8 Volume IV presents photographs on a CD that were taken throughout the workshop. The photographs are only available on CD. Volumes 1, 2, and 4 are all located on one CD. viii

13 Acknowledgements Acknowledgement and thanks are extended to the Los Angeles Department of Water and Power (LADWP) for hosting, co-sponsoring, and coordinating the workshop organization, the American Water Works Association Research Foundation under the direction of Elizabeth Kawczynski for co-sponsoring and helping to organize the workshop and United States participants, the Japan Water Works Association under the direction of Kenei Ishii for cosponsoring and organizing the Japanese participants, the Taiwan Water Works Association under the direction of Jiin-Song Tsai for cosponsoring and organizing the Taiwanese participants, the Multidisciplinary Center for Earthquake Engineering Research under the direction of Donald Goralski for co-sponsoring the workshop, and the many participants from the United States, Japan, Taiwan, and United Kingdom who made presentations, chaired sessions, made recordings, and participated on the panel (listed in Appendix I). Special acknowledgement and thanks go out to Seishi Nonaka, Mitsuo Takasue, and David Lee who served as interpreters. The numerous LADWP management and employees who were instrumental for coordinating, organizing, and enacting this workshop are acknowledged; there are too many to mention, but the primary LADWP personnel responsible for this workshop were Gerald Gewe, Glenn Singley, Martin Adams, Craig Davis, Mark Aldrian, Phillip Lau, Marjar Magee, Linh Phan, Jianping Hu, Elizabeth Shiratori, Sharman Buggs, Reggie Brewer, and T.C. Richard; Dexter Lee and Kien Hoang developed the workshop logo. Doris Lovalo from the American Water Works Association Research Foundation is acknowledged for her contributions toward organizing the workshop. ix

14 x

15 Table of Contents Preface... i US-Japan Workshops on Water System Seismic Practices... iii Third U.S. Japan Workshop on Water System Seismic Practices, Overview...v Acknowledgements... ix Table of Contents... xi Opening Address to the 3 rd U.S. Japan Workshop on Water System Seismic Practices...3 Craig A. Davis, Ph.D....3 Elizabeth Kawczynski...5 Gerald A. Gewe...7 Norio Iijima...9 Ping-Hsin Wang...11 Introduction to the Workshop Format and Organization...13 Technical Session I: Risk Assessment and Analysis I...15 Seismic Upgrades for 20 Suburban Water Utilities in the San Jose Bay Area John Eidinger...17 Government Policies on Earthquake Disaster Prevention Measures related to Water Supply in Japan Yoshihisa Iwasaki...27 Anti-Seismic Measures of Existing Water Supply Facilities: A case study of an antiseismic plan of Inagawa water treatment plant Keiichi Murakami, Kazuo Mishima, Takashi Hanmoto, and Kazuo Ogura...39 San Francisco Public Utilities Commission (SFPUC) - Capital Improvement Program Surinderjeet (Jeet) S. Bajwa...53 Seismic Measures for Waterworks in Yokosuka City Takashi Furuya...55 Technical Session II: Seismic Performance, Predarednes and Readiness...67 Damages and Motion of Pipelines buried in Liquefied Ground Toshio Toshima, Hiroyasu Ohama and Shogo Kaneko...69 xi

16 Seismic Evaluation of Water Supply System in Health Care Facilities Masakatsu Miyajima, Nebil Achour and Masaru Kitaura...79 The Research of Damages of Public Water Supply Pipelines During the Ji-Ji Taiwan Earthquake on September 21, 1999 Ping-Hsin Wang...87 Emergency Operation Planning - How Contra Costa Water is Building Earthquake Response Capabilities In Calm to Excel in Response Under Emergency Stephen J. Welch...99 Emergency Water Supply Facilities of Hachinohe Regional Water Supply Authority Kenetsu Kojima Technical Session III: Seismic Risk Management and Post Earthquake Recovery 115 Seismic Damage Simulation of Distribution Pipeline Based on the Monitoring Data Collected by a Seismometer Network Kazuya Yamano, Katsuhiko Eguchi, and Koichi Murata Prioritization of the San Diego Water Department CIP Michael E. Conner and Frank X. Collins Seismic Practices Evaluation of Kobe Water System using Risk Management Approach Makoto Matsushita Knowledge Management in Engineering - A Methodology that can be applied to Seismic Risk Management Jim Woodhams Emergency Restoration for Water Supply Following Earthquake Disasters in the City of Yokohama Kiyoshi Naito Technical Session IV: Los Angeles Water System History of Los Angeles Water System Seismic Improvements Craig A. Davis and Le Val Lund Technical Session V: System Performance and Management Lessons Learned From the World Trade Center Disaster for Water Supply Management T.D. O Rourke xii

17 Technical Session VI: Risk Assessment and Analysis II Maintenance Management System for Lowering Possible Seismic Damages onto Water Works Facilities Jiin-Song Tsai and Cheryl S.F. Chi An Overview of The Metropolitan Water District of Southern California s Seismic Program Clark Sandberg and Ray DeWinter A Fast Simulation Method for Predicting Seismic Responses of An Extensive Water Distribution Network Nobuhisa Suzuki Multi-Hazard Risk Assessments of Water Systems, Elements In Common with Seismic, Security, and Other Risk Studies Donald Ballantyne A Study on the Development of a Backup System in a Big Urban Area Fukuhisa Iwasaki, Yoshiharu Sorakuma, and Kyoji Wakamatsu Technical Session VII: Seismic Resistant Design The Reinforcement Work of the Embankment of the Yamaguchi Reservoir Hiroshi Yamada and Isao Tahara Seismic Upgrade of East Bay Municipal Utility district s Mokelumne No. 3 Aqueduct Bruce F. Maison Seismic-Proof Design for the Structures of the Water Treatment Plant in Niigata Hitoshi Hasegawa, Isao Hokari, and Kouei Ito A Practical Approach to Mitigation of Earthquake Pipeline Damage William Heubach Technical Session VIII: Panel Discussions Panel Discussion - Important Aspects to Include in a Water System Seismic Improvement Program Introduction Seismic Resistant Design Risk Assessment and Analysis xiii

18 Seismic Risk Management Seismic Preparedness and Readiness Seismic Performances and Post-Earthquake Recoveries Session Closing Closing Remarks Craig A. Davis Jiin-Song Tsai Kenei Ishii Elizabeth Kawczynski Glenn Singley Closing Ceremony Appendix I: Workshop Agenda... A-5 Appendix II: Discussion Results... A-13 Future Directions for Improvement in Water System Seismic Practices: Topics for Future Discussion and Development... A-13 Appendix III: Important Aspects to include in a Water System Seismic Improvement Program... A-23 Appendix IV: Survey of Important Aspects to Include in a Water System Seismic Improvement Program... A-31 Survey Results from Different Organizations... A-34 Survey Results for Different Seismic Improvement Program Topics... A-41 Survey Request on Important Aspects to Include in a Water System Seismic Improvement Program... A-50 Appendix V: Los Angeles Department of Water and Power Facilities Tour... A-55 Tour Itinerary... A-55 Magnolia Trunkline Project... A-57 Van Norman Complex... A-59 Hollywood Water Quality Improvement Project... A-67 Appendix VI: Workshop Participants... A-71 xiv

19 3 rd US-Japan Workshop on Water System Seismic Practices OPENING REMARKS Craig A. Davis Workshop Coordinator Waterworks Engineering Los Angeles Department of Water and Power Elizabeth Kawczynski Senior Account Manager American Water Works Association Research Foundation Gerald A. Gewe Assistant General Manager Water Services Organization Los Angeles Department of Water and Power Norio Iijima Director General Waterworks Bureau Tokyo Metropolitan Government Japan Water Works Association Ping-Hsin Wang Manager Administration Office of the Eight District Taiwan Water Supply Corporation Taiwan Water Works Association 1

20 3 rd US-Japan Workshop on Water System Seismic Practices 2

21 Opening Address to the 3 rd U.S. Japan Workshop on Water System Seismic Practices Craig A. Davis, Ph.D. Workshop Coordinator Waterworks Engineer, Los Angeles Department of Water and Power Good morning. I am very pleased to open this 3 rd U.S. Japan Workshop on Water System Seismic Practices in Los Angeles. The workshop will be held over the next three days here in the John Ferarro Building and around the City of Los Angeles during the planned tour for tomorrow. I want to point out that we held a Welcome Reception last night at the new Otani Hotel for those who were able to fly in yesterday afternoon. The welcome reception was hosted by the American Water Works Association (AwwaRF). I thank and congratulate AwwaRF on a very fun and successful reception. I want to take this time to introduce and thank the workshop sponsors. The Japan Water Works Association has done a terrific job in bringing a large delegation of Japanese participants. This year we have added the participation of the Taiwan Water Works Association (TWWA). Mr. Wang is currently attending for the TWWA, and we will have two more TWWA participants joining us tomorrow. We have also added the Multidisciplinary Center for Earthquake Engineering Research (MCEER) as a co-sponsor this year. MCEER will be hosting a reception prior to the banquet on Wednesday night. We currently have two MCEER representatives attending the workshop and will have a third join us later. I would like to provide special thanks to the Los Angeles Department of Water and Power, of which I am an employee, but I was not the person to make this workshop happen. This workshop happened as a result of our management, especially Glenn Singley, who is a Director of the Water Engineering and Technical Services Business Unit. The primary co-sponsor is AwwaRF. AwwaRF has done a terrific job of helping us at the LADWP to make this workshop happen and we will have some welcoming remarks from AwwaRF in just a few minutes. I will provide some details on how the workshop is organized immediate following all of the opening remarks. However, I need to point out two items at the moment. 1) We are recording the entire workshop, including these opening remarks, and it is very important to speak directly into the microphone so that the sound is adequately recorded; and 2) A map showing restroom locations is provided in the front cover of the preliminary proceedings. At the start of planning for this workshop, several people from the United States and Japan had discussions about what we wanted to do at this workshop. We really felt that it was important and did our best to make this a fun event and to encourage more discussions. We want everyone to have fun and to meet everyone in attendance. It is great to see old acquaintances and to meet new friends. We have made every opportunity in the agenda to provide for open discussions. There are many scheduled breaks to allow people to have fun, relax, talk, and learn. I thank you all for coming and I really appreciate your attendance here. 3

22 4

23 Elizabeth Kawczynski Senior Account Manager American Water Works Association Research Foundation Greetings, welcome, and thank you for coming. It is a pleasure for the American Water Works Association Research Foundation (AwwaRF) to be a part of this exchange related to seismic practices for water utilities. This is our 3 rd workshop on Water System Seismic Practices. Some participants, like me, have been to all three workshops in this series, which started in Oakland three years ago, at the East Bay Municipal Utility District, then moved to Tokyo two years ago, and now we are here in Los Angeles. This is a very important topic for AwwaRF, seismic practices for water utilities, because many utilities that may be impacted by possible earthquakes are very large and are some of the largest AwwaRF subscribers. The information exchange that goes on between people in these workshops is going to make the world a better place tomorrow; it may not happen today, but it is going to happen. I recently heard about how information presented in Tokyo was very useful in helping to a utility in the United States. It is clear there are good things that happen here. I encourage everyone to talk and keep the dialog active so that we can continue to do more of these workshops in the future. I want to thank the Los Angeles Department of Water and Power (LADWP) and especially Craig Davis. He said he did not do it alone, and I know he didn t, but he did a lot! Also thanks go out to Glenn Singley and the rest of the LADWP for wanting to host the event and putting forth all the resources necessary to make it happen. Also, the Japan Water Works Association was kind enough to bring their contingent. We enjoy seeing you all again and I am sure we will have some more good times together. I will now hand the microphone back to Craig Davis who will introduce some of the other people who would like to say hello and thank you. 5

24 6

25 Gerald A. Gewe Assistant General Manager Water Services Organization Los Angeles Department of Water and Power Good morning and welcome to Los Angeles. This is the first time I am 5 minutes ahead of schedule at the beginning of a conference. Congratulations on getting started on time. We are having very good weather. These are about the clearest skies you will see here. We hope to provide you with an informative and wonderful time in Los Angeles. Seismic engineering is rather tortuous in Los Angeles. We have had a couple of large seismic events in my career here. The first one was in 1971, with the Sylmar Earthquake. We were very successful in dealing with the Sylmar earthquake because we had a large staff of employees who could respond to the damage and help with a rapid and successful recovery. A little over 20 years later, in 1994, we had the Northridge Earthquake. At that time we had downsized our workforce by about 2/3 from what we had during our growth days in the city during the 1970 s. In 1994 we were basically built out, and not growing with large subdivisions like we had been; instead the city is now growing and building upward. With the smaller work force, we had to work together with other organizations to solve the problems. One of the key things, besides engineering, is the local organizations. What we found extremely helpful was that we could work with other organizations in the area. Everybody who saw television coverage of the Northridge Earthquake would have thought that L.A. was completely demolished. The television cameras went to the same areas of damage repeatedly. If the television was your only source of information, you would have believed there was no city left. Yet only five miles away from the center of damage virtually nothing changed. By working with other communities away from the center of damage, who were not affected and can send resources in, we accomplished the restoration very effectively and efficiently. One of the biggest disadvantages of pre-planning is the attorneys. We have not yet been able to really successfully achieve a pre-planned set of arrangements because every attorney is afraid of the liability on his client. We ve basically been forced to work as managers until we get to the event, work together until we get it done, and then worry about the paperwork after words. This is not a good way of doing business, but that is the reality in the way we work and we have had to recognize it. As a result, talking about how things can get accomplished following a major event, establishing informal relationships, and the networking that goes on in events like this is crucial for the long term success of meeting our customer s needs. We have learned a lot from our past earthquakes, as I am sure many of you who have gone through earthquakes have as well. We have made changes and redesigned facilities. In the next earthquake we will learn new lessons, but the earthquakes will have far less impact because of the changes we have and are making. Sharing the information we have learned, and in turn learning from each of your as to the changes you have made as a result of your experiences, is important to all of us to providing the best service to our customers. I thank you for coming to Los Angeles and participating. I trust that you will have really good sessions throughout the next three days. 7

26 8

27 Norio Iijima Director General of Waterworks Bureau Tokyo Metropolitan Government On behalf of the Japan Water Works Association Good morning ladies and gentlemen, my name is Iijima, Director General of Waterworks Bureau, Tokyo Metropolitan Government. Welcome to the 3rd US-Japan Workshop on Water System Seismic Practices. I am really glad to be here in this Workshop, and to meet great members like you working on measures of earthquakes. It has been 8 years since Great Hanshin-Awaji Earthquake Disaster occurred in Japan. We had great shock to this disaster took 6,400 lives and 260,000 buildings of collapse. However, on the other hand, this sever earthquake taught us a lot of things and brought us big improvements on the provisions for earthquake disaster in our country. There is a proverb, Danger past, Got forgotten. This means that valuable lessons you learned from the event will fade away as time goes on. Moreover, it is said that Japanese people are easy to heat up and also easy to cool down. I think that we must be attentive with that the tension of the hearts to the earthquake disaster declines elapse of time. The mean time between 2 severe earthquakes at the same region is quite long. In Tokyo, there have not been any severe earthquakes for 80 years since Great Kanto Earthquake occurred in It is difficult to estimate when the severe earthquake will occur, and also to keep a tense atmosphere against earthquake in daily life. Overconfidence can be very dangerous. Both U.S. and Japan have had numbers of experiences on severe earthquakes, and we should make all possible efforts to penetrate the lessons we have ever learned into society. I believe that it is our common mission to forward provisions against this disaster with leadership to minimize the actual damage to the people. Measures against earthquakes can be altered depending on the type of the earthquake, its damage status, or condition of the local community. Also sometimes unique technologies can be developed which is based on the regional peculiarity of technology. In 1891, Nobi Earthquake struck Japan. This became a trigger to form Earthquake Prevention Investigative Committee, and this was the very beginning of seismography in our country. In 1923, as I mentioned, Great Kanto Earthquake struck Tokyo and we experienced the big fire causing fatal damage even more than direct damage caused by the quake itself. In 1946, Nigata Earthquake occurred and liquefaction of the ground was recognized the first time. Miyagiken-oki Earthquake occurred in 1978 caused a severe damage to lifeline systems including water distribution pipes. This earthquake made us realize the importance of taking good measures to protect lifeline systems. Destructive power of Hyogoken-nanbu Earthquake in 1995 was far beyond from our expectations and obliged us to change the way of total ideas on earthquake-proof designs. The history of the measures against earthquakes is also the history of earthquakes occurred in Japan, and the measures have been developed from the experience of earthquakes. In US, experience from those severe earthquakes like San Fernando in 1971, Loma Prieta in 1989 and Northridge in 1994 might brought opportunities to develop the measures. I guess all the nations of us here have the similarity about this point. The provision for earthquake disaster is one of the most important policies at Tokyo Metropolitan Government. We have earthquake disaster prevention plan to proceed with the earthquake-resistant reinforcement of the facilities steadily. Also there is an emergency measure plan 9

28 in order to restore the facilities immediately after the earthquake occurred and make an emergency water service activity possible as quickly as we can. The purpose of the earthquake-resistant reinforcement of the facilities is to minimize damage by earthquakes. Reinforcement works on purification plants, pumping stations and other major facilities of Tokyo Waterworks are under the process at this stage. The measures are implemented according to the result of examinations checking if the facilities satisfy the most advanced seismic criterion. Bank bodies of Reservoirs located near the civilized region are strengthened as earthquakeproof design, and details regarding this case will be presented in this Workshop later on. In addition, the important purification plants have power generation facilities to get prepared for the power failure caused by earthquakes. Regarding the pipes, Tokyo Waterworks now proceed to renew the old distribution pipes or exchanging lead service pipes to stainless steel pipes. At present, the rate of Ductile Cast Iron Pipe and Steel Pipe occupy 96% of the total distribution pipes. And the rate of the stainless pipe becomes 97% in the service pipe. These pipes were adopted for their strength to the earthquake and resistance to corrosion. Moreover, for the emergency water service, about 190 emergency water tanks are secured within a distance of 2km everywhere in Tokyo. These are the major actions Tokyo Waterworks is now taking, and we think that all the effort make sure to minimize damage and bad influences on the resident life of Tokyo. I think there are various ways of measures against earthquakes for every different places, and some of them could be very unique. Exchanging or integrating those various experiences, knowledge, efforts, and technologies bring us huge possibilities to develop measures against earthquakes, and it has been proved by past twice Workshop held already. We can get information of earthquake disaster from the papers. However, there should be a limit in the result. What we really need is the vivid knowledge comes from actual experiences. In this Workshop, people gathered from various places with various experiences and knowledge. We should make the best use of this opportunity to have discussions from various points of view. Valuable ideas could be developed here, and we can share those. Finally, I would like to thank for all the people who put best effort to hold this Workshop. I am really appreciating your cooperation, and I wish this Workshop will be successful with excellent results. Thank you very much. 10

29 Ping-Hsin Wang Manager, Administration Office of the Eighth District Taiwan Water Supply Corporation On behalf of the Taiwan Water Works Association Good morning Lady s and gentlemen. I am from Taiwan. Thank you for your invitation for attending the 3 rd US-Japan Workshop on Water System Seismic Practices. I would like to send you our sincere appreciation and wish good success in the future. Dr. Tsai, the original scheduled speaker, was delayed to arrive and so I am filling in for the opening remarks. This is the first time we are participating in this workshop and we really appreciate this opportunity. We are hoping that in 2005, or shortly thereafter, we can host the next workshop in Taiwan. With these workshops, we are hoping to have the technology exchange in seismic engineering. In the past, these workshops were held in Japan and United States and we are hoping you will give us an opportunity to hold a future workshop in Taiwan. Interpretations provided by David Lee, Senior Engineer, East Bay Municipal Utility District 11

30 12

31 Introduction to the Workshop Format and Organization Craig A. Davis, Ph.D. Workshop Coordinator Waterworks Engineer, Los Angeles Department of Water and Power I will now introduce to you the items you should have already received as a part of this workshop, describe the workshop format, and how it is organized. As you entered and registered for the workshop you all should have received a bag and preliminary proceedings. Please ensure that you received these items. Within the bag there are a number of gifts and supplies for you. We developed a workshop logo this year, which is on the front of the bag and several of the gifts and supplies. In the bag you will find a coffee travel mug imprinted with the workshop logo. We have also provided you with bottled water complements of the Los Angeles Department of Water and Power (LADWP). You are also given a LADWP T-shirt that can be worn on the tour and elsewhere around the world. Also in the bag are several supplies useful for the workshop. We received this morning some gifts from the Japan Water Works Association (JWWA) for the non-japanese participants. The JWWA was kind enough to give each of us a handkerchief, sweat towels as Mr. Ishii described, because he expects us to work very hard this week. The JWWA has also provided us with some handouts. We thank the JWWA and especially Mr. Ishii and Mr. Fukuda dearly for these items. These will be handed out soon. Mr. Yamada, from the Tokyo Metropolitan Government Bureau of Waterworks, has provided us with some gifts and brochures which will also be handed out. Dr. Jiin-Song Tsai from Taiwan was scheduled to present in Session 3 today, but due to his unexpected late arrival we have now made arrangements for him to present in Session 6 on Friday. In Dr. Tsai s place, Koichi Murata from Osaka will be presenting in Session 3 today. You are all provided with name badges. Please keep your badges on at all times and please bring them every day when you return. This will allow you to pass freely throughout the building without security concerns. We have parking validations at the registration table for those who drove their cars. In case of an emergency there will be information provided over the emergency intercom speakers and please exist through the doors and out to the patio. Marjar Magee and Lihn Phan have been a very key people for organizing this workshop and if you have any questions or need any help, please contact them, Elizabeth Kawczynski, or myself. The LADWP felt it very important to gain as much as we can from this workshop. As a result, we have invited several groups of people from the Southern California area to attend the workshop, such as key consultants who help the LADWP and other organizations perform earthquake related evaluations, several local Southern California city water departments, and many LADWP water system managers. We have also taken this opportunity to have approximately 40 LADWP water design engineer staff members to sit in on the session and learn as much as possible from the workshop presentations and discussions. We thank you all for this opportunity to learn from the very valuable information you will be providing us here in Los Angeles. We have also provided you with several displays of LADWP seismic instruments that we use and some Proctor Needles, which are tools invented by the LADWP for performing compaction testing. The displays are labeled and located on the counters in this room. The workshop will proceed in several sessions. There are five presentations schedule for each session, except for one session on Friday that has four presentations. All presentations are schedule 13

32 to use PowerPoint. Mr. Jainping Hu will set up and organize all presentations. If you need any assistance with your presentations please ask Jainping Hu. There is a chairperson table at the front of the room where we will have two chairpersons from two different countries for each session. The Chairpersons have a stopwatch to time each session and we have cards to inform you when 10 minutes, 5 minutes, 1 minute, and 30 seconds remain in your twenty minute presentation. When your presentation time is completed the Chairpersons will inform you with the Stop sign. At that time please conclude your presentation. There is a 10 minute discussion period scheduled for the end of each session period. If any of the presentations are less than their 20 minute allotted time, the chairpersons may allow some questions and answers during the remaining time. It is important that we focus on staying within the scheduled session time. The sessions are being recorded in two ways: (1) we are video tape recording the workshop and I am being recorded right now, and (2) selected U.S. participants have been requested to document in writing the general items that are discussed in each session. We have a general theme of documenting what the participating organizations and individuals believe are the Important Aspects to Include in a Seismic Improvement Program. The written recordings are intended to be prepared for aiding with the Panel Discussion in the final workshop session. The panel members will use their own judgments on what they want to discuss during the panel session, but the concept is to document ideas that people find important throughout the workshop so that we can continue with those topics during the panel session if it is deemed appropriate. The plan is not to have a rigid structure, but to aid as much as possible in encouraging learning and discussion. Much of what I have just stated was previously ed in a document to each of you prior to the workshop. This document was not included in the preliminary proceedings for fear of providing a perception of being too rigid with the workshop structure. However, if anyone is interested in obtaining a copy of this document, please let me know and we will get one to you. The primary objective is to have fun, relax, provide information, discuss, meet new friends, get reacquainted with old friends, and have a good time over the next three days. Lastly, we have a special table for our two most important people in the room our Japanese interpreters. The interpretations will proceed similarly as was done in the 2 nd workshop held in Tokyo. We will not interpret the presentations, unless the speaker requires specific translation support. The questions and answers after each presentation, if there is time, and during the discussion periods at the end of each session will be interpreted. Mr. Wang from Taiwan will have personal interpretations provided by David Lee from the East Bay Municipal Utility District. If there are any particular Japanese interpretations needed during any of the presentations, please make arrangements with the interpreters to do this quietly. We have organized the seating arrangements to best accommodate interpretations for the different languages. 14

33 3 rd US-Japan Workshop on Water System Seismic Practices TECNICAL SESSION I Risk Assessment and Analysis I Seismic Upgrades for 20 Suburban Water Utilities in the San Jose Bay Area Presenter: John Eidinger (G&E Engineering Systems Inc., Oakland, California) Government Policies on Earthquake Disaster Prevention Measures Related to Water Supply in Japan Presenter: Yoshihisa Iwasaki (Health Service Bureau, Ministry of Health, Labor and Welfare, Tokyo, Japan) Anti-Seismic Measures of Existing Water Supply Facilities - A Case Study of an Anti-Seismic Plan of Inagawa Water Treatment Plant Presenter: Keiichi Murakami (Hanshin Water Supply Authority, Kobe, Japan) San Francisco Public Utilities Commission Capital Improvement Program Presenter: Jeet Bajwa (San Francisco Public Utilities Commission) Seismic Measures for Waterworks in Yokosuka City Presenter: Takashi Furuya (Yokosuka City Waterworks Bureau, Japan) 15

34 3 rd US-Japan Workshop on Water System Seismic Practices 16

35 Seismic Upgrades for 20 Suburban Water Utilities in the San Jose Bay Area John Eidinger 1 1 Introduction The San Jose Bay Area (formerly the San Francisco Bay Area) has a metropolitan population of more than 7,000,000 people. The Hetch Hetchy water transmission system delivers water to 30 separate water systems, one of which is the city of San Francisco, and 29 of which are collectively known as the "suburban customers". Figure 1 shows a map of the area, highlight the different regions, and also suggesting the basis for the current $3,500,000,000 seismic and reliability retrofit program being planned and implemented by the San Francisco Public Utilities Commission (SFPUC). Figure 1. Water Systems in the San Jose (Francisco) Bay Area Figure 1 highlights a few issues: 1 G&E Engineering Systems Inc., 6315 Swainland Rd, Oakland, CA 94611; [email protected] 17

36 o Most of the people (1,700,0000 people) served by the SFPUC's Hetch Hetchy water system are in the suburban customer water systems. The City of San Francisco (800,000 people) is the largest individual "customer" of the Hetch Hetchy system. About 70% of the total cost to operate and maintain the Hetch Hetchy water system is paid for by the people in the 29 suburban water systems. o In July, 2003, the United States government re-named the (formerly) San Francisco Bay Area to be now called the San Jose Bay Area. Funny sounding as this might be, this new name reflects that since 1990, the greatest population and economic activity is in the south bay area, dominated by the "Silicon Valley" cities of San Jose, Santa Clara, Mountain View, Sunnyvale, Cupertino, etc. This is where many of the "great technology " companies like Intel, Hewlett Packard, AMD, Apple, etc. are headquartered and located. o The three colors used to highlight individual cities (yellow, orange, red) indicate where the SFPUC "thinks" ("thought") where the its greatest vulnerability for water supply disruption might occur. This map was prepared using the best available information from 1999; since then, a lot of new thinking has suggested a new pattern of water outages, possibly not so severe, especially when considering alternative water supply sources, and more modern understanding of pipeline and tunnel fragilities. o The green lines represent the major pipeline and tunnel corridors of the Hetch Hetchy system. The blue lines show the just a few of the major pipelines of two nearby water utilities, EBMUD and SCVWD, and highlight where current and planned regional connections between these three large water systems might serve to increase reliability of water supply for the bay area. Recognizing that earthquakes can seriously damage the Hetch Hetchy water transmission system, and that loss of water supply can cause great economic damage to Silicon Valley and the other communities of the Peninsula and South Bay, this paper summarizes the seismic risks and economic of water system mitigation for 20 of the largest cities served by the Hetch Hetchy water transmission system. 2 Economics of the Seismic Upgrade of the Hetch Hetchy Water System From 2001 to 2003, 20 of the largest suburban customers have conducted seismic vulnerability studies of their own water systems. I found that when one approaches the typical water utility owner, and asked them if "it is worth it to upgrade their water system for earthquakes?", one will most often get one of the following five stages of response: 1. I don't have a problem 2. I did not know I had a problem 3. I sense that there might be some type of problem, but I don t know how to quantify it 18

37 4. I am pretty sure I have a problem, so I will take a shotgun approach and fix / improve as many parts of the system as my (regulators / city council / rate payers) are willing to pay for 5. I know I have a problem, so I will study it and develop a rational and cost effective approach to address it. In high seismic regions like Coastal California, the author has experience with various water utilities that have provided all of these five stages of response. Some of the larger water utilities serving a million or more people (like the City of San Diego, the Santa Clara Valley Water District, the East Bay Municipal Utility District) have adopted approaches consistent with response 5. Many other water utilities, serving populations from 15,000 people to millions of people have adopted any or all of responses 1, 2, 3 and or 4, with the result that some utilities are spending too little and some are spending too much. One intriguing example is currently taking shape: the seismic and reliability upgrade of the aging SFPUC Hetch Hetchy water system. The Hetch Hetchy system is a water transmission system delivering water from Yosemite National Park (and a few other local supply sources) to about 2,500,000 people in the San Francisco Bay Area. These 2,500,000 people are served by 30 separate water distribution systems, the largest of which (800,000 people) is the City of San Francisco's own distribution system. Ownership, operation and maintenance of the Hetch Hetchy system is by the San Francisco Public Utilities Commission (SFPUC). The remaining 29 water distribution systems (the so-called "suburban customers") purchase water from the SFPUC, and pay for about 70% of the cost to operate, maintain and upgrade the Hetch Hetchy system. At times, the wishes of the 29 suburban customers do not line up exactly with the wishes of the SFPUC. Since the late 1990s, the SFPUC has been studying seismic and other reliability aspects of the Hetch Hetchy system. In January 2000, the SFPUC completed their "SFPUC Facilities Reliability Program" (2000). This effort simulated the overall SFPUC water system reliability in the event of a major earthquake on the San Andreas, Hayward, Calaveras or Great Valley faults. The effort reportedly used the "most current understanding of effects of infrastructure from ground shaking, fault crossing and liquefaction". The analyses resulted in a recommended program of seismic improvements to increase overall SFPUC system reliability. The overall cost of this program was estimated at $3.5 Billion, of which $1.3 Billion was for seismic improvements, and the remainder for reliability improvements. These amounts include no funds to make improvements in the San Francisco City and 29 suburban distribution systems. 19

38 Figure 2. Damage to the SFPUC (Spring Valley Water Company) Transmission System, 1906 Figure 2 shows a map of the SFPUC transmission system as it existed in 1906 and the damage it suffered in the 1906 Great San Francisco earthquake. The modern (year 2003) SFPUC transmission system has about 3 times as many pipelines, many of which follow similar alignments as the pipelines did in 1906, except that newer pipes bypass the marshy area marked by the number "4" in Figure 2. It remains unclear as of 2003 as to exactly how much the 29 suburban agencies are happy to pay for this program. The final cost of the Hetch Hetchy system seismic reliability upgrades will roughly triple the cost to purchase SFPUC water. In July, 2003, one suburban customer said: "my customers like pure Hetch Hetchy water. But, if the cost of Hetch Hetchy water increases to more than $100/acre foot more than what it costs to use SCVWD water, then I am (almost) sure that my customers will want me to drill more wells and buy more water from SCVWD." 20

39 3 Just How Unreliable is the Hetch Hetchy System? Using modern concepts of the seismic vulnerability of water transmission systems (Eidinger and Avila 1999, Eidinger 2001), a risk analysis of the Hetch Hetchy system was performed. Earthquake hazards of ground shaking, liquefaction, landslide and surface faulting were applied to 225 miles of large diameter (mostly 60" to 96" diameter) pipelines and 6 major and 15 minor tunnels. Figure 3 presents the post-earthquake reliability of the Hetch Hetchy system serving the south bay area, assuming that an upgrade of BDPL 3 and 4 pipelines is done to mitigate the hazard where they cross the Hayward fault, and that liquefaction hazards are largely mitigated at a few creek crossings. Total cost of such upgrades is likely less than $100,000,000. Figure 3. Reliability in the South Bay Area, After Hayward M 7.1 Earthquake A few points are made about Figure 3. First, the base map (in blue) represents all the major pipelines and reservoirs of the Hetch Hetchy water in the San Jose Bay Area. Second, the BDPL 4 pipeline (96" prestressed concrete cylinder pipe) has about a 96% chance of suffering none or only slight leaks, given a Hayward M 7.1 earthquake. Third, the BDPL 3 pipeline (78" steel pipe) has about a 85% chance of suffering none or only slight leaks, given a Hayward M 7.1 earthquake. Fourth, by using the existing cross connections between the BDPL 3 and 4 pipelines, and allowing that in-line isolation valves can be turned (if needed) within (no more than) 24 hours after the earthquake, then there is a 95% chance of delivering adequate water supply to all the south bay customers after a Hayward M 7.1 earthquake. In the current condition (pre-mitigation), there is only a 43% chance of delivering adequate water supply to all south bay customers after a Hayward M 7.1 earthquake. 21

40 4 What About the Suburban Customers? With large potential rate increases facing the suburban customers of the SFPUC, the level of awareness about seismic issues has risen from "about" stages 1 or 2, and most are now thinking about responses at stages 4 or 5. From 2001 to 2003, a series of seismic vulnerability analyses have been performed for 20 of the suburban customers. Item Amount Note Average Day Demand 206 MGD 76% of total Hetch Hetchy demand Number of Pump Stations 157 Number of Storage Tanks 202 Miles of Distribution System Pipelines 3,912 Mostly 4" to 27" pipe Wells 90 Treatment Plants 6 Emergency Generators 63 Pipe Repairs, San Andreas M 7.9 Earthquake 1,190 to 3,030 Lower value is more likely Pipe Repairs, Hayward M 7.1 Earthquake 920 to 2,580 Lower value is more likely Seismic Improvement Program $28 to $50 million Table 4. Statistics of 20 Suburban Customer Water Systems The 20 suburban customers that have had seismic vulnerability analyses performed (Hayward, Alameda County Water District, City of Santa Clara, Mountain View, Purissima Hills, Palo Alto, Stanford University, Bear Gulch, Redwood City, San Carlos, San Mateo, Foster City, Coastside County, Mid-Peninsula, Burlingame, South San Francisco, Brisbane, Daly City, Menlo Park, San Bruno) represent about 80% of the total suburban customer demand. Table 4 provides some overall statistics for these 20 suburban customers. The modern Hetch Hetchy water system has about 220 miles of large diameter (mostly 60" to 96" diameter) pipelines within the greater San Jose Bay Area. In consideration of faulting, liquefaction, landslide and ground shaking, these pipes are expected to suffer between 16 and 23 repairs following Hayward M 7.1 and San Andreas M 7.9 earthquakes, respectively. The bulk of these repairs will likely manifest themselves as leaks at air valves or blow offs, but a few full breaks are likely at fault crossings, creek crossings or at unexpected locations. There is even a chance that a major tunnel might collapse. With available in-house repair crews, the SFPUC might be able to patch up the major breaks in 4 to 12 days, and repair all leaks within 1 to 2 months. If the unlikely but not impossible event that a major tunnel should collapse, repairs of the tunnel could last months, in the meantime the water supplies might have to be restricted to no more than about 80% of maximum winter time demands. Given these scenarios, the following seismic improvements have been proposed: 22

41 o $28 to $50 million of seismic improvements within the 20 suburban customer distribution systems. o $1.3 to $3.5 billion of seismic and reliability improvements within the Hetch Hetchy transmission system. As of mid-2003, there remains much work to coordinate the overall transmission / distribution seismic upgrade programs. For example, should a small suburban customer invest $800,000 to construct a well, thereby providing an alternate source of water should all Hetch Hetchy water be lost for days to weeks after a major earthquake? And if that small suburban customer builds that well, should it also accept the allocated cost to improve the major water pipeline transmission system? What might be most cost effective for that one suburban customer might not be the most cost-effective for other suburban customers, or for the SFPUC as a whole, and this brings up difficult political and policy issues. Item EBMUD SFPUC + 20 Suburban Customers Miles of Transmission Pipelines Miles of Distribution Pipelines 3,900 3,700 Tunnels Treatment Plants 6 8 Storage Tanks Pump Stations Small Pipes that cross major active faults ( 18" diameter) Large Pipes that cross major active faults ( 20" diameter) Tunnels that cross major active faults 2 0 Pipe Repairs, Loma Prieta M < 400 Pipe Repairs, San Andreas M 7.9 < 1,000 1,190 to 3,030 Pipe Repairs, Hayward M 7.1 3,300 to 5, to 2,580 Seismic Upgrade, Transmission System $140 million $1,300 million Seismic Upgrade, Distribution System $100 million $28 to $50 million Seismic Improvements, Total $240 million $1,328 to $1,350 mil. Ratio, Distribution to Total 42% 2% to 4% Population served 1,200,000 2,500,000 Cost per person $200 $555 Table 5. EBMUD and SFPUC / Suburban Customer Cost Allocation To provide some insight to these issues, one can examine the allocation of seismic upgrade cost made by EBMUD in their $240,000,000 seismic upgrade program. EBMUD is a utility that owns and operates both a raw water transmission as well as a large potable water distribution system. For EBMUD's case, if one sums up all costs associated with raw and treated water pipelines of 36" diameter and larger (cumulatively, the "transmission system"), EBMUD has spent about $140,000,000 on transmission upgrades. The remaining $100,000,000 was allocated to upgrades of smaller diameter 23

42 pipelines (generally 12" to 30" diameter), water treatment plants, pump stations, storage tanks and emergency response. Table 5 highlights the differences in upgrade costs between EBMUD (actual) and SFPUC / Suburban customers (projected). The age of infrastructure in the EBMUD and SFPUC transmission systems is quite similar. The original EBMUD transmission pipelines and tunnels were put into service in 1929 (Mokelumne 1, Claremont Tunnel); the original Hetch Hetchy pipelines and tunnels were put into service in 1923 to 1933 (BDPL 1 and 2, Coast Range Tunnel). EBMUD's first major transmission pipeline system upgrade was put in service in ~1948 (Mokelumne 2); similar for Hetch Hetchy (BDPL 3). EBMUD's most recent major transmission pipeline system upgrade was put in service in ~1965 (Mokelumne 3); similar for Hetch Hetchy (BDPL 4). 5 Economic Impacts to Suburban Customers A series of seismic vulnerability analyses were performed for 18 water distribution systems that are served by the Hetch Hetchy transmission system. These 18 systems have a combined average day demand of 228 MGD, and serve a population (year 2020) of 1,419,000 people. Allowing for 20 to 30 day outages from the SFPUC transmission system (probably upper bound, more likely 4 to 12 days), and a variable amount of impacts to the local distribution systems (pipe repairs, damaged tanks, failed wells, power outages, etc.), and using the Fire Ignition and Spread models by Eidinger (1996) or the more comprehensive fire ignitions and spread models by Scawthorn, Eidinger and Schiff (in press 2003), the following statistics (medians only) are developed: Item San Andreas M 7.9 Hayward M 7.1 Economic Losses, Year $2003 $1.4 to $1.6 billion $250 to $610 million Fire Ignitions Fire Losses, Calm Winds $85 to $142 million $65 to $110 million Fire Losses, Light Winds $200 to $342 million $153 to $262 million Fire Losses, High Winds $1.1 to $1.4 billion $0.9 to $1.1 billion Table 6. Impacts to 18 Distribution Systems in Scenario Earthquakes (As Is System) Item San Andreas M 7.9 Hayward M 7.1 Economic Losses $93 to $333 million $127 to $535 million Fire Losses, Calm Winds $14 to $57 million $11 to $44 million Fire Losses, Light Winds $85 to $114 million $66 to $88 million Fire Losses, High Winds $1.0 to $1.1 billion $0.8 to $0.9 billion Table 7. Impacts to 18 Distribution Systems in Scenario Earthquakes (Upgraded System) The "upgraded system" evaluation is performed for the same 18 distribution systems, but this time with the assumption that seismic upgrades are in place to reliably assure that no more than a 24 hour outage of delivery of maximum winter time demand rate water from the Hetch Hetchy transmission system to each distribution system. By comparing the difference in losses (economic and fire) from Tables 6 and 7, we can estimate the net benefit (scenario earthquake basis) of the retrofit program. Using the 24

43 midpoint values, and assuming the light wind scenario, the net reduction in losses (i.e., the benefit) is: Item San Andreas M 7.9 Hayward M 7.1 Benefit, Economic Impacts $1,250 million $99 million Benefit, Fire Impacts $172 million $131 million Benefit, Other Impacts $200 million $40 million Total Benefit (Scenario Based) $1,622 million $270 million Table 8. Net Benefits of Seismic Upgrade, Scenario Based Allowing that there is about a 1% chance of occurrence of either of these two or similar scenario earthquakes (San Andreas M 6.8 to 7.9 event that includes the Peninsula fault segment, Hayward M 6.8 to 7.3 event that includes the southern Hayward fault segment), and allowing for other earthquakes on other faults and for smaller earthquakes, and assuming a 5.5% discount rate, and using the benefit cost model for water systems outlined in (Eidinger and Avila, 1999), the net present value of the benefits of seismic upgrades are calculated as follows (all monetary values in millions, year $2003): o San Andreas M 6.8 M 7.9: $1,622. Annual chance: Annual benefit: $16.22 o Hayward M 6.8 M 7.3: $270. Annual chance: Annual benefit: $2.7 o Calaveras, Rodgers Creek, Great Valley, background and smaller earthquakes: Cumulative annual benefit = $9.7 o Total annual benefit over all faults, all magnitudes = $28.6 o Net present value of benefits, 5.5% discount rate, 100 year project life = $28.6 x 18.1 (NPV factor) = $518 In other words, the rate payers of the 18 distribution systems should be willing to pay, in year 2003 dollars, up to about $518,000,000 to seismically retrofit the Hetch Hetchy water system to the point where it can reliably restore water to each system within 24 hours after any earthquake, at maximum winter demand rate or higher. 6 Conclusions and Observations Seismic vulnerability analyses have been performed for 20 of the SFPUC's suburban customers. Cost-beneficial seismic upgrades of about $28,000,000 to $50,000,000 have been identified for these 20 water systems. A comparison is made between the (almost completed) EBMUD seismic upgrade program and the (recently started) SFPUC seismic upgrade program. While there are a number of similarities between the age and quantity of infrastructure between of the two sets of water systems, the cost of the programs is quite different, as well as the ratio of cost between distribution and transmission upgrades. By performing seismic vulnerability analyses for 18 suburban distribution systems served by the SFPUC Hetch Hetchy system, and then performing economic analyses as to the value of seismic upgrades, this paper suggests that these suburban customers should be 25

44 willing to pay up to about $518,000,000 to achieve a no-more than one-day outage of the Hetch Hetchy transmission system after any earthquake. Retrofits and improvements beyond this cost could be justified for non-seismic reliability issues. 7 Units and Abbreviations All monetary values are in year 2003 U.S. dollars, except as noted. BDPL = Bay Division Pipeline EBMUD = East Bay Municipal Utility District Inches (") = 25.4 millimeters M = moment magnitude Miles = kilometers MGD = Million Gallons per Day (US liquid measure). 1 MGD = 43.8 liters per second NPV = Net Present Value SCVWD = Santa Clara Valley Water District SFPUC = San Francisco Public Utilities Commission TCLEE = Technical Council on Lifeline Earthquake Engineering UBC = Uniform Building Code 8 References Eidinger, J., Seismic Fragility Formulations for Water Systems, G&E Report No Revision 1, July 12, 2001 ( and American Lifelines Alliance, G&E Report No revision 0, March 27, Eidinger, J., and Avila, E., Eds., Guidelines for the Seismic Upgrade of Water Transmission Facilities, ASCE, TCLEE Monograph No. 15, January Eidinger, J., Lifeline Considerations and Fire Potential, in Seismic Safety Manual, D. Eagling editor, Lawrence Livermore National Laboratory, September Scawthorn, C., Eidinger, J., and Schiff, A., editors, Fire Following Earthquake, in press, ASCE, AWWA, NFPA, SFPUC Facilities Reliability Program, Phase II Regional System Overview, Final Report, CH2M-Hill, Olivia Chen Consultants, Montgomery Watson, EQE International, January

45 Government Policies on Earthquake Disaster Prevention Measures related to Water Supply in Japan Yoshihisa Iwasaki * ABSTRACT Japan is prone to earthquakes. Japanese water services have therefore had to overcome many, repeated occurrences of damage to facilities caused by major earthquakes, such as the Niigata Earthquake (1964), the Central Japan Sea Earthquake (1983) and the Southern Hyogo Prefecture Earthquake (1995). The current situation is that 96.7% of the Japanese population has access to a water supply system, which has come to be seen as indispensable for daily life. Once the water supply ceases to function, this therefore has serious consequences for people s lives. The Southern Hyogo Prefecture Earthquake in 1995, for example, caused extensive damage to filtration plants, transmission mains and distribution pips. These left approximately 1.3 million households without water supply with a maximum water failure period of as long as ninety days. It was a major, prolonged disaster. Drawing on the lessons of these experiences, the Japanese government has been working out its basic policies on how water suppliers should cope with earthquake disasters. At the same time, it has enhanced the national subsidy system for earthquake disaster prevention measures. This article summarizes Japan s earthquake disaster prevention measures related to water supply facilities, the state of progress, future issues and prospects. I. DAMAGE TO WATER SUPPLY SYSTEMS CAUSED BY EARTHQUAKES IN THE PAST Japan is prone to earthquakes. Japanese water services have therefore had to overcome many, repeated incidences of damage to facilities caused by major earthquakes, such as the Niigata Earthquake (1964), the Central Japan Sea Earthquake (1983) and the Southern Hyogo Prefecture Earthquake (1995) (Table 1). The 1995 Southern Hyogo Prefecture Earthquake, for example, inflicted massive damage on filtration plants, transmission mains and distribution pipes. This left approximately 1.3 million households without water supply with a maximum water failure period of as long as ninety days. It was a major, prolonged disaster (Photos 1 and 2). Besides damage from earthquakes, Japanese water supply systems have also suffered from volcanic eruptions. There are various forms of damage caused by earthquakes. At major facilities such as filtration plants and distribution reservoirs, the typical problems observed are the fracture of expansion joints and leakage at connecting points between structures and pipes. Damage to pipelines includes pipe fractures and cracks due to the use of materials that are less resistant than ductile cast-iron pipe, and the separation of joints that do not have an expansion and restraint function. * Assistant Director, Water Supply Division, Health Service Bureau, Ministry of Health, Labour and Welfare, 1-2-2, Kasumigaseki, Chiyoda, Tokyo, JAPAN

46 Photo 1 Separation of pipes caused by the Southern Photo 2 Emergency water services after the Southern Hyogo Prefecture Earthquake Hyogo Prefecture Earthquake Table 1 Damage caused by major earthquakes Location of earthquake Date of occurrence Magnitude and Scale* Cost of the damage ( million) Number of days of water supply failure** Number of households without water supply Niigata June 16, 1964 M7.5/Scale 5 2,100-55,000 Central Japan Sea May 26, 1983 M7.7/Scale ,321 East off Chiba Prefecture Dec. 17, 1987 M6.7/Scale ,657 Earthquake off Kushiro Jan. 15, 1993 M7.8/Scale ,093 Southwest off Hokkaido July 12, 1993 M7.8/Scale ,907 East off Hokkaido Oct. 4, 1994 M8.2/Scale ,462 Distant off Sanriku Dec. 28, 1994 M7.6/Scale (Approx. 120,000) Southern Hyogo Prefecture Jan. 17, 1995 M7.3/Scale 7 60, Approx. 1.3 million Western Tottori Prefecture Oct. 6, 2000 M7.3/Scale ,300 Geiyo Islands Mar. 24, 2001 M7.3/Scale ,500 South Sanriku May 26, 2003 M7.3/Scale ,792 * On the Japanese scale of seven ** Maximum in the stricken area For Aomori Prefecture only, including damage from aftershocks For Hyogo Prefecture only Including days when tap water was not suitable for drinking purposes II. EARTHQUAKE DISASTER PREVENTION MEASURES FOR WATER SUPPLY SYSTEMS AND GOVERNMENTAL SUPPORT IN JAPAN 1. Earthquake disaster prevention measures in the FRESH Waterworks Plan The Ministry of Health and Welfare (now the Ministry of Health, Labour and Welfare), with the aim of establishing a high-quality water supply system for the 21st century, consulted the Living Environment Commission concerning the main elements of water service policy. In 1990 a proposal entitled Measures for Improving the Quality of the Water Supply System was submitted. Based on this proposal, the Ministry drew up Long-term Goals for Water Supply Development towards the 21st Century, in The plan also goes by the name of the FRESH Waterworks Plan. This long-term plan set the goal of providing a safe and secure water supply, and aims to develop a water supply system that is less susceptible to drought, earthquakes and other natural disasters. More specifically, the Plan determined that the safety of the water system as a whole should be improved through the replacement of decrepit facilities as well as the renovation of the main facilities to ensure that they are earthquake resistant. In addition, in 28

47 order to secure the basis for emergency water supply, the Plan urged that the capacity of distribution reservoirs should be more than the 12 hours of the design maximum daily supply, and that the installation of emergency tanks should be accelerated. These basic directions are still in effect, and have been carried over into the recent measures to be mentioned below. 2. Principles of earthquake disaster prevention measures for water supply systems -- Towards an Earthquake-resistant Water Supply Service The lessons from the extensive damage due to the 1995 Southern Hyogo Prefecture Earthquake taught us that water supply systems are vital in urban areas where there is no other available means of obtaining clean water. Even when the systems are damaged, minimum functionality must be maintained, otherwise it becomes very difficult to acquire water for drinking, for medical care, or for firefighting, which is indispensable in sustaining the lives of the earthquake victims. Ensuring that water supply systems are earthquake resistant was recognized as one of the most urgent tasks. In accordance with this conclusion, the Ministry established in June 1995 a panel of academic experts and intellectuals to assess the possibilities for the establishment of an aseismic water supply system. The panel examined strategies in terms of both hardware and software to establish a water supply system that is earthquake proof, and published a report entitled Towards an Earthquake-resistant Water Supply Service in August of the same year. The principles proposed in this report are introduced below. The Ministry has promoted these principles by upgrading the national subsidy system and by facilitating studies on aseismic engineering. Principles proposed in Towards an Earthquake-resistant Water Supply Service (1) Comprehensive earthquake disaster prevention measures Earthquake disaster prevention measures in relation to the water supply system require an integrated approach. This includes the development of systematic emergency procedures in addition to increasing the resistance of the water facilities to earthquakes. Achieving the required high aseismic standards for water supply facilities is essential to the upgrading of water supply systems so that they can respond flexibly to changes in society. It is also a responsibility of the current generation towards future generations. (2) Efficient operation of emergency procedures Assurance that emergency water supplies will be available and rapid recovery from failure requires well-developed systems and conditions. Since there are many small-scale water systems throughout the country, prefectural governments are required to utilize their capacity to develop a broader-based emergency system while consulting with large-scale water utilities and bulk water suppliers within their jurisdiction. In the case of major earthquakes, the national government should exercise its leadership through close liaison with the prefectures in order to ensure the smooth operation of broader-based emergency measures. 29

48 (3) Upgrading of seismic safety in relation to water facilities Improvements to water facilities require a major investment over a long period of time. Such improvements should be carried out effectively and efficiently, in a focused and planned manner, with the decision-making based on scientific evidence (e.g. a seismic analysis of the facilities). (4) Financial measures in relation to earthquake preparation Expenditures on water facility improvements should, in principle, be born by the water utilities and bulk water suppliers. Without financial support, however, there is no guarantee that the improvements will be sufficient. Such investments are expected to have long-term benefits for future generations. Cost cutting with regard to preventive measures today may result in a broader scale of disasters in the future. It is also necessary to reserve funds to prepare for the complete loss of water rate revenues and the considerable expenditures related to emergency operations. 3. Guidelines on Planning for Aseismic Water Supply Systems (draft) The policy recommendations Towards an Earthquake-resistant Water Supply Service establish the actual steps to be taken. The Ministry of Health, Labour and Welfare, on the basis of this, developed the Guidelines on Planning for Aseismic Water Supply Systems (draft) so that water utilities and bulk water suppliers nationwide can refer to these guidelines in their efforts to promote their own earthquake disaster prevention measures in accordance with local conditions. The Guidelines (draft) cover guidelines for water supply systems in general as well as small-scale water supply systems (Figures 1-1 and 1-2). The former provide the basic steps for the establishment of an effective and efficient plan while combining various measures according to geographical characteristics. The latter refers to specific policies developed in relation to small-scale systems. All measures in the Guidelines are classified and organized in terms of before-the-fact (aseismic facilities) and after-the-fact (emergency operations) activities (Figures 2 and 3). 30

49 (1) Water system damage estimation Seismic diagnosis of structures and equipment Estimation of damage to pipelines Estimation of damage to the entire system (3) Specific approaches to improvements (optional) Impact of the failure of water supplies on the public (2) Target setting for aseismic improvement Targets for aseismic improvements (a) Shorter restoration period (b) Emergency water supply (c) Maintenance of functional capacities Indicators of aseismic improvements Damage assessment of the failure of water supplies (a) Population without water supplies (b) Duration of the failure and restoration Facility improvement Emergency operations Damage control (a) Water sources and structures (b) Pipelines (c) Distribution facilities Damage minimization (a) Pipeline systems (b) Installation and operation of valves (c) Prevention of secondary disaster impacts Prompt restoration (a) Information gathering (b) Recovery work Improved emergency supply (a) Mobile water supply (b) Facility-based supply (4) Development of an aseismic improvement plan Choice of approaches (a) Consideration of aseismic targets (b) Consideration of local characteristics (c) Priority setting Improvement of structures (a) From major water resources to filtration plants (b) Distribution reservoirs Installation of aseismic pipelines (a) Where to supply emergency water (b) Aseismic routes (c) Connection pipelines Development of distribution networks (a) Separation of damaged facilities (b) Supply bases blocks (c) Restoration on a block basis Establishment of emergency supply bases (a) Assurance of the storage volume (b) Establishment of bases (5) Establishment of plans Disaster management plan Urban development plan Evaluation of the effectiveness of aseismic improvement (a) Direct impact on the public (b) Cost effectiveness (c) Use for purposes other than earthquakes Expenses for improvement Action plans for earthquake disaster prevention projects (a) Target date (b) Establishment of action plans Figure 1-1 Contents of the Guidelines on Planning for Aseismic Water Supply Systems (draft) 31

50 (1) Principles of the aseismic improvement of small-scale water systems Characteristics of small-scale systems (a) Natural conditions (b) Social conditions (c) Operational conditions Principles for the aseismic improvement of small-scale water systems (a) Types of aseismic pipes (b) Ensuring water supply for restoration work (c) Consideration for local industries (d) Promotion of wider access and shared systems Special conditions to be reflected in local specific aseismic improvements (2) Islands and mountainous areas (3) Areas whose main water source is groundwater (4) Areas where the main facilities rely on a single water system (5) Areas where demand points have a scattered distribution (6) Areas where there is a low level for the mobilization of capacity (e.g. staff) Figure 1-2 Outline of the Guidelines on Planning for Aseismic Water Supply Systems (draft) Control of damage Aseismic improvement at water sources Aseismic improvement of pipeline facilities Water source facilities Structures Machinery and electrical equipment Pipelines Pipeline accessory equipment Special pipeline configurations Earthquake disaster prevention measures Aseismic improvements to water supply facilities Emergency operations Minimization of damage Prompt restoration of water supply systems Improvement of emergency water supply systems Aseismic improvement of supply equipment Aseismic improvement of the construction process Aseismic improvements to the pipeline system Removal of obstructions Installation and operation of valves Prevention of secondary disasters Information management Speeding up of the restoration operations Ensuring adequate personnel and the proper reception of relief teams Mobile water supply systems Facility-based water supply systems Temporary water supply systems Water supply equipment On-site equipment Issues arising during the process of construction Development of backup systems Development of block systems Development of loop systems Removal of obstructions to the work Installation of valves Operation of valves Assurance of water supply for fire fighting Prevention of landslides Prevention of water pollution Identification of the damaged parts of the system Public information activities Emergency procedures Operation of emergency restoration work (Ensuring water supply for restoration work) Restoration of block systems Ensuring adequate personnel Reception of relief teams Standardization of equipment Ensuring the availability of water supplies Storage of materials Emergency water supply systems Location of facility-based water supply systems Installation of distribution pipes Aseismic storage tanks Water supply to evacuation areas Water supply to temporary accommodations Water supply to medical facilities Figure 2 Systematization of earthquake disaster prevention measures 32

51 Assessment and reinforcement of structures Reinforcement of equipment Filtration plants Conceptual scheme for the improvement of facilities for aseismic water systems Receiving from bulk water suppliers Connection pipes between adjacent utilities Development of multi-channel water transmission Emergency cutoff valve (security of water supply) (backup systems) Installation of large-capacity distribution reservoirs (security of water for restoration operations) Distribution reservoirs Remote supervisory control using TM/TC Diversification of communications Private power generation Multiple channels for power delivery Filtration plants Prevention of pollution at the water source Assessment and reinforcement of embankments (prevention of secondary disasters) Proper spacing between control valves Water supply base Connection to other water systems (backup systems) Replacement of asbestos cement and cast-iron pipes Reinforcement of pipeline accessory equipment Proper location of control valves Installation of monitoring devices Aseismic receiving tank Aseismic improvement of the mains systems Hospitals Aseismic pipes Replacement using aseismic pipes Measures to address soft ground Water supply base Expansion and flexible joint Distribution block system Aseismic pipes Establishment of the water supply base Schools Looping of the distribution mains Aseismic water supply equipment Change in the location of meters and stop valves Figure 3 Concepts of facility improvement for aseismic water systems 4. Outline of the national subsidy system for earthquake disaster prevention measures The following is the list of projects eligible for national subsidies. It includes projects for the aseismic upgrading of water pipelines and for the provision of emergency reservoirs. Projects for the Enhancement of Lifeline Functions (as regards water supply services) (1) Projects for the establishment of emergency water supply bases (To be subsidized up to 10,643 million yen, or one third of the actual costs) Enlargement of the capacity of distribution reservoirs (Installation of reservoirs whose capacity is estimated at more than eight hours worth of the design maximum daily supply) Connection pipelines for emergency use (Mutual support for water supply services among neighboring water utilities in case of emergencies) Water storage facilities (Transmission and distribution pipes with storage function, including high-capacity transmission mains) Emergency cut-off valves (Valves that can be closed in the case of emergency to prevent the flow of water from reservoirs) 33

52 (2) Projects for increasing the seismic resistance of main pipelines (To be subsidized up to 127 million yen, or one half of the actual costs) Reinforcement of existing pipeline networks, not depending solely on post-disaster restoration (3) Projects for modernizing water pipelines (To be subsidized up to 4,200 million yen, or one quarter or one third of actual costs) Renewal of aging asbestos cement pipes as a part of earthquake damage prevention measures * The amounts in parentheses are based on the original budgets for the 2003 fiscal year (National expenditure basis) 5. Engineering development and research for earthquake preparation In the light of the above-mentioned report Toward an Earthquake-resistant Water Supply Service, the Ministry promoted several engineering development and research projects as joint efforts between industry, academia and government, from the 1996 to 2001 fiscal year. These particularly place a focus on technologies for damage identification and damage simulation among items considered to be under infrastructure development. The following is the outline of those projects. (1) Engineering development project for damage prediction and identification ( ) For the purpose of providing a solid foundation for the planning of facility improvements, this project developed the technology to scientifically predict facility damage. It also developed the technology to detect promptly and precisely pipeline damage caused by earthquakes. 1) Simulation technology to predict damage to water supply systems This technology included the creation of databases, establishment of formulae and the development of a computer system for damage prediction in relation to water pipelines. With this technology, water suppliers are expected to promote effective facility improvement and to prepare appropriate restoration plans. 2) Detection technology to identify damage to water pipelines This technology was developed for the effective and efficient detection of damaged sections in the case of emergency, and could be used when the pipe is empty. The adaptability of the technology was examined. (2) System development project for assistance to the post-quake restoration of water supply systems ( ) When water supply systems are affected by an earthquake, water utilities must promptly initiate emergency repairs and water supply. This project is intended to propose an approach to the planning of such emergency procedures. The developed systems can reasonably predict the number of people without water supply based on the simulated damage, and allow for planning the most effective and efficient emergency procedures. This system consisted of the following tools. 1) Severity prediction system based on damage anticipation Utilizing the results of prediction in relation to earthquake damage, this system could assess logically the number of people to whom water is not supplied, and could simulate suitable emergency repair and water supply activities. 2) Prompt damage detection system This system established a technology to detect degraded and leaked pipelines, which could be applied whether the pipe is empty or full, and could automatically detect 34

53 leakages and interruptions in pipelines. 3) Support system for post-quake restoration The system provides support when water suppliers develop disaster prevention plans or crisis-management systems. 6. Current status of earthquake disaster prevention measures (1) Aseismic water pipelines In spite of the extensive efforts of water suppliers, recent research (Figure 4) shows that only 30 to 40% of the mains supply system for raw water and treated water are considered earthquake proof. Coupled with the aging of facilities, the vulnerability of water supply systems to earthquakes has tended to increase overall. Propotion of aseismic pipelines 45% 45% Proportion of aged pipelines 40% 35% 33% 35% 30% 25% 20% 18% 29% 30% 23% 23% 15% 10% 5% 0% Raw water mains Transm ission mains Distribution pipes Average Figure 4 Status of aseismic and aged pipelines * Data relates to water utilities and bulk water suppliers nationwide as of March 31, Proportion of aseismic pipelines: Proportion of aged pipelines: The length of ductile cast-iron pipes with seismic joints plus one half of the length of ductile cast-iron pipes and steel pipes as a proportion of the total length of all types of pipes The length of asbestos-cement pipes and lead pipes plus the length of other pipes that were laid twenty or more years ago as a proportion of the total length of all types of pipes (2) Mutual support between neighboring water suppliers (backup systems) Water utilities under national jurisdiction were surveyed as to whether they have a connection pipe to any adjoining utilities that allows for the joint provision of water. The result showed that only about 20% of water utilities had such a connection. For 80% of those who do have a connection pipe, the backup ratio was below 50% (Figure 5). This indicates that backup systems are still in the process of being developed. 35

54 30 25 No. of facili Less than 10% 10-20% 20-30% 30-40% 40-50% 50-60% 60-70% 70-80% 80-90% 90%- Figure 5 Proportion of water supplies jointly provided among the suppliers (backup ratio) * Data relates to water utilities and bulk water suppliers under the jurisdiction of the national government as of March 31, Backup ratio: The ratio of the capacity of the connection pipes to the design supply of the relevant utility (3) Capacity of the distribution reservoirs (storage hours) The national average of the capacity of distribution reservoirs that is to be used as the basis for the water supply in case of emergencies has not achieved the target, which is an amount of supply sufficient for twelve hours. If present trends continue, it will take about ten years to achieve the target nationwide. Storage hours (h Past 実 績 records Estimation 推 定 Fiscalyear Figure 6 Trend in the capacity of distribution reservoirs * Data relates to water utilities and bulk water suppliers nationwide as of March 31, Storage hours: Effective capacity of distribution reservoirs divided by designed maximum daily supply times 24 hours 36

55 III. FUTURE ISSUES AND PROSPECTS 1. Preparation for anticipated major earthquakes Japan foresees several major earthquakes in the not-so-distant future, including earthquakes in the Tokai, Tonankai and Nankai regions (Table 2). The estimated water system damage from a Tonankai/Nankai earthquake, for example, is such that it would cut water supplies to 14 million people. In addition, due to extreme traffic congestion, restoration will be significantly delayed and the suffering of people without water will be prolonged. Table 2 Estimated number of people without water supply after a major earthquake (Unit: millions) Tokai earthquake Tonankai/Nankai earthquake On the day One day after Two days after One week after Source: Cabinet Office As part of the preparations for a Tokai earthquake, the Large-Scale Earthquakes Countermeasures Law has designated Enhanced Earthquake Preparation Areas. In these areas, an earthquake emergency plan that includes measures for water supply systems should be developed. In the event of a major earthquake, the key issues are the minimization of damage to facilities, the reliability of emergency water supply systems and prompt restoration in the event of damage. These require the promotion of aseismic improvements with regard to water facilities and systematic, well-planned preparations for emergency water supply and other operations, especially in areas where considerable damage is anticipated. The Tonankai/Nankai earthquake is expected to result in the same extent of damage as the Tokai earthquake. The potentially affected areas need to plan their measures well in advance. 2. Timing with facility renewal As is mentioned in the section on the current status of earthquake disaster prevention measures (Chapter II-6), many water facilities in Japan are aging and their renewal in order to maintain their functions continues to be an important task for water suppliers throughout the country. In this context, improved earthquake disaster prevention measures for the facilities should be coordinated with the renewal of the old facilities. This is desirable in terms of upgrading the security of water supply systems as a whole. Efficient, well-planned implementation of such renovations should be emphasized from both the financial and functional points of view. 3. Broader-based backup systems Connection pipelines between adjacent water utilities are an effective means of ensuring water supplies in the event of an earthquake. These pipelines should have sufficient capacity, but many of the existing ones do not. Mutual support agreements that cover a broader area and involve many suppliers provide another option. Since these measures can also be 37

56 advantageous in the event of the need to deal with water pollution and drought, they should be operational in an integrated manner. An example of a broader-based backup system is found in the Master Plan for the Development of Broad-based Disaster Relief Bases laid down for the Keihanshin urban area. The Plan has recognized the effectiveness of backup systems to ensure stable water supply systems, and now it is at the stage of the coordination process to complete the system. Legend Supply of service water Connection pipelines Name of water supplier Kyoto City Hyogo Pref. Osaka Pref. Kyoto Pref. Kobe City Hanshin Water Supply Authority Osaka City Nara Pref. Figure 7 Broader-based backup system with wide-area pipeline connection network 4. Development of conditions for earthquake disaster prevention measures To promote facility improvement for earthquake-resistance and system development for emergency operations (i.e. water supply and restoration), it is necessary to consider several issues as fundamental conditions. Key issues are strengthening of the fiscal base, fundraising from sources external to the water utility account, engineering development and research activities and capacity building for the staff. To develop these conditions steadily, cooperation among the national government, water suppliers and concerned bodies is indispensable. IV. CONCLUSION Earthquakes are one of the risks in society. Water suppliers should take a stance that is as strict towards earthquakes as that towards drought and terrorism. Disaster prevention measures for earthquakes should be considered as a crucial part of crisis management strategies. The Ministry of Health, Labour and Welfare is intending to issue within this fiscal year a Water Supply Vision to provide the direction that future water services should work towards. In this process, earthquake disaster prevention measures will be examined more closely and positioned as a key issue of crisis management. In implementing actual measures, several appropriate indicators, such as the proportion of pipelines that are earthquake-proof and the backup ratio, will be used in combination in order to monitor the state of implementation macroscopically. Policy incentives will also be introduced to accelerate the progress throughout the nation. 38

57 ANTI-SEISMIC MEASURES OF EXISTING WATER SUPPLY FACILITIES A case study of an anti-seismic plan of Inagawa water treatment plant Keiichi MURAKAMI, Kazuo MISHIMA, Takashi HANAMOTO, and Kazuo OGURA ABSTRACT The Hanshin Water Supply Authority can supply 1,128,000m3 of drinking water a day to four cities in the Hanshin district including the City of Kobe. Inagawa water treatment plant which is main facility (Capacity; 916,900m3 / day) is composed of aged structures built in old design standard and new structures built in a new one.aged facilities were heavily damaged by The Great Hanshin-Awaji Earthquake which occurred in January, 1995.Most of facilities were only repaired temporarily because we could not stop the operation due to no margin of capacity in Inagawa water treatment plant which purified approximately 80% water of our authority. We conducted a seismic diagnosis of existing facilities of Inagawa water treatment plant in conformity with the Seismic Design and Construction Guideline for Water Supply Facilities and its Explanation, which had been improved in 1997 after the earthquake. It has been resulted that management center, rapid sand filtration, pump room and reservoir had no resistance against seismic motion level 2 which was the same intensity as The Great Hanshin-Awaji Earthquake. We anticipated this result before the diagnosis, because there was a great difference between old design standard and new one. We assessed that priority of improving seismic capacity should be given to management center which had computer system of controlling all the other facilities and rapid sand filtration which had been heavily damaged in the main four facilities. We considered about how to make the management center and the rapid sand filtration quakeproof. There is almost no space for work of improving seismic capacity in the plant, as distances of each facility are very close.it is difficult that how we make space. We considered three methods as improving seismic capacity of management center. One method was reinforcement of existing management center. Another was that subdivisions of each function of management center were installed on small spaces found in the plant. The other was that new management center was constructed though space for the construction had to be squeezed. Keiichi Murakami,Chief of the Construction Section, Design Division, Construction Department,HanshinWaterSupplyAuthority,20-1Nishiokamoto3-chome, Higashinada-ku Kobe Japan Kazuo Mishima Master of Engineering,Registered, Engineer,Manager of the Planning Division,Construction Department,Hanshin Water Supply Authority 20-1 Nishiokamoto 3-chome,Higashinada-ku Kobe Japan Takashi Hanamoto, Chief of the Project Section Planning Division, Construction Department,HanshinWaterSupplyAuthority,20-1Nishiokamoto3-chome, Higashinada-ku Kobe Japan Kazuo Ogura Planning Division, Construction Department Hanshin Water Supply Authority 20-1 Nishiokamoto 3-chome,Higashinada-ku Kobe Japan 39

58 1. Introduction The Hanshin Water Supply Authority is a regional authority established in 1936 for the supply of water to four cities (Kobe, Amagasaki, Nishinomiya, and Ashiya) in the Hanshin area. Prior to 1956, the authority constructed facilities for the Yodogawa supply system (Yodogawa Pumping Station, Amagasaki Water Treatment Plant, Nishinomiya Pumping Station, Kabutoyama Water Treatment Plant) with a total capacity of 373,000m 3 /day, and in 1972 it completed the Daido supply system (Daido Pumping Station, Inagawa Water Treatment Plant, Koto Pumping Station) with a total capacity of 595,000m 3 /day, representing a total capacity of 968,000m 3 /day. In 1978, it began expanding the capacity of these facilities to 1,289,900m 3 (Daido Pumping Station, Inagawa Water Treatment Plant, Koto Pumping Station, merging of Kabutoyama and Amagasaki facilities to form the New Amagasaki Water Treatment Plant). Currently, with half of the capacity of the New Amagasaki Water Treatment Plant on in reserve, total capacity is 1,128,000m 3 /day, and the system is able to supply approximately 80% of the demand for water from the four cities. The Hanshin-Awaji earthquake in January 1995 resulted in severe damage to the system, primarily to older piping and structures. Subsequent seismic evaluations of structures constructed prior to 1972 concluded that the system would be unable to withstand another similar earthquake (Level 2). The Inagawa Water Treatment Plant (capacity 916,900m 3 /day) is the Authority's primary facility, and its primary equipment (water treatment plant, pumping stations) suffered the most damage during the earthquake. Treatment facilities (e.g. filtration ponds) and managing center constructed prior to 1972 suffered damage, however only emergency repairs have been implemented and immediate seismic protection measures to withstand a Level 2 earthquake are therefore required. A seismic protection plan for the Inagawa Water Treatment Plant was therefore considered, and this plan, and problems involved in its implementation, are introduced below. Fig.1 Outline of Hanshin Water Supply Authority Facilities 2. Damage to Inagawa Water Treatment Plant, and Relevant, Antiseismic Evaluations Construction of the Inagawa Water Treatment Plant commenced in 1959 and partial operation began in The old facility (East and West Systems) was completed in 1972 with a capacity of 595,000m 3 /day. Subsequent expansion work for the new facility commenced in 1991, and partial operation began in The facility was completed in July 1997 with a capacity of 916,900m 3 /day. The new facility was designed from the outset to reduce micro-organic matter and remove odors, and as such was constructed with advanced treatment equipment (see Fig. 2). The horizontal-flow sedimentation ponds in the old facility were modified to an inclined pipe system, and advanced treatment equipment installed in the resulting available space. Water supply using this treatment and treatment equipment commenced in July 2000 (see Fig. 3). 40

59 C/F/ES Ozone GAC-FB C/RF Source Settlers Activated Carbon Media* O 3 Clear Well C/F/ES: coagulation / flocculation / enhanced sedimentation, GAC-FB: granular activated carbon fluidized bed, C/RF: coagulation / rapid filtration, Fig. 2 Improvement of Hanshin Water Supply Authority water treatment train 1) Damage The earthquake resulted in leakage from sedimentation ponds caused by damage to expansion joints, and damage to foundation piling 1). Exploratory drilling in the foundation piles of the sedimentation ponds revealed horizontal cracking at the top of approximately 60% of the piles. These structures were constructed on landfill, and it is thought that the damage was resulted from liquefaction of the soil during the earthquake. As with the sedimentation ponds, filtration pond structures were constructed on landfill, and exhibited ground subsidence and horizontal cracking at the top of piles. Considerable damage to ground floor columns and walls was apparent in the managing center (three floors above ground, one floor below ground), with shear cracks in columns, and vertical cracks in the central area of walls. Slab beams in third floor ceilings exhibited flexure cracking thought to be due to the load of roof-mounted water tanks. Damage to structures not constructed on landfill was slight, and no damage occurred in the new facilities partially completed in July Damage to primary structures, and repair methods, are shown in Table 1. Table 1 Damage of Primary Structures, and Repair Method Location of damaged Condition Year of Construction Flocculation and sedimentation : Inflow duct deformation : Water leakage caused by broken expansion joint : Subsoil run-off : 60% of foundation piles damaged (hair cracking found in the circumference) Emergency restoration method 1963 : Elastic rubber plate for cut-off in expansion joint : Filling urethane foam under the bottom slab Permanent restoration method : Newly built inflow duct and flocculation : Ground improvement Rapid sand filter : Subsoil run-off : Damage in foundation piles 1963 Mortar grout Machinery room Cracking in 6 columns 1963 Epoxy resin injection Wrapping steel plate 2) Implementation of Repair Work after the damages Of the sedimentation ponds, filtration ponds, and managing center structures in the damaged East and West Systems, seismic resistance of the flocculation and sedimentation ponds was improved by improving the foundations. The filtration ponds in which the foundations had subsided and foundation piling was damaged were repaired temporarily by injection of mortar into the voids resulting from subsidence of the foundations. Cracking in the beams and columns in the managing center was repaired by injection of epoxy resin and by wrapping with steel sheeting, however the deformation of structural members continued following the earthquake, resulting in further cracking and distortion window frames. This was accompanied by peeling and spalling of concrete, possibly also due to the age of the building. 41

60 West system facility East system facility New facility AWTF AWTF F/E F/E F/E MC RF RF AWTF r p r RF p F/ES:flocculation/enhanced sedimentation AWTF:advanced water treatment facility RF:rapidfiltration, r:finished water reservovoir p:distribution pump,mc:managing center r Fig. 3 Plan View of Inagawa Water Treatment Plant 3) Seismic Evaluations 3-1) Details of Seismic Evaluations The Seismic Protection Design Guidelines and Interpretations for Water Supply Facilities (Japanese Water Supply Association) were revised as a result of the damage experienced in the Hanshin-Awaji earthquake in The major changes from the previous guidelines (1978) were the need to establish seismic protection according to the importance of water supply facilities, and to incorporate mathematical methods for the two-stage evaluation using Level 1 earthquakes based on the previous level, and Level 2 earthquakes based on the Hanshin-Awaji earthquake. This method is applied not only to new structures, but to existing structures as well, and reinforcing work is required if the required seismic protection cannot be guaranteed. 2) Under the requirements of this revision, seismic evaluations of old facilities constructed prior to 1972 were conducted in 1999, and the flocculation and sedimentation ponds, and advanced treatment facilities, designed under the previous guidelines were examined under the new guidelines. 3-2) Results of Seismic Evaluations Table 2 shows the year of construction, and the results of the seismic evaluation, for the old facilities at the Inagawa Water Treatment Plant. The Level 1 seismic evaluation showed that three 42

61 of the facilities had insufficient seismic protection against a Level 1 earthquake. In particular, the managing center consisted of a large number of beams and columns for which loads occurring during an earthquake would be significantly in excess of the allowable values due to the large differences between the design procedures prevailing at the time of construction, and those in use in the current evaluation. Repairs to the flotation ponds by injecting mortar into the voids under the base plate have provided seismic protection against a Level 1 earthquake. The Level 2 seismic evaluation showed that none of the facilities had sufficient seismic protection due to the fact that the current design standards cannot be satisfied with the design standards prevailing at the time of construction. Table 2 Seismic Evaluation of Old Facilities Facility Year of construction (age as of 2002) L1 L2 Flocculation and sedimentation ponds (modified) Civil work 1998(East system), 2000(West system) Advanced water treatment facility (new) Civil work 1998(East system), 2000(West system) Filtration ponds (mortar injected into voids) Civil work 1968 (34 years) Water treatment ponds Civil work 1968 (34 years) Managing center Construction work 1963 (37 years) Pump building Construction work 1963, 1970 (37, 32 years) O: No problem : Not destroyed, however some structural members were deformed, and there were problems using the facility without repairs following the earthquake. x: Part of the structure was destroyed. L1: Earthquake occurred once or twice while the structure was in use (slightly less than Magnitude 6). L2: Probability of earthquake is low, however magnitude is considerable (Magnitude 7) and with significant effect on structures. 3. Seismic Protection Plan for Inagawa Water Treatment Plant Based on the results of the seismic evaluations it was determined that facilities constructed prior to 1972 were unable to withstand Level 2 earthquakes, and it was therefore decided to implement seismic protection for all facilities held by the Authority. A high priority was given to seismic protection for the Daido system facilities constructed prior to 1972 based on the results of the seismic evaluations, the damage resulting from the Hanshin-Awaji earthquake, and the importance of the facilities, and planning proceeded for seismic protection of the Inagawa Water Treatment Plant, a facility of prime importance to the authority in terms of its function in handling the majority of the water supply. Of the old facilities requiring seismic protection, the horizontal-flow system of flocculation and sedimentation ponds were modified to an inclined pipe system in order to obtain sufficient site area for the construction of advanced water treatment facility, and simultaneously the upper structures designed on the basis of the previous guidelines were renovated. The foundations were improved as a result of repair work on the lower structures. The advanced water treatment facility was designed according to the same standards as the flocculation and sedimentation ponds, and both facilities now have seismic protection up to the current Level 2. The structures requiring seismic protection were therefore the filtration ponds, the managing center, the pump room, and the treatment ponds. Simultaneous seismic protection work on all facilities presented problems in terms of maintaining sufficient water supply and operational expenses, construction of new facilities and advanced water treatment facilities for upgrading of existing facilities resulted in considerable congestion, and it became difficult to obtain sufficient site area for upgrading. An order of priority 43

62 was therefore established for the seismic protection work. Water Treatment Facilities Pumping Facilities Flocculation/ Advanced Water Raw Water Filtration Sedimentation Treatment Facility Reservoir Pump Room Managing Facilities Managing Center Fig.4 Outline of Inagawa Water Treatment Plant The managing center is the core of the control facilities, and as such consists of important facilities such as the integrated control room, the office, chemicals injection equipment, and reverse washing tanks for the filtration ponds. On the other hand, deterioration of the facilities due to age was considerable and it was no longer possible to guarantee Level 1 seismic protection, so that there was a fear that operation of the Inagawa Water Treatment Plant could not be continued in the event of an earthquake of similar magnitude to the Hanshin-Awaji earthquake. Furthermore, the filtration ponds are the final process of the water treatment facilities, and while Level 1 seismic protection is possible, there was a possibility that the total water treatment and treatment capacity of the old facilities (595,000m3/day) could not be maintained in the event of an earthquake of similar magnitude to the Hanshin-Awaji earthquake. The water treatment ponds and pump room form a part of the pumping facilities, and the water treatment ponds are interconnected with, and substitute for, the new facilities so that water pumps are also able to pump from the new facilities. On this basis, the managing center and the filtration ponds were assigned the highest priority in consideration of the importance of the facility, extent of damage, and seismic performance. The methods used in providing seismic protection are described below. 1) Evaluation of Seismic Protection for Managing Center Two methods of providing seismic protection were possible - reinforcement of the existing managing center, and new construction. The reverse washing tanks for the filtration ponds are installed in the top of the managing center. The removal of these tanks and reinforcement of the building to prevent secondary damage was designated as Seismic Protection Reinforcement Proposal (1). Two upgrade proposals were presented, one in which the facility was to be upgraded at the current location, and the other in which a new facility was to be constructed at an appropriate site. The first proposal would have required a temporary managing center and was therefore considered uneconomic. Obtaining an appropriate site for new construction in the immediate area presented difficulties under the current conditions, and a Split Function Proposal (2) was therefore developed in which the functions of the managing center were split into pump room, chemicals injection, office, and wash water tanks, each being installed separately in available space on the site. In order to obtain sufficient space for new construction, a New Construction Proposal (3) was developed in which filtration ponds requiring seismic protection were to be replaced with new filtration ponds operating at twice the rate of the current ponds (Fig. 5). 44

63 2Split Function Proposal 3New Construction Proposal 1 Seismic Protection Reinforcement Proposal Fig. 5 Location of Control Room Upgrade 1-1) Evaluation of Seismic Protection Reinforcement Proposal Structural reinforcement within the managing center would involve large-scale reinforcing of columns and beams, and require significant relocation of existing equipment. It was therefore decided to extend the beams in the managing center and reinforce with additional new columns. As the existing filtration ponds interfere with construction of the new columns, some of the ponds would be removed. Furthermore, structural reinforcement of columns in the managing center requires relocation of existing facilities (Fig. 6), and above-ground wash water tanks are required. The current managing center constructed in 1963 requires reinforcing to provide seismic protection, and it would therefore be upgraded to allow use for a further 20 years. 1-2) Evaluation of Split Function Proposal The current managing center supports four functions - operation and control (control room, computer room, and water quality instrumentation room), chemicals injection (chemical tanks and chemical injection equipment), the office (office and conference room), and treatment. Evaluation of the proposed splitting of functions considered the need to aggregate the appropriate facilities for each function, and to prevent deterioration in the efficiency of operation and control, and therefore examined layout in the available space within the water treatment site. Available space in the vicinity of the current managing center is limited to the area between the East and West system sedimentation ponds and the filtration ponds, and it was decided to use this area as the control space (managing center) important for operation and maintenance, and the chemical injection space (chemical injection building). As the control space is limited it was decided to use the area between the East system sedimentation ponds and the filtration ponds. 45

64 Managing Center Filtration Filtration A A Filtration Filtration Managing Center Protection Reinforcement New Columns A - Fig.6 Artists Conception of Seismic Protection Reinforcement Proposal The chemical injection space is located in the area between the West system sedimentation ponds and the filtration ponds. While it was possible to have the office in the available space on the site, this was considered disadvantageous in terms of operation and maintenance. In order to have the control and chemical injection spaces adjacent to each other, the current managing center would be demolished after the two spaces were completed and the resulting area used for the office (Fig. 7). Chemical injection Managing Center Office Purification Fig.7 Split Function Proposal 1-3) New Construction Proposal Following reconstruction of the West filtration ponds as rapid filtration ponds, the East filtration ponds are no longer required, and would therefore be demolished to make way for construction of the new managing center. As the New Construction Proposal assumes the filtration ponds to include wash water tanks, the operation and control, chemicals injection, and office functions were aggregated, however as this proposal requires construction of the rapid filtration ponds a means of providing washing facilities during construction of the ponds is required. A comparison of the above proposals for seismic protection of the managing center showed that early seismic protection is possible with the Seismic Protection Reinforcement Proposal (1), and that this proposal is beneficial in terms of cost, however it requires measures to compensate for 46

65 deterioration in filtration capacity, and will require upgrading after approximately 20 years. Early seismic protection is also possible with the Split Function Proposal (2), however a construction period of considerable duration is required, and cost is relatively high, so that the splitting of functions will result in complex working flows in terms of operation and maintenance. The New Construction Proposal (3) assumes construction of rapid filtration ponds, and is therefore considered to be in the best option from the point of view of upgrading of facilities and improvements in efficiency of operation and maintenance (Table 3). Table 3 Comparison of Proposals for Seismic Protection of Managing center Details Timing Seismic Protection Reinforcement Proposal (1) Removal of upper wash water tanks, and seismic protection and reinforcement of managing center. Construction able to be commenced at an early stage (requires prior of replacement wash water tanks). Split Function Proposal (2) New Construction Proposal (3) Managing center functions (control room, chemicals injection, office, wash water tanks) split into four and dispersed within available space on site. Construction able to be commenced at an early stage. 47 West filtration ponds reconstructed as rapid filtration ponds, East filtration ponds no longer required and therefore demolished and used as site for construction of new managing center. Construction able to be commenced following construction of rapid filtration ponds. Duration of 3 years 6 years 3 years construction Washing water Y Y N tanks required Y/N Cost 1,300,000,000 3,700,000,000 2,000,000,000 Advantages * Early seismic * Early seismic protection * Upgrading of structures protection of managing of managing center is deteriorated with age. center is possible. possible. * Expenses are low for * Upgrading of structures * Improved efficiency of the time being. which have deteriorated Problems * Measures required to compensate for deterioration in filtration capacity. * Upgrading of structures required following reinforcement for seismic protection. with age. * Costs are relatively high. * Working flows are complex in terms of operation and maintenance. * Seismic protection of filtration ponds delayed. operation and maintenance. * Assumes construction of rapid filtration ponds. Evaluation 2) Evaluation of Seismic Protection for Filtration Ponds The following three proposals for seismic protection of filtration ponds were evaluated. (1) As seismic evaluations confirmed Level 2 seismic protection of upper structural members, the Foundation Improvement Proposal involves continued use of current filtration ponds with improvements to foundations. (2) The Staged Rebuilding Proposal involves upgrading at the current location. (3) As new construction and construction of advanced water treatment facilities would result in congestion, the Upgrade Proposal involves rebuilding of filtration ponds as rapid filtration ponds in order to obtain a site for facilities requiring upgrading 2-1) Evaluation of Foundation Improvement Proposal This proposal makes use of the existing filtration ponds, and involves staged improvements to

66 the foundations beneath the base to prevent liquefaction. The proposal would be implemented in the winter to ensure minimum volume of water in the ponds, and use of the ponds would be halted in stages, the equipment in the ponds (e.g., catchment troughs, filters, bottom catchment equipment) being removed before the foundations beneath the base are improved, and the equipment then restored (Fig. 8). The problems associated with this proposal are as follows. a) While seismic protection is improved, the filtration ponds structures remain unchanged and are not upgraded. b) Construction requires a period of approximately eight years, and the associated effects on operation and maintenance and water treatment and treatment during construction period are considerable. Fig.8 Artists Conception of Foundation Improvement Proposal 2-2) Evaluation of Staged Rebuilding Proposal Staged rebuilding of the current filtration ponds would be limited to the winter to ensure minimum volume of water in the ponds while they are demolished, and new ponds constructed in stages, eventually resulting in construction of new rapid filtration ponds similar to the current filtration ponds (Fig. 9). The problems associated with this proposal are as follows. a) Construction requires a period of approximately eight years, and the associated effect on operation and maintenance and water treatment and treatment during construction period are considerable. b) Restrictions on the timing and method of implementation are more severe than is the case with the Foundation Improvement Proposal, and implementation of the proposal therefore presents difficulties. 48

67 Maneging center Filtratio Flow in Shut down Flow Buil Buil Buil Buil 3BL 2BL 1BL Shut down 2 nd winter Shut down 4 th Removal 1 st winter Fig.9 Artists Conception of Staged Rebuilding Proposal 2-3) Evaluation of Upgrade Proposal The Upgrade Proposal involved demolition of the current filtration ponds and construction in the space obtained of one system of rapid filtration ponds operating at twice the rate of the current filtration ponds. Construction would require a period of three years, thus permitting a reduction in the construction period in comparison to the Staged Rebuilding Proposal above, and minimizing the effects of construction work on the water treatment plant. The fact that one filtration pond system cannot be used during the construction period, thus reducing water treatment capacity by half, and requiring substitute capacity, represents a significant problem associated with this proposal. Water treatment capacity during the construction period would be 755,920m 3 /day, 219,149m 3 /day less than the maximum possible of 975,060m 3 /day (Table 4). Table 4 Water Treatment Capacity during Construction Period Capacity of facility (including future increment)(m 3 /day) 49 Current capacity (m 3 /day) Capacity during construction period (m 3 /day) Maximum capacity (m 3 /day) Total 1,289,900 1,128, , ,060 (reached in 1964) Inagawa water treatment plant New facility Old facility 916, , , , , , , , ,520 New Amagasaki water treatment plant 3 rd winter 373, , ,500 The Foundation Improvement Proposal, the Staged Rebuilding Proposal, and the proposal for improvement to rapid filtration ponds (Upgrade Proposal) were compared. It was concluded that some means of countering the reduction in water treatment capacity of the Inagawa water treatment plant would be required, so that in terms of upgrading of facilities, duration of construction, and costs for maintenance of facilities, the Upgrade Proposal (improvement to rapid filtration ponds) represented the best solution. 3) Comparison by Combined Seismic Protection of Managing center and Filtration Ponds Various methods of providing seismic protection separately for the managing center and filtration ponds were investigated, however it was concluded that seismic protection was required for the entire Inagawa water treatment plant. Table 5 shows the relationship between volume of

68 protected water supply as obtained by combining seismic protection for the managing center and filtration ponds, and the duration of construction. Table 5 Comparison by Combined Seismic Protection of Managing center and Filtration Ponds (1,000m 3/day) D uration of construction Seism ic protection rainforcem ent + Staged rebuilding M anaging cente Filtration Volum e of seism ic protected water suppy Seism ic protection rainforcem ent M anaging cente + U pgrade(high-rate filtration) Volum e of seism ic protected water 900 suppy A m agasaki Filtration Split function + Staged rebuilding M anaging center Filtration Volum e of seism ic protected water suppy Split function + U pgrade(high-rate filtration) Volum e of seism ic protected water suppy M anaging center 510 A m agasaki Filtration Managing center 5N ew construction +Upgrade(high-rate filtration) Volum e of seism ic protected water suppy A m agasaki Filtration M anaging cente Seismic protection for the Inagawa water treatment plant involved investigation of the Seismic Protection Reinforcement Proposal (1), the Split Function Proposal (2), and the New Construction Proposal (3) for the managing center, and the Foundation Improvement Proposal (1), the Staged 50

69 Rebuilding Proposal (2), and the Upgrade Proposal (3) for the rapid filtration ponds. Further investigation of combined seismic protection in terms of maintenance of facilities in the entire Inagawa water treatment plant showed that construction of a new managing center in space obtained by upgrading of the existing filtration ponds to rapid filtration ponds would allow early seismic protection of the treatment facilities. Furthermore, in terms of technical requirements, for example upgrading of facilities and operation and maintenance, this represents the best solution. On the other hand, a point requiring consideration in the near future in association with seismic protection of the Inagawa water treatment plant are the measures required to counter the reduction in capacity of the water treatment facilities during upgrading of the filtration ponds. Furthermore, investigation is required of proposals to ensure the necessary volume of water in light of trends towards reduced demand for water, and the associated lack of urgency for immediate work on future facilities for the Amagasaki water treatment plant, and proposals to provide substitute facilities while those which have deteriorated over the years are provided with seismic protection in a planned manner. Future substitute capacity and availability of funding for work also require consideration. 4. Conclusion For reasons of construction funding and available volume of water, all necessary rebuilding work cannot be conducted simultaneously, and an order of priority must therefore be assigned which considers the state of structures and the time required for construction to ensure that work proceeds in a planned manner. Seismic evaluations of the Inagawa water treatment plant revealed low seismic resistance of structures constructed prior to 1972, and that a planned program of seismic protection for these existing structures was necessary in order to ensure security in water supply. Evaluation of the need for provision of seismic protection for these existing structures was based on the damage sustained in the Hanshin-Awaji earthquake, and the importance of the structure, and it was determined that a high priority should be given to seismic protection of older structures. Evaluation of these older structures, notably managing centers and filtration ponds, and their seismic resistance, showed that an early program of providing seismic protection was required. References 1) Kazuo Mishima, "Restoration and Anti-seimic Measures of Water Supply Facilities of Hanshin Water Supply Authority" U.S-Japan Anti-Seismic Measures Workshop ) Japanese Water Supply Association - Seismic Protection Design Guidelines and Interpretations for Water Supply Facilities, March

70 52

71 San Francisco Public Utilities Commission (SFPUC) - Capital Improvement Program Jeet Bajwa (San Francisco Public Utilities Commission) ABSTRACT The SFPUC s Capital Improvement Program (CIP) has adopted a long-term plan of infrastructure improvement projects that would be implemented during the next ten years. The CIP consists of new projects that either replace or upgrade an old portion of the system in an effort to maintain the reliability of the Regional Water, Local Water and Power systems, and meet future customer needs. This CIP identifies 37 projects for regional water and 40 projects for local water. The total estimated cost of the CIP program for projects is $3.6 billion. Just like many other utilities, the customers of the SFPUC rely on the water, wastewater and power system 365 days a year, 7 days a week, 24 hours a day. Despite this reliance, the SFPUC system is faced with vulnerabilities as a result of aging infrastructure; natural threats such as earthquakes and drought; changing regulations; and increased demand. INTRODUCTION Aging infrastructure The SFPUC system is old. Many of its components were built in the 1800s and early 1900s and are approaching the end of their useful life. As the system ages its vulnerability to failure increases. Exposure to natural threats Many of the SFPUC system components are located on or in the immediate vicinity of three major earthquake faults: the San Andreas, Hayward and Calaveras Faults. Changing regulations Regulatory standards are constantly changing and the SFPUC must respond by upgrading its system to remain in compliance with applicable regulations. Increased demand Water demand among SFPUC customers is expected to increase beyond current supply over the next 30 years. Other challenges facing the SFPUC include improving wastewater treatment facilities to increase reliability, reduce odors and visual impacts; improving alternative water supplies in the City; increasing power generation; improving power system efficiency; and reducing power system vulnerability. The SFPUC has developed the Long-Term Strategic Plan for Capital Improvements, Capital Improvement Program, and Long-Range Financial Plan to address these challenges. The following provides an overview of each of these documents. 53

72 54

73 Seismic Measures for Waterworks in Yokosuka City Takashi Furuya ABSTRACT Yokosuka City is located in the middle of the Miura Peninsula in the southeast of Kanagawa Prefecture, facing Tokyo Bay to the east and Sagami Bay to the west, and has an area of about square kilometers and a population of about 430,000. The Southern-Kanto district, of which Yokosuka City forms a part, is a region subject to frequent earthquakes compared to other regions in Japan, as it is situated on a boundary point where three plates meet (the Pacific Plate, the Philippine Sea Plate and the Eurasia Plate). The Great Kanto Earthquake of September 1, 1923 (M. 7.9) caused more extensive damage to Yokosuka than any other earthquake that Yokosuka has experienced since it was designated as a city in In that earthquake, as Yokosuka was in the area near the epicenter, about 70% of its waterworks facilities were damaged and emergency water supply was needed for a long period of time. Since 1978, the Yokosuka City Waterworks Bureau has been working on seismic measures under the Five Key Strategies for Seismic Measures. After the Great Hanshin Earthquake in 1995, we re-evaluated overall disaster planning in consideration of the potential occurrence of near-field earthquakes. As the waterworks facilities will need to be updated in the near future, we are now designing measures to promote the overall reinforcement of the earthquake resistance of the distribution pipeline network through effective update of distribution pipelines. In this paper, I present the basic policy for seismic measures and an overview of the actual measures taken to date, and address the re-evaluation of water distribution blocks in consideration of expected damage to distribution pipes due to earthquakes and the updated plan for distribution pipelines based on such re-evaluation. Takashi Furuya, Engineer, Management and Planning Division, Yokosuka City Waterworks Bureau 11 Ogawachou, Yokosuka City, Kanagawa, Japan 55

74 1. NECESSITY FOR SEISMIC MEASURES Yokosuka has the possibility of suffering earthquakes that would cause damage (damaging earthquakes), such as (1) earthquakes with epicenters on the Sagami Trough (ones similar to the Great Kanto Earthquake), (2) earthquakes known as Tokai earthquakes with epicenters on the Suruga Trough, and (3) near-field earthquakes with epicenters within plates. As Yokosuka has, within its limits, the Kitatake Fault, the Takeyama Fault and the Kinugasa Fault, all of which are considered as active faults with high activity[1], Yokosuka has the possibility of suffering near-field earthquakes with epicenters in these faults in the future. As Yokosuka s waterworks facilities are located near large cities such as Yokohama City, Kawasaki City and the Tokyo Metropolitan Area and situated on a peninsula, it would take a great deal of time to receive assistance and supplies of emergency materials from other cities should a damaging earthquake occur. Furthermore, as more than 99% of the total water sources come from rivers situated far from Yokosuka the Sagami River and the Sakawa River and less than 1% comes from within the city limits, the water supply would inevitably be cut off for a long period of time if conveyance facilities or transmission facilities were damaged. These facts have also led to the debate on the importance of seismic measures. TOKYO KAWASAKI CITY YOKOHAMA CITY Tokyo Bay Hanbara Water System Arima Water System Kosuzume Water System Sakawa River Water System Miyagase Water System Hasizimizu Water System Sagami River Sakawa River Sagami Bay YOKOSUKA CITY Fig.1. Water sources for Yokosuka 56

75 2. DISASTER PREVENTION PLAN 2.1. Basic policy for seismic measures and actions taken The Yokosuka City Waterworks Bureau prepared the Yokosuka City Waterworks Bureau Disaster Countermeasures Guidelines in 1978 and started to take full-fledged measures against disasters. As a basic policy, the Waterworks Bureau established the Five Key Strategies for Seismic Measures and started to take concrete actions according to the strategies. In light of the damage caused by the Great Hanshin Earthquake in 1995, there was a need to review the disaster prevention plan while taking the possibility of a near-field earthquake into account. The Waterworks Bureau re-evaluated the overall disaster planning accordingly. With the expectation of a near-field earthquake at active faults within the city limits, we examined the relationship between the waterworks facilities damage caused by the Great Hanshin Earthquake and the geological data of Kobe City. We then compared this data with the geological data of Yokosuka City to estimate possible damage and designed the Emergency Restoration/Water Supply Plan based on the estimation. We also decided to promote measures to increase the earthquake resistance of waterworks facilities so as to minimize the number of damaged points[3]. Reinforcement of transmission/distribution mains Development of new emergency water sources Securing drinking water Establishment of water supply system Clarification of roles TableⅠ. Five Key Strategies for Seismic Measures Contents Major measures Major transmission/distribution mains - Construction to increase reinforced for earthquake resistance. earthquake resistance of the Arima water transmission system - Construction to fortify distribution trunk lines against Two wells for emergency use constructed in the city. Water storage facilities installed for purpose of securing as much drinking water as possible within the city. Emergency restoration plan and emergency water supply plan developed. Necessary equipment and materials prepared. Emergency roles of individual employees of the bureau clarified. active faults - Development of water supply sources for emergency use at Hashirimizu and Hayashi - Shitamachi Distribution trunk lines - 100m 3 Tanks (Water storage equipment for emergency use directly connected to water pipes) - Installation of emergency stop valves in distribution reservoirs - Preparation of necessary mechanical equipment and resources - Development of information materials on facilities - Creation of operation manual and implementation of emergency drill - Motorbike investigation group 57

76 Verny Park Sitamachi Koyabe Fig.2-2. Seismic Design of the Otawa distribution trunk line at active faults [2] Otawa Fig.2-1. Utilization of distribution trunk lines Fig m 3 Tank (Water storage equipment for emergency use directly connected to water pipes) 2.2. Emergency restoration plan For the purpose of achieving early restoration of waterworks facilities, it was provided that restoration works shall be implemented under the following basic policies. (1) Major facilities capable of larger supply should be restored first. (2) Restoration work for facilities that can be restored earlier should be implemented first. (3) Restoration work should be implemented in consideration of the operation of facilities and the methods for managing facilities including backup measures. (4) Restoration work should be implemented so as to establish an emergency water supply system using pipelines within the entire area of the city by restoring distribution trunk lines and distribution mains as soon as possible. (5) Due consideration should be given to facilities that will play an important role in the event of disaster such as large hospitals Emergency water supply plan For the purpose of providing emergency water supply, it is necessary to design an emergency water supply plan while taking medical institutions and evacuation centers into consideration. Emergency water supply shall be implemented in the following ways. (1) Direct water supply from waterworks facilities ( Primary water supply points ) Emergency water supply will be provided directly from water storage equipment for 58

77 emergency use that is directly connected to water pipes (100 m 3 tanks), distribution reservoirs with emergency stop valves, distribution trunk lines with the additional function of water storage, and waterworks facilities as emergency water supply sources. There are a total of 68 primary water supply points designated within the city. (2) Transportation of water Drinking water will be transported to emergency water supply points designated at intervals of about 1 km ( Secondary water supply points ) in order to provide emergency water supply. There are 207 secondary water supply points designated within the city, mainly in parks, squares and public facilities (80 in evacuation centers and 127 on the streets). There are 15 water transportation bases designated within the city, mainly distribution reservoirs. (3) Installation of temporary service appliances For emergency water supply, temporary service appliances will be installed in distribution pipes that are not damaged or that are restored. We will provide emergency water supply mainly from Primary and Secondary water supply points with the aim of supplying 3 liters/day per person within about one week immediately after an earthquake. After that period, we will install temporary service appliances in fire hydrants, depending on the progress of restoration work for pipelines, and we will gradually increase the volume of emergency water supply. Within about one month of an earthquake, we will provide at least one service appliance for each household with the aim of supplying more than 100 liters/day per person. TableⅡ. Targets for emergency restoration and emergency water supply Emergency restoration Installation of at least one temporary service appliance for each household within about one month of an earthquake. Emergency water supply 3 liters/day per person immediately after an earthquake, and gradually increase the volume of water supply along with the progress of restoration. More than 100 liters/day per person within about one month. 3. EARTHQUAKE-RESISTANCE IMPROVEMENT FOR WATERWORKS FACILITIES Yokosuka City designed the Distribution Blocks System Basic Plan and the Computerized Mapping System Master Plan in These plans are now the basis for the improvement and management of waterworks facilities. With respect to the management of distribution blocks, the whole water supply area is divided into five large blocks, which are divided into 26 medium blocks by distribution reservoir, which are then divided into 198 small blocks. The supply population in each small block is about 59

78 2,000 to 4,000 persons. Water is distributed from distribution trunk lines and distribution mains generally via two or more injection points. We have been operating the Mapping System since 1989 and using it to obtain map information, pipeline information, water supply information, etc. for overall waterworks services. In 2002, it became possible to conduct hydraulic/water-quality analysis for all pipelines using the system, and we have been utilizing it in order to identify problems with the existing pipelines and to design an update plan. Under the Earthquake-Resistant Water Pipeline Plan developed in 1998, we promote measures to increase the earthquake resistance of pipelines by earthquake-resistant materials for pipes when installing new distribution mains, distribution sub-mains, small service pipes connected to important emergency institutions, and small pipes with diameters greater than 150mm. Filtration Plant A Large Block A Medium Block A Medium Block Reservoir Reservoir Distribution trunk lines and mains Distribution sub-mains Distribution sub-mains Small Blocks Small Blocks Fig.3. Distribution Blocks 4. EXPECTED EARTHQUAKE DAMAGE TO PIPELINES AND PROMOTION OF EARTHQUAKE-RESISTANCE MEASURES 4.1. Ground classification For the purpose of increasing the earthquake resistance of pipelines, we identified the danger level of the ground in various areas in the event of an earthquake based on the organized boring data on 2,800 boreholes within the city and classified the ground in each area into Classes A, B and C in the order of danger. Since earthquake damage to pipelines is governed by the displacement of the surface ground, we selected boring data on 64 boreholes that could be considered representative of the ground conditions in Yokosuka and calculated the ground strain 60

79 and the displacement amplitude in the event of an earthquake by the Response Displacement Method. As a result, we found a close relationship between the ground strain and the thickness of the surface ground. In light of ground collapse (danger of liquefaction, landslides, active faults) due to earthquakes for which the Response Displacement Method is not applicable as well as uneven settlement at ordinary times (reclaimed land, banking/cutting boundary), we developed a danger ranking level and conducted zoning for the whole area of Yokosuka City. We then divided the whole area of the city into 1,704 meshes of 250m in length (each mesh is formed by equally dividing a third mesh under the standard regional mesh system of the Geographical Survey Institute into 16 sectors) and classified these meshes into five categories depending on the percentage of Class A grounds contained in each mesh. Computerized Mapping System pipelines data Boring data Material, Diameter, Length Standard damage rate ( Ds ) Correction by type of ground ( Z 1 ) Setting the damage rate for individual pipelines Data for analysis Map data Important institutions data Software for pipe networks analysis Composition of data Chart on the risk ranking of grounds Percentage of Class A grounds ( Figure of 250m meshes) Setting the damage probability for individual pipelines Flow quantity comparison of individual pipelines Estimation of damaged pipelines by Monte-Carlo simulation analysis method Used for giving priorities for pipeline update Fig.4-1. Procedures for expecting damage to pipelines 4.2. Expectation of damage to pipelines We examined the ground conditions and installation of pipelines in Yokosuka according to the data on damage due to the Great Hanshin Earthquake, and estimated the number of damage cases and the damaged points Calculation of the number of damage cases Based on the data on damage due to the Great Hanshin Earthquake provided by Kobe City, Nishinomiya City and the Hanshin Water Supply Authority, we decided the standard damage rate (Ds) by pipe material/pipe diameter and calculated the number of damaged points by the following 61

80 formula. C = Ds Z 1 L (1) C: number of damaged points; Ds: standard damage rate (number of damaged points/km); Z1: correction by type of ground; L: length of pipeline (km) Correction by type of ground We corrected the data on damage due to the Great Hanshin Earthquake in light of the difference in ground conditions between the Hanshin area and Yokosuka City. We plotted, by type of ground, the total damage data provided by the Kobe City Waterworks Bureau on topographic charts and geological charts, and considered correction by type of ground according to the relationship between ground and damage. When we divided the ground in Kobe City into seven types and assigned it to Yokosuka s ground classes, A, B and C; 78% was assigned to Class A, 17% to Class B and 5% to Class C, which revealed that most damage occurred on Class A ground. On the basis of these percentages, we classified the ground assigned to Class A into five subtypes and set the correction coefficient (Z 1 ). TableⅢ. Standard Damage Rate by pipe material / pipe diameter and Correction by type of ground Diameter Standard Damage Rate (points/km) Percentage of Class A Grounds Z 1 (mm) ACP VP CIP DIP SP DIP(seismic-proof) Less than ~19% ~39% ~59% ~79% ~100% 1.25 Greater than Calculation of the number of damaged points We calculated the damage rate for all pipelines in light of the danger level of the ground in which the pipelines were installed. More specifically, we corrected the standard damage rate by pipe material/pipe diameter according to the ground conditions for individual pipelines (about 60,000 pipelines for the pipe network calculation). We calculated the number of damage cases within the whole area of the city by multiplying the damage rate by the length of pipelines by pipe material/pipe diameter, and collected the numbers for each distribution block. 62

81 Estimation of damaged points In estimating damaged points in pipelines, we used the Monte-Carlo simulation analysis method. This is a method to estimate damaged points, on the assumption that earthquake damage to pipelines will occur according to the Poisson distribution, by comparing the damage probability based on the length of each pipeline with a computer-generated random number between 0 and 1. We calculated the damage probability for individual pipelines in each medium block (distribution reservoir) using the following formula. PF = 1-e (-R L) (2) PF: damage probability for pipeline; R: damage rate for pipeline (number of damaged points/km); L: length of pipeline (km) If damage occurs, PF will be greater than the random number. As a damage pattern, we estimated damaged points by comparing the damage probability with a random number for individual pipelines within each medium block. We repeated this estimation 100 times and identified spots with frequent damage as estimated damaged points so as to ensure consistency with the number of damaged points in each block calculated in An example of estimation is shown below. The flags in Fig. 4-2 indicate expected damaged points Fig Expectation of damaged points in pipelines In the 2100 Block containing 10 small blocks and having a pipeline of 237 km in length, 44 spots are expected to be damaged in small distribution pipes and distribution sub-mains, mainly in coastal reclaimed areas. In the 4300 Block containing 8 small blocks and having a pipeline of 70 km in length, 63

82 which has ground with relatively high earthquake resistance and which is located in an upland area, 10 spots are expected to be damaged, mainly in 50mmφVP small distribution pipes. 4.3 Promotion of earthquake-resistance measures and disaster prevention support Fig. 4-3 shows the distribution of the average flow volume through pipelines at ordinary times, which is calculated by hydraulic analysis. The average flow volume is large for the pipelines indicated with bold lines, and these pipelines are considered important within each distribution block. With respect to the pipelines both indicated in this figure and identified in Fig. 4-2 as containing expected damaged points, it is necessary to replace them with earthquake-resistant pipes. The anticipation of damaged points plays an important role in designing plans for implementing disaster prevention measures. Based on such anticipation, we are able to distinguish pipelines that would have a great impact from those that would not, and design an action plan concerning the method of passing water through undamaged pipelines as well as the method of investigating water leakage, selection of locations to install temporary service appliances for emergency water supply, the method of selecting points for supplying transported drinking water, and the order for implementing restoration work. It is also important to carry out such disaster prevention process and to encourage persons in charge in divisions concerned to consider what they should do on a daily basis Fig Flow volume distribution 5. UPDATE AND IMPROVEMENT PLAN FOR DISTRIBUTION PIPES Due to the necessity to promote effective updating of pipelines using limited funds, we evaluated and gave scores to individual pipelines in respect of (1) aging level, (2) water pressure, 64

83 (3) water quality, (4) readiness for accidents, and (5) readiness for earthquakes, and set priorities for updating. The scoring methods were as follows. (1) Aging level We gave scores for cast iron pipes (CIP) etc. that were installed before (2) Water pressure We gave scores on the degree of risk and the degree of impact of reduced water pressure, on the basis of hydraulic grades obtained by hydraulic analysis and the average flow volume as indicators. (3) Water quality We evaluated the degree of risk of deteriorated water according to the decrease in density of residual chlorine and the occurrence of retained water, and gave scores on the basis of pipe material/pipe diameter and flow velocity as indicators. We estimated the decrease in density of residual chlorine by calculating the density of residual chlorine in pipelines by hydraulic/water-quality analysis. We calculated the coefficient necessary for hydraulic/water-quality analysis by on-site measurement. (4) Readiness for accidents We evaluated the degree of risk of accidents at ordinary times from the perspectives of possibility of accidents and impact of water failure. As for possibility of accidents, we gave scores on the basis of water leakage record and the corrosion level expected from the corrosion research as indicators. As for impact of water failure, we gave scores on the basis of (i) the average flow volume obtained by hydraulic analysis and (ii) the importance of the roads under which pipelines were installed. (5) Readiness for earthquakes We evaluated the importance of individual pipelines in the event of an earthquake, according to their connection to important facilities such as large hospitals and disaster measure facilities in the event of an earthquake and the degree of damage risk due to earthquakes. As for connection to important facilities, we identified pipelines that were connected to important facilities by hydraulic analysis and gave scores according to the number of facilities to which they were connected. As for the degree of damage risk due to earthquake, we gave scores on the basis of (i) the damage rate of pipelines, (ii) the average flow volume obtained by hydraulic analysis, and (iii) the importance of the roads under which pipelines were installed. We set priorities for updating all pipelines based on the sum total of their scores in these five items. By improving pipelines from the comprehensive perspective considering earthquake-resistance measures, we are promoting the establishment of a safe and functional water service system that is prepared for disasters. 65

84 6. FUTURE SEISMIC MEASURES In order to minimize damaged points due to earthquakes and reduce the time necessary for emergency restoration and emergency water supply, it is important to promote earthquake-resistance measures. As it is necessary to implement effective update and improvement for waterworks facilities that need update, we will select pipelines to be updated according to priorities. As for transmission pipes, we will consider the feasibility of backup between water systems and decide the transmission routes to be updated according to priorities. As for distribution pipes, we will re-evaluate the pipeline networks in each distribution block and set priorities for effective update of pipelines based on the re-evaluation, and thereby promote reinforcement of overall earthquake resistance. In addition to these measures, it is important to prepare a system to cope with disasters and clarify the role of the individual employees of the Waterworks Bureau on a daily basis. For the purpose of encouraging individual employees to understand their roles independently, the Yokosuka City Waterworks Bureau implements emergency drills twice a year. However, since there is a limit to the number of employees, we need help from other organizations so as to implement emergency restoration and emergency water supply promptly. While we have held drills mainly for our employees so far, we should plan drills for the participation of other people who would be assisting in such event. In particular, we will provide emergency water supply mainly from 100m 3 tanks installed in evacuation areas for the three days immediately after an earthquake. Employees will install emergency service appliances in 44 tanks within the city, and then we will ask citizens who come to individual evacuation areas for emergency water supply work. It is important to clarify the role of the Waterworks Bureau and that of citizens and publicize what we would ask for citizens in the event of an earthquake, as well as to encourage them to participate in emergency drills. We should put emphasis on disaster measures in cooperation with local residents, residents associations/neighborhood associations, and volunteer groups in the future. REFERENCE [1] The Research Group for Active Faults of Japan, Active Faults In Japan (1991) [2] Igari, Katsudansō wo tsūka suru Ōtawa haisui kansen chikuzō kōji no hōkoku (Report on the Laying of Distribution Trunk Lines in Ōtawa Passing Through Active Faults), 47th National Meeting for Reading of Waterworks Research Papers (1996) [3] Furuya, Yokosuka-Shi Suidōkyoku no saigai taisaku keikaku ni tsuite (Yokosuka City Waterworks Bureau Disaster Prevention Plan), The Fourth International Symposium on Water Pipe Systems (1997) 66

85 3 rd US-Japan Workshop on Water System Seismic Practices TECNICAL SESSION II Seismic Performances, Preparedness, and Readiness Damages and Motion of Pipelines in Liquefied Ground Presenter: Toshio Toshima (Japan Ductile Iron Pipe Association) Seismic Evaluation of Water Supply System in Health Facilities Presenter: Masakatsu Miyajima (Kanazawa University, Japan) The Research of Damages of Public Water Supply Pipelines during the Ji-Ji Taiwan Earthquake on September 21, 1999 Presenter: Ping-Hsin Wang (Taiwan Water Supply Corporation, Yuang-Shang, Taiwan) Emergency Operation Planning - How Contra Costa Water is Building Earthquake Response Capabilities In Calm to Excel in Emergency Presenter: Stephan Welch (Contra Costa Water District, California) Emergency Water Supply Facilities of Hachinohe Regional Water Supply Authority Presenter: Kenetsu Kojima (Hachinohe Regional Water Supply Authority, Aomori, Japan) 67

86 3 rd US-Japan Workshop on Water System Seismic Practices 68

87 DAMAGES AND MOTION OF PIPELINES BURIED IN LIQUEFIED GROUND Toshio Toshima,Hiroyasu Ohama,Shogo Kaneko ABSTRACT Many of the water pipelines suffered great damages in the 1995 Kobe Earthquake, especially in the liquefied ground. After that the 2000 Tottoriken-Seibu Earthquake also caused a total of 892 damages on the water pipeline, including water distribution pipes and service pipes in the service area of Yonago City Water Works. The results of damage analysis revealed the high damage ratio in the liquefied ground such as reclaimed ground. This fact reconfirmed the importance of the countermeasures against liquefaction in considering the seismic improvement of water pipelines. This paper describes the results of surveying the damages and behavior of water pipelines buried in the liquefied area at the 2000 Tottoriken-Seibu Earthquake, and also the experiment results using the large scale shaking table with respect to the behavior of pipeline at the lateral spreading ground due to liquefaction. Ductile iron pipelines suffered 17 damages due to slip-out of joint buried in the Takenouchi industrial complex, which was widely liquefied reclaimed ground. These damages occurred at the places where large ground deformation such as ground surface cracks and ground subsidence were found. According to the investigation on the movement of each joint by inserting the TV camera into the target pipeline, it turned out that the joint expanded largely at the place where ground deformation was found. The following results in regard to the behavior and the force acting on pipe body in the pipeline with joints were obtained by the shaking experiment which generated lateral spread of the ground in the pipe axis direction. a) Each joint fully expands one after another. b) The total amount of expansion of pipeline is about half of ground displacement. c) The force acting on pipe body during liquefaction of the ground is less than that in the saturated soil, and depends on the effective load. Toshio Toshima, Section Manager, Ductile Iron Pipe R&D Dept. KUBOTA Corporation2-26 Ohamacho Amagasaki City, Hyogo, Japan, Hiroyasu Ohama, Engineer, Water Works Diagnosis and Information System Dept.KUBOTA Corporation, Shikitsu-higashi, Naniwa-ku, Osaka, Japan, Shogo Kaneko, Engineer, Water Works Diagnosis and Information System Dept. KUBOTA Corporation, Nihonbashi -Muromachi, Cyuo-ku, Tokyo, Japan,

88 INTRODUCTION Water pipelines were severely damaged in the Kobe Earthquake (magnitude 7.2) in 1995, particularly in the liquefied areas. In the Tottoriken-Seibu Earthquake (magnitude 7.3) in 2000, although the conveyance and transmission pipes managed to remain intact, the distribution and service pipes were damaged at a total of 892 locations within the service areas of Yonago City Waterworks Bureau, which supplies water to Yonago City (J.M.A Seismic Intensity Scale: 5 Upper), Sakaiminato City (J.M.A Scale:6 Upper), and Hiezu Village. According to the results of an analysis of distribution pipes with more than 75 mm nominal diameter, it was found that, as was the case in the Kobe Earthquake, the damages were more frequent in the liquefied areas such as reclaimed grounds. 1) Thus, the measures against liquefaction are particularly important in trying to improve the earthquake resistance of water pipelines. In devising the measures against ground liquefaction to protect the buried pipelines, it is necessary to study the motion of the pipelines and the forces applied to them. After the Kobe Earthquake, the motion of the ductile iron pipes buried in the liquefied areas was examined. 2) We, on the other hand, conducted an experiment using a large laminar shear box to examine the motion of the pipes and the forces applied to them when there is land subsidence resulting from liquefaction. 3) This paper discusses the results of our subsequent survey of the motion of ductile iron pipelines buried in the liquefied areas in the Tottoriken-Seibu Earthquake, as well as the results of our experiment to examine the motion of the pipes and the forces applied to them when there is lateral spread of the ground caused by liquefaction. DAMAGES AND MOTION OF PIPELINES IN THE TOTTORIKEN-SEIBU EARTHQUAKE Outline of damages on water pipelines In the Tottoriken-Seibu Earthquake, while none of the conveyance or transmission pipelines in the service areas of the Yonago City Waterworks Bureau (Yonago City, Sakaiminato City, Hiezu Village) was damaged, there were 280 cases of damages on the distribution and service pipes under the roads, and 612 cases in the residential areas. Prof. Hosoi analyzed the damages on the distribution pipes with a nominal diameter of 75 mm or more, and found out that there were 65 cases of damages on the subject pipes. Figure 1 shows the rates of damages on the pipelines according to the degree of liquefaction and their comparison with the similar rates obtained in the Kobe Earthquake. 4) As in the case of the Kobe Earthquake, the rates of damages were higher in the liquefied areas than elsewhere. 70

89 R ate of dam age(dam ages/km ) Tottoriken-Seibu Earthquake Kobe Earthquake Totalliquefaction P artly liquefaction N o liquefaction Figure 1. Rates of Pipeline Damages According to the Degree of Liquefaction Damages on Pipeline in Liquefied Areas There were 34 damages of slip-out of joints, all found in liquefied areas, with regard to ductile iron pipes with general joints. The damages were concentrated particularly in Tomimasu housing complex area, where the ground was prepared by filling a previous sand hole, and in Takenouchi industrial complex area constructed on reclaimed ground. Liquefaction in Tomimasu housing complex was localized. Plotted on Figure 2 are the points of ground deformation and pipeline damages (6 of them). The damages on the pipelines were witnessed in the places where there was cracking or sinking of the ground caused by liquefaction. The entire area was almost liquefied at Takenouchi industrial complex, and the liquefaction led to significant ground deformation, such as. cracking, ground subsidence, and sinking of roads. Plotted on the map in Figure 3 are the places of ground deformation and the points of pipeline damages (18 of them). The damages occurred at places where the roads cracked or sank. Most of the damages were found along the pipelines extending from east to west. 71

90 Types of ground deformation Sand Boil Cracks Opening Level difference slope sinking of road Points of damages on pipeline Points of damages on accessories m Figure 2 Ground Deformation and Pipeline Damages in Tomimasu housing complex Types of ground deformation Pipeline1 Opening,cracking Sinking of roads Lateral spread (Finding by our team) Sinking of roads Pipeline2 Pavement opening and cracking Sinking Pipelines with general joints Pipelines with earthquake resistant joints Points of damages Pipeline m Figure 3 Ground Deformation and Pipeline Damages in Takenouchi Industrial Complex 72

91 Pipeline Motion in Liquefied Areas Prof. Hosoi inserted a TV camera inside a ductile iron pipeline (φ150) at Takenouchi industrial complex and measured the spaces inside each joint (see Figure 4) to examine the amount of joint expansion. The measurements were taken at 2 locations, at Pipeline 1 (extending east-west) and Pipeline2 (extending south-north), as shown in Figure 3. Damages of slip-out of joints were found along Pipeline1. Furthermore, there were more cracks and displacement of curbstones on Pipeline1 than on Pipeline2. Figure 4 shows the measurements of joint expansion. The amount of joint expansion of Pipeline1 (east-west) was large at 40 mm maximum, while that of Pipeline2 (south-north) was small at 10 mm maximum. These results just coincides with the fact that the damages were more severe on the east-west pipelines than on the south-north pipelines. A m ount of joint expansion(m m ) A Space Pipeline 1 Point of dam age (Trace of repairing) B (m ) A m ount of joint expansion(m m ) C Space Pipeline 2 D (m ) Figure 4 Amount of Joint Expansion The Yonago City Waterworks Bureau investigated the motion ofφ150 ductile iron pipelines equipped with earthquake resistant joints in the same Takenouchi industrial complex (Pipeline3 in Figure 3) 5). Figure 5 shows the results of measurement of joint expansion/contraction. The amount of expansion/contraction was large among the pipes found under the places where the pavement was significantly deformed by land subsidence, and rising of 50 curbstones. This finding indicated that the expansion and 20 contraction of the joints helped 10 to absorb the seismic motion of 0 the ground Locations of sinking of pavem ent -60 Locations of opening or rising of curbstones Distance from the reference point(m ) Joint expansion/contraction(m m ) Expansion C ontraction Figure 5 Amount of Joint Expansion/Contraction 73

92 EXPERIMENT ON PIPELINE MOTION UNDER LATERAL SPREAD OF THE GROUND 6) Experiment Method Experiment Apparatus Figure 6 shows the apparatus used in the experiment. A box filled with soil was mounted on an shaking table and a lateral spread was artificially generated along the pipe axis inside the box to investigate the motion of the pipe and the soil, as well as the forces applied to the pipe. The soil box was divided into two layers. The upper layer represented liquefied ground (relative density: 40%). The lower layer represented non-liquefied ground (relative density: 80%). An 8 degree slope was formed along the border between the upper and lower layers to facilitate the flow of the liquefied ground. 4m J1 J2 J3 J4 J5 0.5m Movable wall Dynamic actuator 2m 1.35m P1 P2 P3 P4 P5 Liquefied ground Non-liquefied ground 8 Shaking table(6m 4m) Soil box Shaking direction Figure 6 Experiment Apparatus Experiment Method The soil box was shaken by a sine wave (5 Hz, 200 gal maximum). Immediately after liquefaction had started (approx. 22 seconds after starting the shake), the movable wall was pulled back by 260 mm at 10m/s by the dynamic actuator to cause a lateral spread in the direction of pipe axis. The shake of the soil box was orthogonal to the pipe axis so that the motion of the pipe would not be affected by inertia. 74

93 Test Pipe A total of 5 earthquake resistant joint ductile iron pipes(φ75 length 600mm) (P1 to P5) were buried at 0.5m from the ground surface. Figure 7 shows the structure of the model for the actual earthquake resistant joint used in the experiment. The joint section contracts/expands by ±10 Displacement guage Load cell φ75 ductile iron pipe mm, and when the joint is fully expanded, the clamp automatically prevents slip-out of joint. The load cell and the displacement gauge were installed at each joint to measure the force applied to the joint and the amount of expansion of the joint. Figure 7 10mm 10mm Clamp Structure of Model Joint Horizontal Displacement of Pipe and Ground Figure 8 represents the amount of expansion of each joint (J1 J5). The joint J2 began expanding first, in approximately 25 seconds since the shake was started, and fully expanded in 32 seconds. The other joints gradually began to expand from around 30 seconds after the beginning of shake, and fully expand consecutively in the order of J4, J3, J1 then J5. The amount of horizontal displacement of the pipe was obtained by adding up the amounts of displacement of the joints. Figure 9 shows the measurement results of horizontal displacement of Pipe 1 to Pipe 4 (P1-P4) and the corresponding Joint expansion(mm) Fully expand J2 8 J3 6 J5 4 J4 2 J Time(s) Figure 8 Joint Expansion displacement of the ground (G1-G4). The amount of horizontal ground displacement was 70~200 mm, and the amount of displacement increased toward the movable wall. The fact that the amount of displacement of the pipes was smaller than that of the ground indicate that slippage probably occurred between the pipe and the ground immediately the liquefied ground had begun spreading. Figure 10 shows the rate of horizontal displacement of Pipe 4 (P4) against the horizontal displacement of the ground near the movable wall (G4). The displacement of P4 was smaller than that of G4, with the former being 0.4 to 0.6 of the latter. 75

94 Horizontaldisplacem ent(mm) Time(s) Ground displacem ent G1 G2 G3 G4 Pipe displacem ent P1 P2 P3 P4 Pipe displacem ent/ Ground displacem ent Fully expand Time(s) Figure 9 Horizontal Displacement of Pipes and Grounds Figure 10 Rate of Pipe Displacement Against Ground Displacement Force on Pipes Force on Joints Figure 11 shows the measurement results of the forces applied to the joints. In approximately 37 seconds after starting the shake, the joint fully expanded, and the force began to apply to the joint immediately after that. The maximum amount of force was recorded in about 40 seconds from the start, or 3 seconds after the joint fully expanded. This indicates that when there is a lateral spread of the ground, there is a slight time lag between full expansion of the joint and the maximization of the applied force. Figure 12 are the results of obtaining the maximum value of the force applied to the pipe by subtracting the amount of force on the adjacent pipes. While a maximum force of around 1.2kN was obtained for P5, it was found that little force was applied to P1 to P4. This finding indicates that the entire pipeline was pulled by the force on P5, and this force on P5 was successively transmitted from one joint to another. Relationship between Effective Stress and Force on Pipe Force on Joint(kN) Maximum force J2 Fully expand J Time(s) Forces A pplied to the Pipes(kN ) Forces A pplied to the J oints Pipe1 Pipe2 P ipe3 Pipe4 Pipe5 Movale Wall Figure 11 Measurement of Forces Applied to the Joints Figure 12 Maximum Force on the Pipes Figure 13 shows the measurements of excess pore 76

95 water pressure of the ground around P1 and P5. While the excess pore water pressure around P1 gradually dispersed after the liquefaction had set in, the pressure around P5 dispersed abruptly all of a sudden. The excess pore water pressure around P1, P3 and P4 also gradually dispersed like that around P2. This indicates that since liquefaction continued while the excess pore water pressure remained high in the case of P1 to P4, almost no force was applied to the pipes themselves. In contrast, since the strength of the surrounding ground was recovered in the case of P5, more force from the ground was applied to the pipe than in the case of other pipes. Figure 14 shows the relationship between the effective ground stress and the force applied to pipe 5, for the period from 25 seconds to 40 seconds after the start of shake, when the excess pore water pressure was dispersed. The effective stress is normalized by the initial overburden pressure, and the force applied to the pipe was normalized by using the value (2.0 kn) of the saturated sand ground obtained in a separate preliminary experiment. The figure shows that the force on the pipe increased as the effective stress of the ground increased. Excess pore water pressure(kpa) Pipe Pipe Time(s) Figure 13 Measurements of Excess Force applied to pipe * *1 N orm alized by the force of saturated sand *2 N orm alized by the initialoverburden pressure before liquefaction Effective stress of the ground *2 Figure 14 Relationship between Effective Stress and Force on the Pipe CONCLUSION The results of the survey of the motion of ductile iron pipes buried in the liquefied areas at the Tottoriken-Seibu Earthquake were presented along with the results of the shaking experiment regarding the motion of the pipes and the force applied to them in the event of lateral spread of the ground caused by liquefaction. Below summarizes the findings of the survey and the experiment: 77

96 1)The rate of damages on pipelines was greater in liquefied areas in the Tottoriken-Seibu Earthquake 2000 as was also the case in the Kobe Earthquake. 2)There occurred the damages on pipelines and the expansion of the pipes was large at the locations where the roads were cracked or sank in the liquefied areas. 3)The joints fully expanded successively one after another after the occurrence of lateral spread. 4)The final amount of joint expansion was approximately half of the ground displacement. 5)The force applied to the pipe by liquefaction was smaller than that in saturated soil and was dependent on the amount of effective stress. References [1] Yoshihiko Hosoi, Analysis of Damages of Water Pipelines in the Tottoriken-Seibu Earthquake 2000, Journal of Japan Waterworks Association, Vol.71, No.2, (in Japanese) [2] Hisato Miura, Survey of the Motion of Earthquake Resistant Joint Ductile Iron Pipes in the Kobe Earthquake, Journal of Japan Ductile Iron Pipe Association, No. 61, (in Japanese) [3] Toshio Toshima, Hiroyasu Ohama, Damages on Water Pipelines and Ground Deformation in the Kobe Earthquake in 1995, The 4 th International Waterworks Symposium, (in Japanese) [4] Japan Waterworks Association, Analysis of Damages on Water Pipelines in the Kobe Earthquake in 1995, (in Japanese) [5] Yoichi Mishima, Mitsuaki Kato, Masamitu Ebara and Takeshi Ishida, Survey of Motion of NS Type Ductile iron Pipelines in Liquefied Reclaimed Areas in the Tottoriken-Seibu Earthquake in 2000, Journal of Japan Ductile Iron Pipe Association, No. 70, (in Japanese) [6] Sakatoshi Mori, Toshinori Kawabata, Hiroyasu Sato, Toshio Toshima, A Research on the Motion of Ductile Iron Pipes in Lateral Spread Ground, The 52 nd Japan Waterworks Research Conference, (in Japanese) 78

97 SEISMIC EVALUATION OF WATER SUPPLY SYSTEM IN HEALTH CARE FACILITIES Masakatsu Miyajima, Nebil Achour and Masaru Kitaura ABSTRACT This paper is focusing on seismic safety of water supply system in health care facilities. First, damage to lifelines in health care facilities during the past damaging earthquakes in the world was shown and the effects of the damage on functions of the facilities were considered. Second, the performance of hospitals in Kobe City just after the 1995 Hyogoken Nambu Earthquake was examined by using the results of questionnaire survey conducted by authors. According to the results of the questionnaire survey, it is clarified that the performance of water supply strongly affects the treatment in hospitals just after an earthquake strongly and the water is necessary for not only a treatment but also cooling machines, washing appliances and hands of doctors and nurses, ordinary life of inpatients and so on. Moreover, a method of seismic evaluation of water supply system in health care facilities was studied. Masakatsu Miyajima, Professor, Department of Civil Engineering, Kanazawa University, , Kodatsuno, Kanazawa, Japan Nebil Achour, Graduate Student, Graduate School of Natural Science and Technology, Kanazawa University, , Kodatsuno, Kanazawa, Japan Masaru Kitaura, Professor, Department of Civil Engineering, Kanazawa University, , Kodatsuno, Kanazawa, Japan 79

98 INTRODUCTION After the occurrence of an earthquake, a city may suffer damage to many or all of its infrastructures. Hospitals and health care facilities, which are supposed to take care of injured people, in some cases cannot even function. In this paper the importance of lifelines in health care facilities and hospitals is discussed and some examples of past and recent earthquakes are considered such as Kobe 1995, Taiwan 1999, Turkey 1999, Peru 2001, and Algeria The conclusion of this research will be a new method to protect hospitals from disasters and in particular, earthquakes. SYSTEM OF HOSPITAL A hospital is a very complicated system, consisting of two different parts: a physical part and a human part. The physical part is composed of structural and nonstructural elements while the human part is comprised of the staff (doctors, nurses, administrative staff and support staff). A hospital is different from any other facility because it cannot function if any of its elements are affected. The following diagram shows the different parts of a hospital: Doctors Nurses Support staff Administration Patient in Hospital Services Patient out Structural elements Nonstructural elements Others Figure 1. System of hospital 80

99 PERFORMANCE OF HOSPITALS IN PAST EARTHQUAKES Kobe earthquake, Japan 1995 After the earthquake of Kobe 1995, the city suffered damage to all its facilities. Although mass destruction was widespread the earthquake left its mark on two hospitals in particular. Both hospitals experienced damage to both structural and nonstructural elements. In the Kobe Medical College Hospital, some local damage was found in a section of wall as a result of the interaction between its two buildings, concrete beside one joint burst and damage to the oxygen distribution system caused a drop in the pressure of the system. The failure of the water system caused the emergency generators to switch off, as the water had been used to cool the engines. The hospital is composed of three large buildings all of which were connected to the same water system and none of which had any storage facility for water. There was no problem with the telecommunication system within the hospital. However the whole city s system was hampered. As in the Kobe Medical College Hospital, water in the Hyogo Medical Centre was also used for cooling the emergency electric power generators. The loss of water caused the emergency generators to shut down. However at this second hospital a water tank was placed on the roof of the hospital. However due care was not taken and as a result of the shaking the tank fell and caused flooding of part of the hospital. As a result of the damage to these two hospitals, they couldn t offer the necessary appropriate treatment to their patients due to the lack of water, electricity and gas. The loss of water was devastating and forced some hospitals to transfer patients to other hospitals. The loss of communications within the city made contact between hospitals very difficult [1]. Kocaeli earthquake, Turkey 1999 The case in Turkey was worse than that of Kobe. Hospitals were reported to be under siege from thousands of injured people. The majority of them were treated in tents set-up in the parking lots. In this paper 4 hospitals are considered: Izmit SSK Hospital, Izmit State Hospital, Adapazari SSK Hospital and Adapazari State Hospital. In all these facilities, the electric power was cut. In some however, emergency electric generators were sufficient to continue the supply of electricity. Internal and external telecommunications were cut; in some cases walkie-talkies were the only tool of communication. All the oxygen cylinders toppled over and were displaced; however no explosion or leaking was reported. Medical equipment fell down as a result of bad anchorage. The majority of injuries were treated in parking lots, corridors, transferred to other hospitals (in Istanbul, Bursa and Ankara) or simply turned away because there was absolutely no room, no supplies and even no physicians available to treat them [2]. 81

100 The Chi-Chi earthquake, Taiwan 1999 More than 4,000 health care facilities were affected as result of the Chi-chi earthquake of 1999 in Taiwan; 165 of them were hospitals. Nonstructural elements were considered to be the main factor that hampered the functioning of the major hospitals. The following are some examples of damage caused by the earthquake to the following hospitals; Christian Hospital in Puli; Veterans Hospital in Puli and Shiu-Tuan Hospital in Tsushan. All these facilities suffered communication failure, falling debris, damage to pipes, toppling of gas cylinders and liquid storage tanks, damage to medical equipments among others. In all the mentioned hospitals no emergency evacuation plan had been implemented. Trauma to the majority of patients was the result of their evacuation and relocation from one hospital to another. The capacity of the Veterans hospital had been reduced by 50% of it original capacity and in the Christian hospitals the reduction was about 10% of its original capacity. The Shiu-Tuan hospital, which was built near the Chelungpu fault two years before the occurrence of the earthquake, was entirely evacuated and then closed [3]. Peru 2001, India 2001 and Algeria 2003 In Peru as result of the earthquake of 2001, a hospital in Arequipa had to be partially evacuated since the occurrence of severe structural damage. In the earthquake of India, 2001, damage to hospitals occurred. All three hospitals in the region of Buhj, India completely collapsed as a result of being so close to the epicenter of the 2001 earthquake. This tremendously affected the treatment of patients. In Algeria, after the quake of 2003, one small hospital in Boumerdes was completely destroyed. Other hospitals in the region of the capital, Algiers, were unable to treat such large numbers of injuries due to their lack of preparedness. Damage to Hospitals after Some Past Earthquakes Table 1 summarizes all the previous damage experienced in the cited hospitals. Not only damage to each component but also interaction between each component is important such as cooling water for an engine of emergency generators. A correlation diagram of function of each lifeline during earthquake should be illustrated in the next step. QUESTIONNAIRE SURVEY IN KOBE CITY A questionnaire survey for performance of hospitals just after the Kobe earthquake, Japan 1995 was conducted. One hundred and nine questionnaires were delivered to the hospitals located in Kobe city. Twenty-six questionnaires were sent back by facsimile, so the withdrawal rate of questionnaires was 24%. Since the 82

101 TABLE 1. DAMAGE TO HOSPITALS AFTER SOME PAST EARTHQUAKES Kocaeli earthquake (M=7.4) Kobe earthquake (M=7.2) Peru earthquake (M=8.4) Chi-chi earthquake (M=7.6) Izmit Adapazari Kobe Arequipa Puli Tsushan Cause of Disruption and Evacuation Izmit SSK Izmit State Adapazari SSK Adapazari State Kobe Medical College Hyogo Medical Centre Peru Christian Veterans Shiu-Tuan Backup Power Outage X X X X X Water Supply Outage X X X X X Gas Service Outage X X X Broken Piping, Water X X X X Leakage Toppling of Gas, Liquid X X X X X X X X Storage Tanks Elevator Damage X X X Communication Failure X X X X X X HVAC Anchorage Failures X X X Mechanical Damage Medical Damage Equipment Equipment X X X X X X X X Falling Debris X X X X X X X X Emergency Evacuation X X X Plan Not in Place Transfer of patient X X X X X X X X Structural damage X X X X X X X X statistical analysis cannot be conducted because of small number of samples, the results of questionnaires are shown and an importance of water in hospitals is discussed. Figure 2 illustrates the performance of water supply after the event. The hospitals of more than 60% could not obtained water at all and 27% were not affected of suspension of water supply. Figure 3 shows a situation of treatment at the hospitals in water suspension. Twenty percent of all hospitals could not treat at all and half of hospitals were damaged to their activity by water suspension. These figures suggest that the shortage of water directly affected the performance of hospitals after the event. According to the answers, water was needed for not only a treatment but also cooling machines, washing appliances and hands of doctors and nurses, ordinary life of inpatients and so on. Figures 4 and 5 show the volume of required water in a surgery and hemodialysis. These figures indicate that about 50 liters in a surgery and about 83

102 ordinary supply suspenssion shortage 0% 20% 40% 60% 80% 100% Figure 2. Performance of water supply just after the earthquake no treatment ordinary treatment with difficulty 0% 20% 40% 60% 80% 100% Figure 3. Affects of water suspenssion on treatment others 150L 100L 50L 10L Frequency Figure 4. Volume of required water in a surgery 84

103 others 150L 100L 50L 10L Frequency Figure 5. Volume of required water in a hemodialysis w ater tank w ell others 0% 20% 40% 60% 80% 100% Figure 6. Emergency water in hospitals 150 liters in a hemodialysis are necessary. Figure 6 illustrates preparedness for water in hospitals before the earthquake. The hospitals of more than 60% store the water in a water tank that installs at underground or the roof. The capacity of water tank was not enough for long suspension of water according to circumstances. There were some cases that the underground water tank was broken because of no seismic design. According to the results of the questionnaire survey, it is clarified that the performance of water supply strongly affects the treatment in hospitals just after the earthquake and the water was necessary for not only a treatment but also cooling machines, washing appliances and hands of doctors and nurses, ordinary life of inpatients and so on. 85

104 METHODOLOGY The methodology is studied to evaluate the performance of water supply system as a lifeline just after an earthquake. First, a critical system and components are identified and assessed. The interaction between each component should be considered here. Next, the fragility curve of each critical component is established by using the data of damage in the past earthquakes. The location of the whole facility with regard to seismological zones and the location of each component within the facility itself should be considered. Then the risk analysis is conducted. The final step requires the judgment of the best way to reduce the risk. Many possibilities are available such as: relocation, replacement, strengthening, substitution, etc [4]. CONCLUDING REMARKS The lessons we can get from the cited earthquake experiences above are as follows: health care facilities, which are typically expected to receive and treat injuries, can be affected by malfunction of lifeline after earthquakes; and closing or reducing the capacity of a hospital has remarkable impact on treatment of patients and victims alike. Therefore preparation of hospitals is an obvious step that should concern both physical and functional parts of lifelines in all hospitals. With regards to the physical part, both structural and non-structural elements have to be protected. As it was previously stated, this methodology considers life saving as its main goal unlike other methods that consider economical savings as more important. REFERENCE [1] Shinozuka, M. et al., 1995, The Hanshin-Awaji Earthquake of January 17, 1995 Performance of Lifelines, Technical Report NCEER, [2] Mark A. Pickett, 2000, Hospital with Emphasis on Lifeline, Izmit (Kocaeli), Turkey earthquake of August 17, 1999, Including Duzce earthquake of November 12, 1999 Lifeline Performance, ASCE, Technical Council on lifeline, [3] George C. Lee et al., 2000, The Chi-Chi, Taiwan Earthquake of Sept.21, 1999, Reconnaissance Report, Tech. Report MCEER, [4] Gayle S. Johnson et al., 1999, Seismicity Reliability Assessment of Critical Facilities, A handbook, Supporting Documentation, and Model Code Provisions, Technical Report MCEER

105 The Research of Damages of Public Water Supply Pipelines During the Ji-Ji Taiwan Earthquake on September 21, 1999 Ping-Hsin Wang ABSTRACT On September 21, 1999, 1:47 am, seconds, Taiwan experienced an immensely strong earthquake with a magnitude of 7.3 on the Richter scale. The epicenter of the earthquake of was located 12.5 kilometers west of Sun Moon Lake; depth of the hypocenter was 1.1 kilometers, which is considered a shallow earthquake. More than 10,000 aftershocks occurred in the ensuing days; many of these aftershocks registered magnitude greater than 6 on the Richter scale. The quakes epicenters were largely concentrated in Hsuehshan (Snow Mountain) and the Central Mountain Range. Aside from the toll it took on lives and properties, the water supply system of the calamity areas that were under the Taiwan Water Supply Corporation sustained very serious damage, especially the water supply systems of Taichung County (City), Nantou County, and other areas that were located along the fault lines; water supply facilities in these areas collapsed, leaving them totally without water. The damage and destruction brought about by this natural calamity is something that has been rarely seen in Taiwan hundred-year old history. It brought to light the inability of water supply facilities to resist or cope with earthquake; in reality, the consequences go far deeper than this. In view of this, the Taiwan Water Works Association is aggressively pursuing researches related to these issues. This study analyzes the results from various analyses, including: A. Damage survey statistics based on the various types of water supply system. B. Damage survey statistics according to the magnitude of the earthquake C. Damage survey statistics based on specific areas along the fault lines. D. Damage survey statistics according to the various types of piping materials. E. Damage survey statistics according to the nature of earthquake damage Based on the analysis of survey statistics, the study aims to further explore the reasons behind the damage to the water supply facilities during this earthquake. Aside from the intense magnitude of the earthquake and the fact that the facility was directly on the fault lines, which may be considered natural factors, other factors, including manmade ones that are related to the design, engineering process, management, and piping materials are also important factors contributing to the serious damage on hand. What is important now is to focus on how to improve water facility design, management, supervision, and assess construction in order to maintain a standard of quality. At the same time, attention must be given to the selection of appropriate materials, taking into consideration resistance to earthquakes; these measures may lead to a reduction in the loss of life and property. Ping-Hsin Wang, Manager, Administration Office of the Eighth District Taiwan Water Supply Corporation, 265 Gin-Shang West Road, Yuang-Shang, I-Lan County 264, Taiwan, R.O.C. 87

106 INTRODUCTION The fault area of the Ji-Ji Earthquake lies along the Chelungpu and Hsuangtung Tamaopu Faults. Tamaopu s Hsuangtung fault, which lies to the west of the quake s epicenter moved and pushed westward, resulting in the displacement of the Chelungpu Fault. The Chelungpu Fault stretches from Fengyuan, Dantzu in the north, passing through Taiping, Dali, Wufeng, Tzaotun, Nantou City, Mingchien, Chushan, and ending at Yunlin County s Kukeng. An influence parallel line reaching 983 cm /sec 2, it produces a destructive area of about 100 kilometers square; with the upper plates pressing down the lower plates, a great burst of energy was released from below the ground. Tungshih, Guohsing, Chungliao, and Ji-Ji were located along the Tamaopu Fault, and sustained great damage. The Chelungpu Fault extended 45 degrees eastward from the northern tip of Fengyuan to Tzaolan in Miaoli County; the fault has also, together with Tamaopu Shuang Tung Fault, pushed northward together, trapping Tungshih, Shihkang, and other areas, where it created the most serious damage.(see pictures 1 and 2) Picture 1 Fengyuan First Water Filtration Plant Damage Situation of 2000mmSP Distribution Pipes (crushed and crumbled) Picture 2 Fengyuan First Water Filtration Cracked Joint in the 2000mm DIP Filtration Pipe The scope of the research is the governmental released report on the areas affected by the September 21 earthquake (Hsinchu to the north, northeastern part of Kaoshiung County to the south, areas near Changhua to the west, and the Central Mountain Range to the east; a length of 150 km. And a width of 80 km.); it does not cover the whole province (see Illustration 1). The mountainous areas and remote areas where pipes have not been installed have been deducted from the actual number of pipes installed in the water supply area, which has been modified to become more accurate. Upon revision, the actual network acreage of areas supplied by the pipes is pegged at kilometers square (about hectares). 88

107 Damage Survey Analysis of Various Types of Water Supply Systems 86 water supply systems comprise the scope of the survey. Among these, 42 of them show no sign of damage, the other 44 shown varying degrees of damage. The degree of damage is classified to the decimal system. Looking at the preliminary damage systems, details of the damage survey analysis can be seen in Table 1. (llustration 1) Table 1. Damage Survey Analysis Statistics Number of Damaged Systems in Varying Degrees Damage rate(case/km) Distribution Service for consumer Remarks 100X 6 10X 7 X 8 16 X/ X/ X/ No damage Damage Survey Analysis of Earthquake Magnitude If the damage rate resulting from an earthquake below a magnitude of intensity 4 is to be used as a basis in comparison to the damage rates of intensity 5 and 6, the ratios are as follows: distribution pipelines, 1:13.7:137; service pipeline for consumers, 1:74.5:403. The analytical results of the damage system based on the survey (which does not include damage sustained by other systems), again based on the damage rate resulting from an earthquake below a magnitude of intensity 4 that is used as a basis in comparison to the damage rates of intensity 5 and 6, the ratios are as follows: distribution pipelines 1:1.8:10.3; service pipelines for consumers, 1:5.0:19.3. This proves that there is a direct relationship between the damage rate and the magnitude of the earthquake; the damage rate increased as the magnitude of the earthquake grows stronger. Details of the damage survey analysis can be seen in Illustrations 1 and 2. 89

108 (Illustration 2) (Illustration 3) Survey Analysis of the Damage Rate of Each Type of Piping Material Vis-à-vis Specific Earthquake Magnitude Level Table 2.Survey Analysis of the Damage Rate of Each Type of Piping Material Vis-à-vis Specific Earthquake Tremor Level Kinds of Pipes Below Intensity 4 Intensity 5 Intensity 6 CIP SPS PVCP PSCP GIP PVCP-PE FRP SSP DIP PCCP HDPE ACP Categories Others Remarks Numerical value of 0 means no damage; there is no such material. ABSP number of cases with no damage. 90

109 In summary, the damage sustained by each type of piping material is as follows: HDPE for intensity 6( case/km), ACP for intensity 5( case/km), CIP for intensity 4( case/km); these materials are ranked on top. For distribution pipes, because there are only 4 kilometers of ACP left and pipes made of up for replacement due to constant leaking as a result of bad connection, although the extent of damage is wide, they are widely used enough to merit analysis. As for GIP and CIP, the former is primarily installed in mountainous areas, which make sit more special; CIP is not included in the new focus. Aside from the abovementioned four types of pipes, the damage sustained by pipes made of DIP and PVCP, which are the most common types, are more serious. As for PVC and DI pipes, which are the most frequently used types, DIP is more resistant to earthquakes than PVCP, theoretically speaking. However, we can see from Table 2, the damage sustained by DI pipes during the Ji-Ji Earthquake, at a magnitude of intensity 6, was , higher than PVC pipes at At a magnitude of intensity 5, damage sustained by DIP (0.0406) is lower than PVCP (0.0707). At a magnitude of intensity 4 and below, DIP did not sustain any damage, PVCP sustained damage of The fact that DIP cannot withstand earthquakes with a magnitude of intensity 6; many of the joints of these pipes become disconnected should be carefully studied for improvements. Damage Survey Analysis of Specific Area Along the Fault The study discovered that the damage sustained by the water pipes along the Hsuangtung Fault is the most serious, followed by that of the Chelungpu Fault as shown in Illustration 4. (Illustration 4) Damage survey analysis has been done according to different types of piping materials, pipe diameter, nature of sustained damage, and usage. Results are shown in Tables 3-1, 3-2, 3-3, 3-4, 3-5, as well as Illustrations 5 and 6. 91

110 Table 3-1.Damage Survey Analysis on the Types of Piping Material for Distribution Pipes Type of Length of pipes Pipes km Case/ km HDPE ACP DIP GIP CIP PVCP SP C FRP PSCP PVCP-PE ABSP 21 0 RCP 58 0 SSP PCCP 24 0 Others Categories Remarks: ABSP, RCP, SSP, and PCCP have no cases of direct damage Table 3-2 Damage Survey Analysis of Server Pipeline s Main Pipes According to Type of Pipe Material Service Type No. of cases Case/ km PVCP LP PBP GIP Damage rate of consumer server pipeline has a union of case/household per thousand, which is not that serious. Results are shown in Table 3-3. Table 3-3. Damage Survey Analysis Illustration of Server Pipeline Parts of Categories case/house Categories case hold per thousand Union Stop cock Corporation cock Meters Others Stub

111 Tale 3-4. Damage Survey Analysis of Distribution Pipe Diameter Case/ Diameter Cases KM 以 下 以 上 Table 3-5. Damage Survey Analysis of Pipe Diameter of Server pipeline Diameter Cases Case / KM (Illustration 5) 93

112 (Illustration 6) Damage Survey Analysis According to Forms of Earthquake Damage Table 4. Damage Survey Analysis of Forms of Earthquake Damage Forms of Earthquake Distribution Service Damage No. of cases % No. of cases % Tremors % % Collapse % % Sinking due to % % Others % % Total % % 94

113 Conclusion From the cross analysis of the different types of pipe material and types of damage, the study has gained an understanding of the interactive relationship between the two. The results have been classified as follows: On damage rate of distribution pipeline Damage rate for distribution pipes which are below 450 m/m is 21 times (0.8574/ )that of the distribution pipes above 500 m/m. This is because building standard requirements for distribution pipes are stricter; the structural design of the big pipes being better has no bearing on it. Disconnection of joints on DIP type of pipes, which accounted for 60% ( case/km), is ranked No. 1, followed by damage sustained by main pipes, which accounted for 30%; other damaged parts at 9% and joint damage at 1%. From this, one can see that the main reason for DIP damage is due to joint disconnection. This situation exists because the joints currently used are A and K types; these types have lower earthquake resistance, especially Type A that easily disconnects. Plans must be made to procure S, S2, and NS, as well as other earthquake resistant anti-disconnect joints to be more cost-effective. On damage rate of outer consumer pipelines 13,510 cases of damage sustained by outer pipes for consumers account for 78.13%, which is the main reason behind the delay in repairs during this earthquake. 99% of the consumer service pipelines are PVC pipes; damage sustained is four times that of distribution pipes(0.8698/ ). The service pipeline joints use TS; so that the damage or disconnection of consumer service pipeline and the parts of the pipes around the joints are significantly serious. The joints of the TS connectors lack of flexibility; consequently, during an earthquake, damage is sustained not only be the joints, but also by the surrounding parts of the pipe. It is therefore concluded that improvement need to start from the joint. Comparison of the damage rate of PVCP and DIP pipe materials and pipe joints The damage rates of PVC pipes during this earthquake are as follows: distribution pipelines case/km. The damage sustained by the main pipes at case/km ranks first, followed by damage sustained by corporation cock at case/km. At the same time, the study discovered that the damage sustained by PVC pipes is lighter than that of DIP pipes. The reason for this is that connectors usually use rubber ring type connectors (RR) similar to T-shaped connector. Although the disconnection discovered in the joint and plugs is only 5%, however many serious damages were sustained by the main body of the pipe. From this, one can see that the earthquake resistance (more than Intensity 6) of this type of pipes is below: Details of the abovementioned situation are shown in Tables 5-1, 5-2, and

114 Type of Distribution Pipe (case) Table 5-1. Cross-Analysis of Damage Rates of Different Pipeline Materials and Damage d Pipes (main line) % Classification- Distribution Pipes(Case/km) Damage d parts (missha pen pipes) % Joint Disconn ection % Joint Damage % Categor ical Damage % Others % Total HDPE ACP DIP GIP CIP PVCP SP FRP PSCP PVCP-PE ABSP RCP SSP PCCP Categories Others Remarks Ranki ngs Red-colored letters represent the focus of the study; no further study is made on the other special reasons behind the damage sustained by pipes made of other materials. Service - Types of Pipe Table 5-2. Cross-Analysis of Damage Rates of Different Pipeline Materials and Classification- Service pipelines for Consumers ( Case/Km) Damage d pipes % Damage d parts % Joint disconn ection % Joint Damage % Categor ical damage % Others % Total PVCP LP PBP GIP HDPE Red-colored letters represent the focus of the study; no further study is made on the other special reasons Remarks: behind the damage sustained by pipes made of other materials. Ranki ngs 96

115 Service Category Table 5-3. Cross-Analysis of Damage Rates of Different Pipeline Materials and Damage d pipes Classification- Categories(Case/household per thousand) % Damage d parts % Joint disconn ection % Joint damage % Categor ical damage % Others % Total Union Stop cock Corporatio n cock Water meter Others stub Remarks Ranki ngs Red-colored letters represent the focus of the study; no further study is made on the other special reasons behind the damage sustained by pipes made of other materials. Factors Affecting the Damage Rates of Water Pipes We have tried to look at the damage occurring in the aftermath of this earthquake by cross analyzing soil liquefaction, landslide, ground cracks, and other occurrences as well as the damage rates, the results of which are seen in Illustration 7. In terms of damage sustained by the distribution channels, the area damage rate of liquefaction and ground cracks, case/km, is the most serious; the main affected line was the Wufeng water supply system. In terms damage sustained by service pipelines for consumers, area damage rate of liquefaction plus ground cracks plus landslide, case/km is the most serious; main affected lines are the water supply systems of Tungshih, Nantou, Mingchien. Upon further inspection of the heavily damaged mentioned above, it was discovered that the following factors affect the damage rates of these pipelines: 1. Long-term ground displacement, the larger the displacement the higher the damage rates. 2. The larger the ground tremor motions are, the higher the damage rates. 3. For pipeline materials, material can be extended sustain less damage. 4. For pipe joints, steel joints sustain more damage than softer ones. 5. Continuous pipelines are more earthquake resistant than segmented ones. 6. Pipes which are rusty and eroded or where the materials are already aging may have higher damage rates. 7. The larger the pipes are, the lower is the damage rate. 8. Pipeline in weaker ground or in paces with geological structure can have higher damage rates. 97

116 (Illustration 7) The study, which focuses on the changes in the form and shape of the channels, the ground as well as its location changes, should be given priority in when designing and planning measures for future earthquake damage control. This is especially true when considering the use of certain types of pipeline materials, joints, and as well as the drawing up of construction regulations and strategies. Acknowledgment I would like to extend my gratitude to the Taiwan Water Works Association and Taiwan Water Supply Corporation for their assistance and also to Mr. Chia-Yan Liu, who helped prepare the drawings. References 1. Central Geological Survey, Ministry of Economic Affairs. Geological Survey Report on the September 21 Earthquake, Committee on Public Construction, Executive Yuan. A Study on the Earthquake Resistance of Underground Life Maintenance Systems, Taiwan Water Works Association, R.O.C. Report on the Damage Sustained by Water Supply Pipelines During the September 21 Earthquake, Taiwan Provincial Government. Observation Report on Damage Sustained During Kobe Earthquake (on the part of water supply and gas), Liu, Chia-shiao, Wang Ping-Hsin, Taiwan Water Works Association, R.O.C.: Water Supply Journal, 1 st issue (Trend of pipe joints), Wang, Fu-Sang, Taiwan Water Works Association, R.O.C: Taiwan Water Supply Magazine, 2003 edition (with illustrations), Training materials for quality assurance engineer, Committee on Public Construction, Executive Yuan: Analysis and Solution for Engineering Quality Issues, Wang Ping-Hsin, Taiwan Water Supply Corporation: 22 nd Statistical Report on the Water Supply Business in Taiwan,

117 Emergency Operation Planning - How Contra Costa Water is Building Earthquake Response Capabilities In Calm to Excel in Response Under Emergency Stephen J. Welch, P.E., S.E. ABSTRACT For the Contra Costa Water District (CCWD) the responsibility of managing a Northern California water utility with over 450,000 customers means ensuring the highest level of service economically feasible under all conditions. As with most utilities in California, the planning and operations of a utility system in earthquake country means added attention to the details of design and construction of facilities to ensure system operation following a major earthquake event. Following the October 17, 1989 Loma Prieta Earthquake, CCWD aggressively increased its planning for system operation following a major earthquake event. Although the CCWD system experienced only minor damage from the Loma Prieta earthquake, the event instigated a significant program to increase CCWD earthquake readiness. From 1994 to 2002 CCWD implemented a $120 million capital improvement program to improve its various raw and treated water systems to ensure its core water delivery functions could continue following the design level seismic event. Having completed its capital improvement program, CCWD recently has updated its overall organizational approach to emergency response to a major seismic event. The result has been a program that invests over $40,000 per year for emergency response and preparedness training (approximately 25-percent of CCWD s training budget). In addition, CCWD has recently completed the design and is moving into construction of a $450,000 state-of-the-art Emergency Operations Center designed to the State of California s Office of Emergency Services standards for a 2,500-year return period earthquake. With this new facility and regular, high quality emergency response training, CCWD plans to ensure that not only will its physical facilities be reliably operational following any foreseeable seismic events, but that its organization as a whole will be prepared to respond with effectiveness in worst case scenarios. OUTLINE OF CONTRA COSTA WATER DISTRICT The Contra Costa Water District (CCWD, District) provides raw and treated water service to over 450,000 customers in Contra Costa County, California. The eastern part of CCWD s service area is currently one of the fastest growing areas in California with over 200- percent growth in population in the last decade. Its primary water conveyance system is a 65- year old, 85-kilometer long, open, concrete lined canal which serves two CCWD water treatment plants, and various cities, municipalities, industries and agricultural customers. CCWD was established in 1936 to provide water to the central and northeastern regions of Contra Costa County, California (Figure 1). CCWD draws its water from the Sacramento-San Joaquin Delta under a contract with the U.S. Central Valley Project (CVP). As part of the CVP, the 85-kilometer Contra Costa Canal was built in By means of the canal, water is diverted 99

118 from Rock Slough (13-kilometers east of Antioch, California) through a 7-kilometer unlined channel into a 78-kilometer concrete-lined facility. Four pumping stations lift water 37.8-meters above sea level to the canal's Antioch summit, after which gravity propels the water to its terminus in Martinez. The canal runs through Oakley, Antioch, Pittsburg, Concord, Walnut Creek, Pleasant Hill, Pacheco, ending at a terminal reservoir in Martinez (Figure 2). Figure 1 Vicinity Map P i t t s b urg M a r t i n e z P l e a s a n t H i l l W a l n u t C r e e k Co n c o r d C l a y t o n Antioch Brentwood Los Vaqueros Reservoir Figure 2 Service Area CCWD has four raw water storage reservoirs: Martinez, Contra Loma, Mallard, and Los Vaqueros, totaling a storage capacity of 1,233,487-litres (103,070 acre-feet). It also has two water treatment plants; the Bollman Water Treatment Plant in Concord, with a capacity of 284 million litres per day; and the Randall-Bold Water Treatment Plant in Oakley, with a capacity of 150 million litres per day. Half of CCWD customers receive treated water directly from CCWD, 100

119 the remaining from 6 local agencies who treat and distribute CCWD water. CCWD s treated water distribution facilities include 31 pump stations, 40 storage reservoirs, and 1,400-kilometers of pipelines. PLANNING FOR THE WORST In 1994, following the Loma Prieta and Northridge earthquakes, CCWD undertook a comprehensive seismic assessment of its water conveyance, treatment and distribution systems. The study focus was to identify strategic improvements throughout the CCWD system to minimize water service interruption after a maximum credible earthquake from relevant faults. The study identified over $120 million (1994 U.S. dollars) of improvements required in the system, including improvements to existing pumping and piping, as well as construction of additional pumping, piping and raw water system improvements. CCWD s resulting Seismic Reliability Improvement Program (SRIP) was created to implement the major improvements in a timely manner. The purpose of the SRIP was to identify a combination of capital and operational improvements that allows CCWD to provide reliable water service to present and future water customers. Starting in March 1994, CCWD began assessment of its overall water storage, treatment and distribution system. The objectives of the assessment were to maintain public health and safety, meet water needs of existing and future customers, and ensure operation of critical facilities following an emergency or natural disaster such as an earthquake. The resulting SRIP planned a cost-effective package of system improvements to better CCWD s system for future growth demands and post-earthquake fire and domestic service needs. The SRIP initially defined reliability criteria and seismic design criteria consistent with the program objectives. This first step included defining the study area (CCWD existing and planned raw and treated water service boundaries, see Figure 2); the planning period (out to the year 2020); the specific facilities to be studied, and the analysis stress events by which to model (Concord fault, M6.5; Great Valley fault M7.0). The criteria was intended to be comprehensive of both expected deficiencies of the normal operating system, and emergency based deficiencies resulting from a stress event. A key aim of this critical step was ensuring that CCWD created a study foundation that provided no more, nor less, benefit and cost than the desired outcome. Since 1994 the District has effectively implemented the complete SRIP with the completion of a backbone pipeline known as the Multi-Purpose Pipeline (MPP) this year. CCWD believes that this completed system will fully meet the above defined design criteria. Nevertheless, CCWD realized that even with completion of this monumental seismic improvement program at great cost to the District s rate payers, the investment itself would not provide the desired reliability without a trained and well prepared organization to run it under emergency conditions. Therefore, starting in 2001 at the completion of over $120 million of improvements, CCWD undertook an increased effort to train its staff for worst-case scenarios response (with an added focus to human created disasters following the U.S. September 11, 2001 New York Trade Center terrorist attacks). While CCWD invested wisely towards improved water delivery systems under seismic loading, by the beginning of 2002 it had not yet constructed the necessary infrastructure to house and assist emergency response to such events. Though the District was and is in full compliance with State of California Office of Emergency Response requirements for emergency response standards, it had not yet increased the level of expected performance from its administrative 101

120 facilities, nor its people, as it had with its water delivery systems. The result was the realization and commitment to expand its earthquake response capabilities. This commitment involved both an investment in the capital facilities necessary to organize and administer an earthquake response effort, as well as a dramatic increase in training and scenario drills to increase staff skills (both general management and field level staff) under stressful events. BUILDING THE HOME BASE In 2001 CCWD dedicated a significant share of its annual training budget towards emergency response training (approximately 25-percent of the budget). This investment was an obvious first step to ensure the District would make use of its hardened water facilities. But the District s Emergency Operation Center was an existing Maintenance staff lunch room. And, although the District s main Administration Center was designed for a greater level earthquake than this Maintenance lunch room, the building had no available space to house the response team. CCWD needed to work quickly to ensure that it would be ready to respond, including safe and operable response facilities, before the emergency forced the issue. CCWD took a two-step approach to gain this objective. As the District was in the process of designing for 2002/2003 construction a 2,140 square meter administration building for its Planning, Engineering and Construction staff (Figure 3), the first step was bringing this new building to Emergency Operation Center (EOC) seismic standards. This new facility was on a faster pace than any future dedicated EOC, and the need for the facility was more immediate (the District s lease on its current office space for over 100 employees was ending). The additional cost to design the entire structure to exceed normal building code requirements (an additional 4-percent), and expand its full capabilities to include immediate occupancy after the design earthquake, drove the decision to design the facility for a 2,500-year return cycle seismic loading level. Additionally, the District constructed one of the new building s conference rooms to include all the necessary provisions (communications, emergency power, EOC equipment and space) to meet State of California standards for an EOC. The facility was completed and occupied in early But this is not the end of the improvement story. Because this facility is normally fully occupied and in use, and the EOC constructed is in effect valuable meeting space, this EOC is only a backup EOC. Figure 3 New Engineering Administration Building (Backup EOC) 102

121 The second step to improving the level of emergency response reliability was the construction of a dedicated EOC (Figure 4). Instead of creating another situation as CCWD had with the previous lunchroom EOC, subject to the disturbance and wear and tear of daily use, the District opted to convert the existing Maintenance building, previously used as the makeshift EOC, into a fully functioning, state-of-the-art EOC. Again, the District has designed this facility with a 2,500-year design earthquake return cycle (2-percent chance of exceedence in 500 years). The District opted using the Federal Emergency Management Agency (FEMA) 356 standard, and designed the facility for a designated immediate occupancy BSE-1 level seismic event, and a life safety BSE-2 level seismic event. This required significant structural improvements to the existing building such as shearwall construction, roof diaphragm strengthening, and various building framing connection improvements. Figure 4 Emergency Operations Center The 160 square meter facility is designed with all State EOC required communication equipment, backup power, satellite television, recording equipment, furnishings, white boards, safety and first aid equipment, computer, printing and mapping capabilities. Additionally, the facility includes amenities such as kitchen facilities, storage cabinets, restroom and seating areas for life activities necessary under a prolonged earthquake event response. Unlike the backup EOC, this new EOC will be fully dedicated to emergency response. The structure is designed to ensure immediate occupancy following the design event. It additionally will be equipped with pre-designated displays of the District s overall system and key facilities, and all computers will be equipped with key response software such as CCWD s system hydraulic model, mapping software, and interface with system controls (SCADA). The facility will be equipped with emergency supplies such as water, cots, blankets, cameras and film, batteries and battery powered lighting and radios. The total construction cost of this project is estimated at $450,000. Considering that this facility will now allow the District to make full use of its $120 million SRIP investment, the cost is a general bargain. The project is on schedule for construction completion by early The 103

122 District has constructed over 6,000 square meters of office space over the last four years, so the District s strong experience in office construction ensures the construction of this project on schedule and budget. With the completion of this project, the District will have support facilities at least above par with the actual water system facilities performance. And as with the funds invested in the actual system, and with the increased and on-going training the District provides its staff, this facility will significantly leverage District resources. RESPONSE ORGANIZATION CCWD s emergency response planning started in earnest in The initial planning involved quarterly meetings of key staff in the District s Maintenance Building to discuss the general approach to respond to emergency events of any kind, including earthquakes, chemical releases, flooding, and critical system failures. By 2001 this emergency planning had evolved into an organized approach that included use of the same Maintenance facility, but with a moderate amount of equipment, furnishings, and communication equipment to better respond. The key focus of the response team up until 2001 was earthquake preparedness. The greatest single emergency risk the District believed it could face was a catastrophic earthquake. However, with the September 11, 2001 terrorist attacks on the U.S., the District shifted its primary response focus from earthquakes to terrorist attacks. Although CCWD maintained the strong reality that earthquakes are a major risk to the system, until early 2003, the District shifted the focus of its emergency response planning and training from earthquakes to terrorist response. One of the key lessons in this shift however is that response to either emergency, earthquake or terrorist attack, is quite similar. As is outlined below, the emergency response organization works for either scenario. Additionally, most of the facility improvements made with the $120 million SRIP, as well as the major investments in an EOC and backup EOC, provide the same benefits regardless of the type emergency. The SRIP improvements had not only hardened District facilities against earthquakes through strengthened structures, but it had also made the facilities more durable to vandalism and terrorist attack, including strengthened access (for example through stronger roofs and overall structures.) Additionally, and probably most importantly, the SRIP had added significant redundancy to both the District s treated water and raw water systems. Again, regardless of the emergency event, this improved redundancy for earthquake has an almost direct translation into improved redundancy for terrorist attack. The result is that the $120 million SRIP investment proved to have an even greater return on investment through the flexibility, improved security and deterrence to terrorism provided while focusing on earthquake response. CCWD s Emergency Operations Plan (EOP) is based on the concept that operations during disasters should leverage off of the capabilities that exist within the core organization under normal operations. The approach utilizes the same core competencies of the departments and personnel for normal operation, but formats the organization into a team more conducive to emergency response. The District has chosen a response plan modeled on the State of California s Standardized Emergency Management System (SEMS). The value of this approach is that not only does the District build off the intelligence and lessons developed by the State, but it assures it can better interface with other organizations within its service area that complied with these same standards. The District in effect formed a cost effective emergency response network simply by complying with the State s SEMS standards. 104

123 The objective of the EOP is to provide prompt and effective District-wide response to emergencies, protect the public welfare by minimizing the effects of emergencies, coordinate response activities with interrelated public and private agencies, and expedite repair and recovery of operations to normal condition. Focus situations for this team include earthquakes, hazardous materials incidents, wildfires, flooding, contamination, terrorism, and unplanned power outages. As noted, District staff train regularly to respond to these scenarios using the SEMS organization. The training consists of response simulations, but also lecture and seminar style coursework. The training also involves the continuous development and improvement of various checklists and contact lists used by the individual functions within the team. In addition, all the training and exercises include written recording of the events to ensure lessons learned through the training are built-upon for the team s overall advancement. These lessons are then reviewed prior to the next training exercise, and are maintained as reference by the individual staff. The actual emergency response team organization consists of two branches of staff: command and functional (Figure 5). The command staff consists of response commander, safety, risk management, liaison, public information and financial support staff functions. The functional staff consists of operations, planning and intelligence, administrative logistics, and finance functions. Each function has clearly defined roles and responsibilities. The key to the success of this organization is that the assignments to these positions is based upon the individual manager and department expertise related to these functions, yet these functions are consistent with the necessary type of organization for emergency response. EMERGENCY OPERATIONS TEAM POLICY GROUP Board of Directors Represented by General Manager EMERGENCY OPERATIONS DIRECTOR Director of Operations & Maintenance SAFETY Safety Officer LIAISON Environmental Compliance Officer SUPPORT O&M Analyst PUBLIC INFORMATION Director of Public Affairs OPERATIONS SECTION CHIEF Water Operations Manager PLANNING/INTELLIGENCE SECTION CHIEF Director of Engineering LOGISTICS SECTION CHIEF AGM Administration FINANCE/ADMINISTRATION SECTION CHIEF Director of Finance Figure 5 105

124 In addition to this organization, the District has entered into various mutual aid agreements with adjacent public agencies such as Contra Costa County, Marin Municipal Water District, Alameda Flood Control, Santa Clara Valley Water District, San Francisco Water Department, Solano County, East Bay Municipal Utility District, and Alameda County Water District. These agreements provide for a variety of mutual aid including staff, materials, equipment, as well as emergency intertie connections with adjacent water systems, as applicable. Of course the aid is based on a cost reimbursable basis. In addition, the District entered into preestablished contracts with various suppliers and distributors for materials and equipment (for example pipeline material and emergency power) to speed procurement during the emergency, ensuring that needed supplies are maintained for the District for use in the event of an emergency. The major benefit of CCWD s emergency response approach is that through the continuous training, exercise, and planning, CCWD has identified significant areas of improvement during the relative calm of normal operations to ensure more effective response under the pressure of an emergency. Through the EOP, the District identified the need of better facilities for an EOC, and the need to have a well-equipped and ready backup should the primary EOC be debilitated or overwhelmed. The result was the upgrade of an already planned administration facility for use as a backup EOC (including upgrading the seismic loading criteria), and the conversion of an existing non-performing building into a dedicated EOC (designed to the latest criteria for an emergency response facility). The overall outcome of this work is a better prepared District organization, not only for the normal challenges of daily operation, but for the unforeseen disasters known to be out in the future. CONCLUSION CCWD s primary objective is ensuring the highest level of water service economically feasible under all conditions. Through planning and training the District has developed an approach that not only has improved the overall infrastructure of the District to better withstand earthquakes and other disasters, but has improved its staff and organization to provide a significant increase in earthquake readiness. From 1994 to 2002 CCWD implemented a $120 million capital improvement program to improve its various raw and treated water systems to ensure its core water delivery functions could continue following design level seismic events. Simultaneously the District invested in training and advancement of its human resources to ensure that not only did its physical assets perform under the stress of a major earthquake, but so would its human capital. The resulting emergency response team and Emergency Operation Plan have provided the vehicle for that objective. The program increased staff capabilities, and identified staff support facility needs. The District has aggressively moved to meet all these needs through increased expenditures in training, the construction of new EOC facilities (both primary and backup), and purchase of necessary support equipment and supplies. The result is a better prepared District organization ready to respond with effectiveness to the worst foreseeable events. 106

125 Emergency Water Supply Facilities of Hachinohe Regional Water Supply Authority Kenetsu Kojima ABSTRCT Hachinohe Regional Water Supply Authority (HRWSA) has been actively developing a quake-resistant water supply system. The water supply facilities including distribution pipelines have been developed with due considerations on quake-resistance. Especially for water distribution pipes, HRWSA firstly in Japan adopted quake-resistant ductile iron pipes, S-type pipe. Quake-resistant pipelines have been installed with a total length of 351 km so far with a diameter ranging from 75 to 1,500 mm. In addition to development of quake-resistant pipelines, HRWSA has provided water storage tanks and water supply valves, which are built next to or into the distribution pipelines, to function as water supply stations in emergency. This paper presents the basic concept of the above-mentioned facilities and emergency water supply system including water storage tanks and water supply valves. 1. Current Status Since 1968, the Hachinohe Region has been hit by 11 big earthquakes, which have seismic intensity of larger than 4 of the Japanese scale. Hence, the pipes and other water supply facilities have been constructed with due considerations on quake-resistance. 2. Problems Since development of quake-resistant pipelines requires a huge amount of investment cost as well as long construction period, HRWSA has installed quake-resistant pipes for only some 20% out of total length. In this connection, provision of emergency water stations is urgently necessary. 3. Solution In addition to development of earth-resistant pipelines, HRWSA has been developing the emergency water supply system by provision of 1 water storage tanks at evacuation sites and 2 emergency water supply valves at quake-resistant distribution pipelines. 4. Conclusion HRWSA has been steadily implementing the measures against earthquake. In 1996, it decided to adopt quake-resistant pipelines for both of new installation and replacement of the transmission and distribution pipelines. It will make sure to establish the emergency water supply system to secure direct water supply to evacuation sites by fully utilizing the facilities mentioned above. Kenetsu Kojima, Division Chief, Water Services Division, Hachinohe Regional Water Supply Authority, Minamihakusandai Hachione City, Aomori, Japan

126 1. INTRODUCTION The Hachinohe Regional Water Supply Authority (HRWSA) is a middle size waterworks system located in the southeast of northern tip of Honshu Island. Current service population is 330,000 and the service area stands at 790 km 2 with maximum water supply of 114,000 m 3 /day in At the 2 nd Water Supply Workshop held in 2001, HRWSA reported about the development of quake-resistant water supply system titled The Measures against Earthquake based on Earthquake Simulation. This time, HRWSA presents the outlines of its emergency water supply system including water storage tanks and water supply stations. Photo.1. Main Office of Hachinohe R.W.S.A 2. CURRENT STATUS Hachinohe City Water Supply Bureau, the predecessor of HRWSA, developed quake-resistant ductile iron pipes, S-type pipe, in collaboration with a cast-iron pipe manufacturer. It stemmed from the experiences and lessons learned at Tokachi-oki Earthquake which occurred in May 1968, i.e.; water wagons failed to comply with the needs of people. Since then, HRWSA tried to develop S-type pipe, which aims to resist earthquake not by rigid strong pipe materials but by flexible structure of pipelines accommodating the movement of the ground. These distribution pipelines were required not to be pulled out nor to be broken at joints by external force when an earthquake occurred. Finally, a chain-like structure joint, S-type, was invented. It was designed to have characteristics of flexibility and ultimately not to be pulled out by a stopper. 108

127 Fig.1. S-type Mechanical Joint The said distribution trunk line rings round Hachinohe City, and makes the followings easy: to establish emergency water supply stations equipped with emergency water supply valves; to use the pipelines as an emergency big water storage tank with a total volume of 11,000m 3 ; and to supply water for wide area including newly developed areas. Fig.2. Looped Main Pipe In 1985, HRWSA became a regional water supply system by integrating the waterworks of one city, eight towns and two villages and has been developing a quake-resistant water supply system to secure lifelines by applying S-type pipes to the distribution trunk delivering water in 109

128 its respective service areas. It also has abandoned some of its small aged facilities and now developed into an efficient water supply system. Furthermore, in designing service reservoirs and pumping stations, strict horizontal seismic coefficients were applied as shown in Table.1. TABLE.1. DESIGN SEISMIC COEFFICIENT IN EXISTING FACILITIES Design year Facility Applied horizontal seismic coefficient 1988 NISHIYAMA Service Reservoir HIBARINO Standpipe MABUCHI Standpipe NANGO Standpipe HAKUSAN Service Reservoir (expansion) HAKUSAN standpipe MUKAIYAMA Standpipe (expansion) NANKODAI Service Reservoir (expansion) HAKUSANN Water Reservoir (expansion) 0.3 HRWSA has provided emergency water storage tanks in the service areas where quake-resistant pipes have not been installed. While, for evacuation sites where quake-resistant pipes have been installed nearby, emergency water supply valves have been provided to secure water supply in times of emergencies. These measures were taken from the lessons learned in the two large earthquakes of Sanriku-Harukaoki Earthquake in December 1994 and Great Hanshin Earthquake in January Water supply status during the emergency operation in Sanriku-Harukaoki Earthquake was as follows: (cumulative) period of water suspension: 137 hours households affected by water suspension: maximum 30,000 households water supply by water wagons operation period: 7 days emergency water supply stations: 138 stations water tank wagons (as 1m 3 /unit): 347 units( 50 units per day) water supply amount: 1,443m 3 ( 206 m 3 per day) mobilized staff: 527 persons ( 75 persons per day) 3. DEVELOPMENT PLAN OF EMERGENCY WATER SUPPLY FACILITIES AND CURRENT STATUS To secure emergency water supply stations is an urgent requirement, because HRWSA has installed quake-resistant distribution pipes for only some 20% of its approximate 1,800km long pipelines. The policy for developing emergency water supply facilities is as follows: 1to provide emergency water storage tanks at evacuation sites; HRWSA provides water storage tanks in order to keep drinking water for areas where 110

129 quake-resistant pipes are not installed. 2 to establish water supply stations; HRWSA establishes water supply stations mainly for water wagons by developing water reservoirs and emergency water supply valves. 1) Emergency Water Storage Tanks Sites selection and capacity decision for emergency water storage tanks are as follows ; a) Sites selection Presently, out of 221 open-air evacuation sites within the service area, 58 sites (or 26%) are more than 2 km away from the service reservoirs or distribution pipes equipped with emergency valves. Among them, priority was given to the sites where the past earthquakes seriously damaged. Especially, towns and villages along the Pacific Ocean were considered as priority sites taking account of anticipated damages simulated with the expected seismic intensity, magnitude 8.2. The outline of the simulation was already reported in the 2001 Workshop. b) Capacity Decision - Emergency water requirement was assumed to be 3L/day per capita for the first 3days, the first stage of emergency water supply operation. This volume was referred to The Guidelines for quake-resistant water supply system planning prepared by the Ministry of Welfare, predecessor of Ministry of Welfare and Labor, as shown in Table2. - The capacity of a water storage tank shall be less than 1/3 of water consumption in the downstream area of the storage tank to prevent water quality from deteriorating in the tanks at normal time. - Water storage tanks shall be located within approximately 2km of radius, taking account of accessibility of people on foot in case of emergency. TABLE.2. REQUIRED WATER SUPPLY FOR EMERGENCY Days after Required per capita Distance to water Proposed method of water earthquake consumption station supply 0~3days 3L per capita per day approximately within 1km Quake-resistant water storage tanks and water tankers ~10days 20L per capita per day approximately Temporary water supply within 250m faucets near main distribution ~21days 100L per capita per day approximately within 100m ~28days 250L per capita per day (the same volume before disaster) approximately within 100m pipes Temporary water supply faucets on the branch distribution pipes Water supply faucets from temporary distribution pipes 111

130 Proposed Site Item Emergency Water supply per capita per day (the first stage) Population in a 2km radius of storage tank Required capacity (3L x 3days x population) Average supply volume at downstream of the tank TABLE.3. STORAGE TANK CAPACITY TOYOSAKI ISHIBACHI KOUYOU Elementary Elementary Elementary School, School, School, HACHINOHE KAIJOU MOMOSEKI City Town Town KINOSHITA Elementary School, SHIMODA Town 3L 3L 3L 3L 3L 3,500 4,000 4,800 4,900 5, m 3 ( 40m 3 ) 36m 3 ( 40m 3 ) 43.2m 3 ( 50m 3 ) 44.1m 3 ( 50m 3 ) Central Community Center, NAGAWA Town 46.8m 3 ( 50m 3 ) 150.5m 3 /day 120.7m 3 /day 180.2m 3 /day 180.3m 3 /day 408.0m 3 /day Construction of the emergency water storage tanks has completed at five sites under this plan. Fig.3. Water Storage Tank for Emergency 112

131 2) Emergency Water Supply Valves Emergency water supply valves are installed in quake-resistant distribution pipes aiming at efficient operation of water wagons and securing water supply to nearby residents in emergency. The valves have been already provided at 31 sites, because its installation was easy by drilling small holes in the pipes and the measure was economical. Connection pipelines for four water supply stations located at elementary schools and junior high schools as evacuation sites were completed before the year This project is one of the quake-resistant measures as reported at the 2001 Workshop. Fig.4. Water Storage Tank And Water Supply Valve for Emergency 4. FUTURE DEVELOPMENT PLAN The Disaster Relief Law stipulates that a local authority such as a prefecture or a municipality shall be primarily responsible for water supply in emergency. Although it is not the government, HRWSA provided five emergency water storage tanks as quake-resistant measures. However, HRWSA concluded that emergency water supply valve is better than small emergency water storage tank with a capacity of 50m 3 or similar small capacity for the reason of cost-benefit performance and it is not going to build any other emergency water storage tanks without support from a local government. In 1996, HRWSA decided to adopt 113

132 quake-resistant pipelines for both of new installation and replacement of pipelines. Following this policy, HRWSA will make sure to establish emergency water supply system to secure direct water supply to evacuation sites by fully utilizing such quake-resistant facilities. Fig.5. Water Supply System for Emergency 5. CONCLUSION The main point for emergency water supply system has been how to secure drinking water when water supply is cut off right after an earthquake occurs. Bottled water may be more economical just for drinking purpose, because in Japan affected area is considered to be limited and access roads are expected to be passable. Past experiences, however, imply that people needed a considerable volume of domestic water for toilet flush, kitchen, laundry, etc. rather than drinking water. It is also important to secure water for fire fighting. Taking the facts mentioned above into consideration, further we need to pursue a water supply system which is free from water suspensions or which can be sustained with fewer interventions. REFERENCES (1) Japan Water Works Association, 1997, Design Criteria and Explain For Water Works Facilities Quake-Resistant Construction Method (2) Japan Water Works Association, 1990, Design Criteria For Waterworks Facilities (3) Japan Water Works Association, 1999, Criteria For Water Works Facilities Quake-resistant Plan (Draft) (4) Hachinohe Regional Water Supply Authority, 1998, Fresh 21 st Century Plan 114

133 3 rd US-Japan Workshop on Water System Seismic Practices TECNICAL SESSION III Seismic Risk Management and Post Earthquake Recovery Seismic Damage Simulation of Distribution Pipeline Based on the Monitoring Data Collected by a Seismometer Network Presenter: Koichi Murata (Osaka Municipal Bureau, Japan) Prioritization of the City San Diego Water Department s Capital Improvement Program Presenter: Michael E. Conner (City of San Diego Water Department) Seismic Practices Evaluation of Kobe Water System using Risk Management Approach Presenter: Makoto Matsushita (Kobe City Waterworks Bureau, Japan) Knowledge Management in Engineering - a Business Improvement Methodology Applied to Seismic Risk Management Presenter: Jim Woodhams (Thames Water Utilities, UK) Emergency Restoration for Water Supply Following Earthquake Disasters in the City of Yokohama Presenter: Kiyoshi Naito (Yokohama Waterworks Bureau, Japan) 115

134 3 rd US-Japan Workshop on Water System Seismic Practices 116

135 Seismic Damage Simulation of Distribution Pipeline Based on the Monitoring Data Collected by a Seismometer Network Kazuya Yamano, Katsuhiko Eguchi, and Koichi Murata ABSTRACT In 3 fiscal years starting from 2000, Osaka Municipal Waterworks Bureau developed the Disaster Information Management System, for the purpose of assisting the quick and precise decision regarding restoration activities, which carry out management concentration and common use of disaster related information. After the occurrence of direct seismic damages to water supply facilities, early restoration is necessary to decrease the influence of water suspension on the civic life, city activities, economic activities, and so on. However, when pipeline is damaged widely, it is difficult to carry out the restoration activities efficiently during a few days after the damages occurrence, because it takes time to pinpoint stricken areas and grasp a hazard scale. From the above aspect, the Disaster Information Management System includes a function for seismic damage simulation of distribution pipeline based on the monitoring data collected by a seismometer network. This function contributes to carry out the restoration activities more efficient through the quick planning, with the simulation result, of the formation of restoration teams and the arrangement of restoration materials, even in the period when we have the lack of information about the pipeline damages. This paper presents the outline of the configuration of equipment, operation, and expected effects of this function. Kazuya Yamano, Manager for Earthquake Preparedness, Engineering Div., Osaka Municipal Waterworks Bureau, Nanko-kita, Suminoe-ku, Osaka JAPAN Katsuhiko Eguchi, Chief of Planning Section, Planning Dept., Engineering Div., Osaka Municipal Waterworks Bureau, Nanko-kita, Suminoe-ku, Osaka JAPAN Koichi Murata, Chief of Water Distribution Section, Water Distribution Dept., Engineering Div., Osaka Municipal Waterworks Bureau, Nanko-kita, Suminoe-ku, Osaka JAPAN 117

136 1. INTRODUCTION Since its foundation in November 1895, Osaka Municipal Waterworks Bureau has been expanded via 9 major improvement and expansion projects in response to the increased demand for water supply in the growing city of Osaka. Especially, in order to cope with the ever-growing demand for water supply in the 1950s, two new water purification plants were constructed in 1957 and Currently, Osaka City has the water supply capacity to the level of 2.43 million m 3 /day, supports the civic life and highly advanced industrial activities in Osaka City. The water supply network has been, and is, subject to continuous upgrading to ensure the reliability and stability of water supply to the people of Osaka. Improving Earthquake Resistance of Key Facilities Establishing a Water Supply and Distribution Center Network Improving Earthquake Resistance of Headquarters Necessary for Disaster Relief and Recovery Activities Improving the Information and Communication System s Reliability 8 Basic Elements Im proving M utual Com patibility among Different Distribution Systems Countermeasure against Power Failure Establishing Stable Water Supplying Routes to Manmade Island Expanding Emergency Material Stocking System Figure1. Osaka Waterworks Bureau Earthquake Preparedness Plan Kobe earthquake occurred in 1995, damaging the waterworks facilities of Kobe and other devastated areas. Nobody had anticipated such a massive earthquake, which measured 7 on the Japanese seismic intensity, in the area. The earthquake in Kobe prompted not only the Japanese Government, but also local municipalities to take action to protect water facilities against massive earthquakes. The process by which the damaged water supply lines were restored and resumed their supply of water to the earthquake-damaged area helped us to realize that all the information sporadically coming in to the emergency headquarters piece by piece, including which part of water supply lines are damaged by the earthquake, should be arranged in an easy-to-understand manner so that appropriate decisions can be made in a timely manner. In order to prepare for possible earthquakes, Osaka Municipal Waterworks Bureau devised Osaka Municipal Waterworks Earthquake Preparedness Improvement Plan 21 (Seismic Water Osaka Plan 21) in March 1996, as a series of directive measures against earthquake disaster (Figure.1). In line with the Plan, the Bureau established the 'Disaster Information Management System of Waterworks Bureau' from 2001 to 2002, which effectively shows the conditions and degree of restoration of water facilities and emergency water supply information on a map and/or in a form of tables, aiming at supporting decision making by the disaster emergency headquarter in a timely manner. 118

137 Seismic damage simulation of distribution pipeline based on the monitoring data collected by a seismometer network, as shown in the title of this paper, is an integral part of the Disaster Information Management System of Waterworks Bureau. Using this system, even immediately after an earthquake detailed damage information is not available for a certain period of time, how the city's water supply lines are damaged due to the earthquake can be estimated, based on the estimated ground surface velocity of the city area hit by the earthquake. This function enables a decision to be made on the order of investigating the damage to water supply lines by location and when and how a water supply line restoration party should be formed to start restoration work, whereby early restoration of the damaged water infrastructure can be expected. Accordingly, this paper introduces Disaster Information Management System of Osaka Waterworks Bureau and how the system is used to estimate the damage to water supply lines using seismic monitoring data, as well as the manner in which the result of the simulation is utilized in actual emergency restoration activities. 2. OUTLINES OF THE DISASTER INFORMATION MANAGEMENT SYSTEM OF OSAKA WATERWORKS BUREAU 2.1 System Functions Function of Seismic Damage Simulation When an earthquake hits, the distribution of ground surface velocity of the entire city and location of damage to the water supply lines around the city are estimated and plotted on a map of the whole of Osaka City, based on seismic scale data collected by seismometers scattered around the city, ground surface velocity information monitored by the City's seismic monitoring system, characteristics information of the ground in Osaka City and other information released by the Meteorological Agency of Japan such as epicenter information, the method of simulation of which is shown in the next chapter. The system will play an important role in estimating where water supply lines are damaged even immediately after an earthquake when no such information is available. This is one of the most noteworthy features of the Disaster Information Management System of Osaka Waterworks Bureau. Function of Emergency Restoration and Water Supply Activity Management The location of low water or water suspension, conditions of having water supply equipments in place, damage to water intake/water purification/water distribution plants and water supply lines which are collected during the initial phase activities and restoration activities, as well as estimated dates of restoration are displayed in the total area or detailed area map of Osaka City. This function is highly useful in conducting emergency water supply activities and restoration of water supply lines and water intake/water purification/water distribution plants in an efficient manner. Pictures taken by a digital camera at the damaged site can be transmitted via digital communication equipment to each information terminal on the System. 1. Emergency Water Supply Information Management Date and conditions of having water supply equipments (temporary water tanks, water supply tanks), in such important places as evacuation sites and hospitals, and volumes of water refilled and supplied at each location are registered and managed in the System, which is useful in effectively grasping the status of water supply at each focal point where water supply is temporarily or totally terminated. At the same time, this function is used to decide the priority of the water suspension areas 119

138 in installing water supply stations, and facilities requiring emergency measures, in a timely and orderly manner. 2. Facility Restoration Information Management Facility restoration status at water intake/water purification/water distribution plants is shown on a plan and on a water purification process flow chart. In addition, the volume of water reservoir at each reservoir as well as emergency water inspection results can be monitored on the System. Functional assessment data taken at each water purification center are inputted, calculated and shown in the System. This function is also used to select which waterworks facilities need to be restored most urgently. 3. Restoration Status Information Management for Distribution Pipeline Details of water pressure measured by telemeters scattered around Osaka City are collected and tabulated to estimate the location of low water or water suspension just after an earthquake when there is no reliable information available. By doing so, which damaged distribution pipelines require investigation most urgently can be decided. The status of damaged water supply lines, anticipated restoration periods, as well as low water and water suspension information can be showed on the map. Personnel Management Function This function is used to control the Osaka Bureau's staff emergency activities in order to assign the right person to the right position depending upon the degree of damage caused by an earthquake as well as the number of staff available for emergency restoration and water supply activities. 2.2 Disaster Information Communication Terminals, etc. When a disaster occurs, an emergency headquarters is set up at Waterworks Bureau s main office, which is assigned to perform as a center for controlling emergency activities. As soon as the headquarters are established, water purification plants, maintenance offices, and service offices, acting as satellite hubs, will be assigned to various jobs including emergency water supply, restoration of damaged waterworks facilities (water intake/water purification/water distribution) and pipelines. In order to facilitate smooth communication between the emergency headquarters and each hub, as well as to share necessary information regarding the disaster among them, an information system terminal is installed at Waterworks Bureau s main office, water purification centers, maintenance offices and service offices to connect them. The Disaster Management Office is located at the main office where various information communication equipment such as radio wireless, telephone and fax machines are arranged in order to collect diversified information including earthquake, disaster, river condition and updated reports from Osaka City Disaster Countermeasures Headquarters and news agencies and so on, all information of which are to be controlled in an integrated manner. 2.3 Others Water distribution and water quality information in Osaka City is very important as basic reference to make plans as to which part of the city should have their water supply partially or totally stopped due to planned water supply line works, as well as to make emergency plans against unexpected damage to water supply lines and sudden failure of water distribution pumps. As such, 120

139 Osaka City has so far collected and monitored water distribution information from telemeters installed all over the city (Water Distribution Information Management System). The above information is also important as basic data for the emergency headquarters and hubs around it, when undertaking emergency measures in times of disaster. Currently, Osaka City is trying to interface the newly installed disaster information system with its existing water distribution information so that water distribution and water quality information can be referenced in a centralized manner. 3. FUNCTION OF SEISMIC DAMAGE SIMULATION The Disaster Information Management System of Osaka Municipal Waterworks Bureau analyzes earthquake motion according to Japanese seismic intensity measured by seismometers installed all around the City, together with epicenter information released by the Japanese Meteorological Agency, to estimate the incidence of pipe damage in Osaka City - information which is then used to support smooth initial emergency restoration activities such establishing which part of a devastated area, most urgently, needs investigation of pipe damage and/or temporary damage recovery activities. This is especially important just after an earthquake has hit, when no relevant or reliable information is available. In this chapter, the outline of the function to estimate pipe damage from an earthquake is described and a model to estimate the location of damaged water supply lines is examined. 3.1 System Configuration Digital Mapping and Geographic Information Systems Input Data Output Data Seismometer Seismometer Distribution Information Management System INS lines Seismic Monitoring System Seismic Observation Data Media (Offline) Distribution Pipeline Data Seismic Damage Simulation Function Water Distribution Conditions Modem Epicenter Data Japan Meteorological Business Support Center Seismic Damage Simulation Server Damage Simulation Data Map Ground Distribution Pipelines Past Earthquake Series of Data Data Necessary for Damage Simulation Disaster Information Server Through the Web Simulation Result PC Terminal Disaster Information Management System Figure 3-1. Earthquake Damage Simulation System Configuration and the Flow of Data 121

140 Figure 3-1 shows the system configuration of the seismic damage simulation function and the flow of data. When an earthquake hits, it is first necessary to analyze the seismic vibration using epicenter data released by the Japanese Meteorological Agency, which is delivered to us by the Japan Meteorological Business Support Center, and seismic monitoring data using Osaka Municipal Waterworks Bureau's Seismic Monitoring System. The result of the above analysis, together with information on the water supply line status in Osaka City, which is collected via the Distribution Information Management System, is used to estimate the location and degree of damage to the City's water supply lines. This information is then transmitted to the server of the Disaster Information Management System of Osaka Municipal Waterworks Bureau. 3.2 Earthquake Damage Simulation Method Earthquake Motion Analysis For the purpose of this analysis, there are two possible methods: one is to determine active faults subject to earthquakes and examine the process of their destruction in detail, while the other is to analyze maximum velocity and maximum acceleration from the latitude, longitude, and depth of the epicenter and magnitude of earthquakes. As noted previously, the aim of estimating earthquake damage is to support initial phase activities smoothly just after the occurrence of an earthquake, when no relevant or reliable information is available. As such, Osaka City adopted the latter method, taking into consideration various factors such as the applicability of the data of each earthquake, and the time required to analyze the data. In order to curtail the time required for analysis, all the seismic information for which the epicenter is anticipated has been stored in the database. Consequently, anticipation of earthquake damage can be completed in around a minute. The data stored in the database has been further augmented by incorporating actual measurement data from the Earthquake Monitoring System to increase its accuracy. 1. Identification of the Epicenter The epicenter is a point where destruction of an active fault starts, and is designated as the origin of the earthquake. However, no earthquake occurs with a point as its origin, but originates from a plane of a fault. Therefore, wherever available, active fault data (plane data) is estimated from the epicenter data by using active fault information stored in the database. 2. Simulation of Earthquake Motion by Distance-Decay Formula Next, based on the epicenter data (point data) or the fault data obtained via the epicenter identification process, the shortest distance between the epicenter or the fault and each one of the Osaka City areas divided in approx sub-areas is calculated. Then, the earthquake motion effect of base layer is analyzed using the Annaka's Distance-Decay Formula. Annaka's formula [1]: log m M + Ch H Cd log10{ R + C1 exp( C2M } + C0 C m =0.725,C h = ,C d =1.918,C 0 =-0.519,C 1 =0.334,C 2 = Vel = C ) 122

141 Where M stands for the magnitude of the earthquake, R for the shortest distance between the fault and a sub-area in the City (km), H for the depth of the epicenter or the center of the fault, and Vel for seismic velocity at the base layer. Earthquake motion at the ground surface is analyzed using an amplification factor corresponding to the ground data after a response spectrum at the base layer. 3. Method for Increasing Accuracy of Earthquake Motion by Earthquake Monitoring Data Water supply facilities are expected to maintain their function after the occurrence of an earthquake. In order to do so, it is necessary to grasp the incidence of damage caused to each facility as soon as possible and efficiently operate water supply facilities, taking into consideration the seriousness of damage to the facilities. For this purpose, a seismometer is installed at 11 water intake station/water purification plants/water distribution plants located in and out of Osaka City to collect and accumulate seismic data. In this function of Seismic Damage Simulation, the relationship between observation value via the above Seismic Monitoring System and the result of the above analysis is determined using the least squares method. By compensating for the difference, accuracy in earthquake method analysis data is enhanced. Simulation of Pipe Damage As the pipe damage from an earthquake is more dependent on the velocity amplitude of seismic wave than its acceleration amplitude, according to the results of studies made so far [2], the overall incidence of pipes damaged from an earthquake is calculated by multiplying the total of pipe length by the damage ratio (places/km), which is calculated from the estimated ground surface velocity of seismic wave obtained from the earthquake motion analysis. Average damage ratio for individual materials and joints of pipe and correction factor decided by the diameter of a pipe and ground type were established at The Osaka Municipal Waterworks Bureau Examination Committee of Anti-Seismic Measures for pipes, including academic experts. 1. Average Damage Ratio(D 0 ) TABLE 3-1. Relationship between Maximum Seismic Vibration Velocity and Average Damage Ratio Equations Types of pipe(materials and joints) D 0 =0 Ductile iron pipe with S, SⅡ, PⅡ and KF type joints (Anti-seismic joints) and steel pipes with the diameter of φ 800mm or above installed in or after D 0 =0.004(V max -20) Ductile iron pipe other than above Cast iron pipe regeneration by applying coating agents inside the pipe D 0 =0.010(V max -20) Cast iron pipe with mechanical type joint D 0 =0.016 V max Cast iron other than above and vinyl pipe The following relationships as shown in TABLE 3-1 below have been determined for individual materials and joints of pipe with the simulation that maximum velocityv max of seismic wave on the surface and average damage ratio D 0 have a linear relationship. In this determination, an investigation of data of pipe damage caused by The Kobe Earthquake[3], made up by Japan Water Works 123

142 Association, was used as reference data showing the relationship between seismic wave and pipe damage. In making up the table shown below, the data of the places where liquefaction occurred was omitted. 2. Correction by Pipe Diameter and Ground Type The incidence of pipe damage by an earthquake depends upon the pipe diameter and geography of the area. In order to reflect the variation of damage ratio by pipe diameter, etc., a correction factor against the above average damage ratio was determined taking actual measurement data from the Kobe Earthquake as a reference. TABLE 3-2 shows the correction factor CR 1 against pipe type and diameter, and TABLE 3-3 for the correction factor CR 2 against the geography. TABLE 3-2. The ratio of the damage ratio for each type of pipe and each pipe diameter Pipe φ100 φ200 φ300 φ500 ~φ75 Type ~150 ~250 ~450 ~ DIP SP CIP Ratio against Average Damage Rate: CR 1 Ratio against Average Damage Rate:CR TABLE 3-3. The ratio of the damage ratio at Reclaimed Land Pipe Reclaimed Type Land DIP SP CIP Ratio against Average Damage Rate: CR 2 Ratio against Average Damage Rate:CR 2 3. Simulation of Damage to Pipes From the above, Damage Ratio (D) (places/km) was determined with the calculation formula of D=D 0 CR 1 CR 2. For every 1900 sub-areas in Osaka City, the incidence of pipe damage from an earthquake and its local distribution were estimated by multiplying the average damage ratio by the total of pipe length for individual materials, joints and diameter of pipe. 3.3 Examination of the Earthquake Damage Simulation Model Preconditions The Seismic Damage Simulation Model is examined using the epicenter information and data on pipe damage in Osaka City caused by the Kobe Earthquake in The epicenter information and number of pipe damage are shown in TABLE 3-4 below. The distribution of pipe damage is shown in Figure

143 Epicenter Depth Magnitude 7.2 Number of damaged water supply lines in Osaka City TABLE 3-4. Epicenter Information Used for the Analysis (From the data of the Kobe Earthquake) Lat ' Lon 'E Measurement data by Japan Meteorological Approx.18km Agency 264 places Number of Places which repaired water leakage within two weeks after the Earthquake Actual Damages Observed Figure 3-2. Place of Actual Pipe Damage (by Kobe Earthquake) Examination of Earthquake Damage Simulation 1 (Number of Pipe Damage) The simulation of earthquake damage caused to Osaka City using the above data revealed that 215 places may incur damage from an earthquake in Osaka (excepting the places where the simulation result comes to less than 0.5 places). In this simulation, the pipe data used was as of the end of March 2002, the latest information according to the database obtained by Osaka City, and not that of 1995 when the Kobe Earthquake occurred. When taking into consideration the fact that due to the upgrading works of cast iron pipes 125

144 after the Kobe Earthquake, the total length of cast iron pipes has decreased to approx. 22% of the total length of all pipes, it is estimated that approx. 205 places may be subject to damage in Osaka, and the difference is converged within 5%. Thus, though direct comparison study of pipe damage is difficult, it may be said that it reflects the probable number of pipe damage. Examination of Earthquake Damage Simulation 2 (Distribution of Pipe Damage) Figure 3-3 shows the correlations between the number of places where pipe damage were observed in Osaka City due to the Kobe Earthquake and the analyzed value per administrative unit. The number of places where pipes were damaged are adjusted taking the replacement of cast iron pipes from 1995 to As shown below the number of pipe damage in Osaka City and analysis data has strong correlations, which means the analysis data represents almost identical distributions with the actual number of pipe damage by the Earthquake. N um ber of dam ages observed in O saka (places) R 2 = Analyzed value (places) Figure3-3. Comparison of Actual Pipe Damage to Analysis Data (per Administrative Unit) Summary From the above examination, Osaka Municipal Waterworks Bureau's Model of Seismic Damage Simulation can be utilized effectively for emergency water distribution and temporary damage recovery activities just after the occurrence of an earthquake, when relevant information is not available. 126

145 4. OPERATION OF EARTHQUAKE DAMAGE SIMULATION FUNCTION City D isaster Prevention Inform ation System W ater D istribution Inform ation M anagem ent System Disaster Information Management System of Waterworks Bureau Seism ic Dam age Sim ulation Function Disaster Anticipation of Fire H azards Distribution Plant O peration C onditions Result of W ater Supply Line D am age Simulation Low Pressure and W ater Suspention Area Efficiency Improved by the Earthquake Damage Simulation Function Information Automatically Shown in GIS Obtained by the System Distribution plant: Normal Water Pressure in City Area:Low Pipeline Damage Ratio:High Yes Through Investigation Through Investigation of M ajor W ater Supply of M ajor W ater Supply Lines Lines Identification of Identification of Damage Area Damage Area W ater Distribution Contro l W ater Distribution Contro l Priority:Fire Extinguishing W ater Priority:Fire Extinguishing W ater Estim ated Recovery Period W ater Distribution Contro l Valve H andling for Damage Em ergency W ater Damage Investigation Suspention and W ater Investigation PersonnlA location Distribution C ontro l PersonnlA location Information Blank Period Confused Information Period Judged by Maintenance Offices Decided by Disaster Countermeasure Headquarters Figure 4.Flow Chart of Emergency Water Distribution and Pipe Recovery Activities Just After an Earthquake The foundation of earthquake preparedness program for waterworks is to secure a stable supply of water that is necessary to sustain citizen s lives. Therefore, in order to ensure a sufficient supply of water, it is necessary to practice effectively emergency water distribution and temporary damage recovery activities when an earthquake occurs. This chapter explains a series of emergency water distribution and temporary damage recovery activities to be put into practice by Osaka Municipal Waterworks Bureau just after an earthquake has hit, when no relevant information is available. Figure 4 above is a flow chart indicating initial phase activities, just after an earthquake, utilizing the function of seismic damage simulation. The Disaster Information Management System of Osaka Municipal Waterworks Bureau aims at collecting, mapping and sharing segmental information held by different departments, based on Osaka Municipal Waterworks Bureau s manual for Disaster Prevention Operations for earthquake. For this purpose, intranet and other information technologies are to be utilized. The Earthquake Damage Simulation Function is an integral part of the system. Pipe damage caused by an earthquake is often spread over a wide area, and it is not easy to correctly assess how much the damage is spread, especially just after an earthquake has hit, when no relevant or reliable information is available. 127

146 The Disaster Information Management System of Osaka Municipal Waterworks Bureau is useful at such a confused information period, as it can show, on the screen map, the result of estimated damage to main pipes calculated using the Earthquake Damage Simulation Function, and the result of estimating the area of water suspension based on the telemeter data. At the same time, the relative priority of main pipes that will need an assessment of the damaged pipe can be shown by setting a certain standard. When it is confirmed that, compared to before and after an earthquake, the water supply condition is stable in terms of water pressure at each distribution plant using the Water Distribution Information System, the damage investigation into transmission pipes and main pipes may be omitted. This system is quite useful in planning appropriate emergency water supply activities, as it can quickly and easily identify which part of the pipeline network are damaged by an earthquake. Not only does this system give damage information on main pipes, but it also offers information on the entire pipes around Osaka City, which enables estimation of the incidence of pipe damage caused in the whole city. Consequently, an estimation of time for recovering the pipes can be made available, and an effective and to-the-point personnel allocation for investigating the damage can be built up. 5. CLOSING This paper describes Osaka City s Disaster Information System and the integral part played by the Function of Seismic Damage Simulation. Further, the application of the System to actual restoration activities is also explained. Preparation of disaster measures in advance ensures an effective support of initial phase activities just after an earthquake has hit, when no relevant or reliable information is available. Such measures will also contribute to the early recovery of earthquake-damaged facilities. In addition, this system can give an estimate of the damage under scenario earthquake ground motions, which can contribute to anti-seismic measures for water supply facilities in the form of effective infrastructure investment, which this paper does not describe in detail. Simulation of potential damage by an assumed earthquake year by year is available by updating pipes data, which can be used to examine whether the investment in upgrading pipes is being effectively undertaken or not. REFERENCES [1] Proposal of the attenuation equation of peak ground motion and response spectrum using JMA-87 type accelerogram', T. Annaka, F. Yamazaki and F. Katahira, Proceedings of the 24 th JSCE Earthquake Engineering Symposium, , [2] Water Supply Line Restoration Strategy and Future Anti-Seismic Measures'(Osaka City Waterworks Bureau) 1998 [3] Damages and Analysis of Water Supply Lines Caused by Kobe Earthquake in 1995'(Japan Water Works Association) Vol.80, p

147 PRIORITIZATION OF THE SAN DIEGO WATER DEPARTMENT CIP Michael E. Conner City of San Diego Water Department Frank X. Collins Parsons INTRODUCTION The CIP originated in 1996, when a Strategic Plan for Water Supply was initiated with the support of a City wide Public Advisory Group (PAG) and the City Council Committee on Natural Resources & Culture (NR&C). The PAG and NR&C participated in water supply workshops, discussing the need for significant capital improvements to ensure a cost effective, safe, and reliable water supply. The strategic plan process identified the need to implement a Capital Improvements Program. The CIP Program Management Plan was developed in 1997 and established a list of capital projects to be implemented starting in 1998 and continuing through The projects are being financed through the sale of investment quality municipal bonds and water rate revenues. The repayment of the bonds and costs associated with the new CIP are being paid for by an increase in water rates approved by the City Council. Approval of the water rate increase to finance the first phase of the CIP was granted in Due to limited financial resources the CIP could not incorporate all of projects identified during the initial development of the program. Subsequent studies have added projects to the CIP including the projects identified as a result of the Seismic Assessment Project. A priority listing of CIP Projects needed to be made to best utilize the funding available to upgrade the water system. THE WATER SYSTEM The San Diego Water Department (SDWD) Distribution System is composed of about 3,200 miles of distribution pipe, 3 water treatment plants, 51 pump stations, 31 treated water reservoirs, 14 pressure tanks, all serving a total of 99 pressure zones. The 3 water treatment plants get raw water either from the City s own raw water reservoirs, or directly from the San Diego County Water Authority aqueducts, which receive water from the Colorado River and the State Water project through the Metropolitan Water District. These aqueducts are also used to fill some of the City s raw water storage reservoirs. The City imports approximately 90 percent of its annual water supply from the San Diego County Water Authority. 129

148 SEISMIC ENVIRONMENT The City of San Diego is exposed to potential earthquakes from a number of local area faults. The major faults include Rose Canyon, Silver Strand, La Nacion, Coronado Banks, San Diego Trough, San Clemente, Newport Inglewood, Elsinore, San Jacinto, and San Andreas. Figure 1 shows the major San Diego Faults considered in this analysis. Five possible earthquakes represent a bounding set of events that could generate the worst impacts to the water system. These possible earthquakes are shown in Table 1. Table 1. San Diego Area Possible Earthquakes Earthquake Magnitude Estimated Where Source M W Recurrence (yrs) Rose Canyon/Silver Strand La Jolla to Mexico Rose Canyon La Jolla to Mission Bay Silver Strand Downtown to Mexico Elsinore East of San Diego La Nacion ,000 Alvarado WTP to Mexico Figure 1. Major Faults Near San Diego Service Area 130

149 Figure 2 shows the mapped liquefaction and landslide zones. The liquefaction zones are characterized as being high or low susceptibilities. The landslide zones are characterized as being in one of nine susceptibilities. Each landslide or liquefaction zone can be digitized as a polygon with a typical boundary accuracy of approximately 20 feet. Figure 2. Liquefaction and Landslide Zones in San Diego Analysis of City of San Diego Water Department Facilities A Water Department s Seismic Assessment Project analyzed the City s entire water system including the raw water piping, water treatment plants, and the treated water distribution system. Additionally, an analysis reviewed the reliability of power after an earthquake. The analysis of the electrical grid was compared to the Water Department s pumped pressure zones. It was concluded that most pump stations would have electrical service restored within 24 hours of a major seismic event. The Seismic Assessment Project identified improvements needed throughout the water distribution system. Most of the pump stations, reservoirs, and water treatment plants appear to be rugged enough to withstand a probable earthquake. However, many pipelines are subject to damage in the case of such an earthquake. Buried Pipe Vulnerability Assessment The inventory of water system infrastructure was exposed to simulated earthquakes using a Monte Carlo simulation model (SERA). The SERA model incorporates the location of faults, attenuation models (which account for spectral variations with distance from the fault and local soil type), landslide models (which account for the proportion of each 131

150 slide zone that can actually slide), liquefaction models (which assess the chance of liquefaction given local accelerations and soil profile types), and the amount of surface fault offset at locations where pipes cross faults. For each site location, the peak ground acceleration, response spectral ordinate, peak ground velocity and permanent ground deformation is calculated. Pipe damage algorithms were adapted from published works by Eidinger (Eidinger 1998, Eidinger, in press). For fault crossings, the amount of offset and the pipe material are the critical parameters in determining whether the pipeline will break. Other parameters (soil backfill, angle of pipeline crossing, depth of burial) were also considered. Key Findings of the Seismic Assessment Project Significant findings are as follows: Major transmission pipelines (Murray 1, Murray 2, Alvarado 1, Alvarado 2) from the Alvarado WTP, that proceed westerly through the San Diego River drainage valley are very susceptible to breakage. The hazard which causes this is liquefaction. All possible earthquakes impose similar damage patterns in this area. This will significantly reduce the ability to deliver water to the downtown area, Mission Bay, and Point Loma. Under the listed posisble earthquakes, significant damage is expected to the pipelines that traverse from the Downtown San Diego area to Point Loma / Mission Bay area. This is caused by liquefaction and fault offset. This will seriously affect the port area next to Downtown; retail water customers in Coronado (Navy, Cal American Water); the airport area; Point Loma area; Mission Bay area; Pacifica Beach area. For earthquakes on the Rose Canyon fault, there will be significant pipeline damage in the La Jolla Soledad Torrey Pines areas. This damage will be caused by liquefaction, fault offset and landslide hazards. For all listed possible earthquakes, damage to most parts of the northern part of the water system will be moderate, and widespread long duration outages are not as likely. This reflects the generally good geology in these areas (few liquefaction areas), lack of surface faulting hazard, and distance from causative faults. Short term repairs will then continue with the medium and small diameter (20 inch diameter and smaller), until all pipes are repaired. The work crews would continue to work 12 hours per day 7 days a week until this effort is completed. Long term repairs are made following the completion of all pipes. A long term repair consists of returning to the pipe damage location, re-filling the street, compacting the soil, and making final street paving repairs. Improvements in the system will focus on restoring fire service to all high risk areas within hours after any earthquake. 132

151 Mitigation Analysis The proposed target performance goal for the SDWD is to be able to restore water to essentially every customer, via the buried pipeline distribution system, within 7 days after a Rose Canyon M 6.5, Silver Strand M 6.5, and La Nacion M6.6 or 15 days following a Rose Canyon/Silver Strand M7.2 or Elsinore M7.4 earthquake. As can be seen in Figures 3 (Rose Canyon/Silver Strand earthquake scenario), the SDWD will not currently meet these goals. The beneficial effect of mutual aid will tend to reduce the time needed to restore water service to customers, however, unless other pre-earthquake mitigation strategies are implemented, the SDWD will not meet the established goals. This suggests that the SDWD will need to initiate a Seismic Improvement Program (SIP) to implement pre-earthquake mitigation strategies which would allow it to more rapidly restore water service to as many customers as possible. System Days Lost 100% Goal Ç É B Customers with Water Service (Percent) 90% 80% 70% 60% 50% 40% BÉ Ç BÉ Ç Mitigation Ç É B Distribution Pipes Repaired Backbone Pipes Repaired System Stabilized 30% 20% B No Mutual Aid 10% É With Mutual Aid Ç Best Mutual Aid 0% Days After Earthquake Figure 3. System Restoration Curve (Rose Canyon-Silver Strand M 7.2 EQ) 133

152 VUNERABILITY ASSESSMENT The events of September 11, 2001 heightened the public s awareness of the security of the nation s infrastructure to terrorist attacks. As a result, HR 3448, Title IV, Section 1433 (hereinafter referred to as HR 3448 ), was signed into law. HR 3448 requires that each community water system serving a population of greater than 3,300 persons shall conduct an assessment of the vulnerability of its system to a terrorist attack or other intentional acts intended to substantially disrupt the ability of the system to provide a safe and reliable supply of drinking water. HR 3448 mandates that water systems serving a population of 100,000 or more, complete vulnerability assessments (VA) by December 31, Those water utilities serving 100,000 customers or more that received the Large Drinking Water Utility Security Grants set aside by the Department of Defense Appropriations Act of 2001 were also required to complete the VA by December 31, 2002 or six months after award of grant. The City of San Diego assembled a project team certified in the Sandia National Laboratories Risk Assessment Methodology for Water (RAM-W), which is a methodology acceptable to the U.S. EPA for this work. The first step of the VA was to divide the system into the three major supply systems that correspond to the treatment plant that is the primary supply for the area (Miramar (the northern system), Alvarado (the central system) and Otay (the southern system). The project team, along with select City staff, utilized this approach to identify essential facilities and assets within the City s supply, transmission, treatment, and distribution system necessary to meet an overall mission objective. Site inspections of critical facilities were then completed to determine how an adversary might attack the asset and what current physical protection system elements are in place to detect, delay and respond to such an attack. Using the risk equation, relative risk values were then determined for each evaluated asset. The risks along with possible actions to help mitigate consequences of an attack or to improve the physical protection systems were then presented to the City management staff for review. The ultimate goal was to provide a well-balanced risk for all critical assets and a prioritized implementation plan and schedule for necessary upgrades. In-depth findings and details of the evaluation methodology are presented in a confidential vulnerability report on file with the City. PRIORITIZATION Due to limited financial resources the CIP could not immediately incorporate all of projects identified during the initial development of the program and subsequent studies including the projects identified as a result of the seismic evaluation. A water rate case was developed in the Spring of 2002 to support a bond issuance for capital improvements from Fiscal Year 2003 through At that time, CIP and Water Operations staff analyzed a list of candidate projects. It was a hastily gathered process in which projects were ranked either high, medium, or low priority. There was a rush to get City Council authorization for a water rate increase combined with a window of opportunity for the 134

153 City of get the best rate on its bond financing. The Water Department completed the bond issuance with a high rating and relatively low interest obligation. However, it became evident that a more systematic approach was needed to establish future capital improvements. A prioritization process would provide the documentation necessary for bond council, rating agencies, and future grant opportunities. Given that the needs of water utilities are constantly changing, a systematic, reproducible process would document priorities at any given time. This is extremely important when there is a difference between forecasted budgets and the actual funds available for capital improvements. The prioritization list was made using the following approach. The Water Department formed an ad hoc team of senior personnel from Operations, CIP Program Management, Finance and Planning, and Water Research & Development. The panel was intended to represent the diverse interests of the divisions within the department. However, all panel members had to have a strong background in system operation, capital financing, and long term planning. The panel developed the prioritization criteria based on various water system needs. Thirteen items were identified and subsequently reduced to seven criteria. Water Operations was given double voting power in identifying the top criteria. This was done so that the planning and engineering functions would not outweigh the interests of long-term operations. Additionally, the operations and maintenance costs for pump station, reservoirs, and treatment plants usually exceed the capital costs. After the top seven criteria were selected, the team reached consensus on each of their relative weights. Criteria & Weights * Health & Safety 18 points * Regulatory Compliance 18 points * Operations Need/Reliability 18 points * Future Demand/Expansion 15 points * DHS Compliance 13 points * Project Status 9 points * O&M Cost Reduction 9 points The Water Department then updated its project lists to include all current information such as schedule, budget, project requirements and status. The panel then scored all 145 projects (FY 03 07) in a group setting using a range of 1 10 for each criteria. This approach resulted in a prioritized list of CIP projects. The results of the prioritization were presented to the Water Executive Team (WET) for their review. The WET, made up of the Director and Deputy Directors, considered the political climate as well as prior commitments to City Council and other City Departments. Additional scenarios and alternatives were explored based on these considerations before the prioritization was finalized. 135

154 SUMMARY Prioritization of the Water Department s CIP required consideration of a variety of needs including the Seismic Assessment Project and the Vulnerability Assessment. This was all done under the guiding principle of providing safe and reliable water to the customer at the lowest possible cost. The Water Department decided to move forward with the Minimum scenario which was based on the least number of scheduled CIP projects from being deleted or deferred. Based on the criteria selection process and the relative importance of other critical CIP projects, several seismic improvements related mitigation projects were deferred until funds could be made available. Additional emphasis was placed on projects mandated by the State of California, Department of Health Services and those with grant funding. The need for additional treated capacity combined with State and Federal mandates for enhanced water quality have resulted in water treatment facilities receiving the highest priority. The category of water transmission piping has the second largest amount of CIP funding commitment. These projects are increasing the distribution capacity and reliability. Expanding the recycled water distribution system will decrease San Diego s reliance on imported water. Unfortunately, the needs of the City s water treatment, storage, and delivery system far outweigh the resources currently available to improve it. The prioritization of the Water Department s CIP will be an ongoing refinement in order to maximize ratepayer value for each upgrade. References Eidinger, J., Lifelines, Water Distribution System, in The Loma Prieta, California, Earthquake of October 17, Lifelines, US Geological Survey Professional Paper 1552-A, pp A63-A78, A. Schiff Ed Eidinger, J., (ed.), Fragility Formulations for Water Systems, prepared for the American Lifelines Alliance, Report , (in press). 136

155 Seismic Practices Evaluation of Kobe Water System using Risk Management Approach Makoto MATSUSHITA 1 ABSTRACT Kobe Water has been promoting various seismic practices since the 1995 Hanshin-Awaji Great Earthquake. It has passed eight and a half years and has evaluated the improvements by the means of percents of budget spent and project completed, however, these indicators could not explain well about the current situation exposed under seismic risks. Originally seismic practices should be evaluated by the reduction of seismic risk. In order to solve this problem, application of Risk Management approach would be of use. Using Risk Management Approach all the seismic risks should be identified and then risk planning will determine how to cope with the risks to be reduced, to be avoided, to be transferred and to be held. Also it is well recognized the importance of dialogue with customers about risks, it is called risk communication. In actual risk communication will be performed through accountability and disclosure of information, it will make easy to understand for water utilities to determine the dimension of seismic investment by knowing customer s attitude. Risk Management approach will bring us the further mutual understanding with customers on investment for seismic practices. Finally all the seismic risks will be under control of Kobe Water. And we will be able to provide the best choice of strategies in case of an emergency. In this paper the author describes the details of seismic practices evaluation by using this new approach. 1 Deputy Director, Tobu Center, Kobe City Waterworks Bureau , Tanaka-cho, Higashinada-ku, Kobe, , Japan Tel.: , Fax: [email protected] 137

156 INTRODUCTION Kobe is a beautiful port city situated along the north coast of Osaka Bay and attacked by a strong near-field earthquake, the 1995 Hanshin-Awaji (Kobe) Great Earthquake, at 5:46 am on January 17. It has passed eight and a half years and we have been making efforts for reconstructing an earthquake-resistant water system. Recovery planning is stressed on facilities, which can resist for disaster and can be repaired easily water system. Facility aging and replacement are taken into consideration, these are mentioned in a report in the last workshop in August 2001, Tokyo. In Kobe water service was started in At that time city area was small and water supply capacity was only 25,000m 3 per day. Water resource was two small dams. As for the distribution system, pipe material was mostly cast iron. While Kobe was growing in both city area and population, main water source was changed from own reservoirs to Hanshin Water Supply Authority, which supplies drinking water to four cities including Kobe. Now Kobe Water serves 1.5 million customers in Kobe City area over 550km 2. Distribution system was also extended accompanied with city growth. Main facilities are two concrete tunnels, 120 service reservoirs, 48 pumping stations, 4500km of distribution pipes and telemeter/ tele-control system. Kobe Water s main concern is traditionally about distribution system. Pipe replacement has been an essential problem since 1960s. Pipe breaks were often occurred due to heavy traffic load. Colored or red water and pressure down were also occurred due to pipe deterioration and heavy water demand. Kobe Water was trying to solve these problems by replacing weak pipes to ductile iron and cement/epoxy lining pipes. However, we are still exposed by the risks of earthquake and other disasters. In the 1995 event water service had been interrupted for three months in the most damaged area. The main reason of prolonged recovery was that the damages to distribution pipes and service connections were so tremendous. It was difficult to find out leakage points of buried pipes below the road surface. Water pressure was lost because of many scattered pipe breaks occurred simultaneously. In this paper the author will explore the application of Risk Management (RM) approach for water system seismic practices. Particularly as an example of risk assessment, Kobe Water s distribution system is introduced. Many social events requiring RM have occurred in recent years, BSE and SARS problems are typical examples. These events show us importance of risk communication with customers and prior identification of various risks. We should study how to control seismic risks in normal time and use it effectively in an emergency. I think this is the effectiveness and advantage of RM approach. RISK MANAGEMENT APPROACH FOR WATER SYSTEM SEISMIC PRACTICES In this section the author will describe the procedure of RM and how to apply for seismic practices of water utilities. Fig.-1 shows general procedure of RM. First step is to identify all the risks, then to classify in four categories, namely risks to reduce, to avoid, to 138

157 transfer and to hold. Next step is to assess all the risk and evaluate current situation and to know the risks should be controlled. From this result seismic practices will be modified and water system is optimally controlled under the seismic risks. Risk communication is becoming important because seismic practices would strongly require the understanding of customers. We can minimize the disaster risk of water system with collaboration of customers. Recently it is considered to be important for utilities to determine the investment, which sometimes leads to financial problems of water utilities. Adequate feedback is necessary for this management cycle to perform well. Risk Identification Seismic Practices Risk Planning Risk Assessment - To reduce - To avoid - To transfer - To hold Feed back Risk Communication Fig.- 1 Procedure of Risk Management RISK IDENTIFICATION The first step of RM is to identify all the seismic risks. For this purposes it would be useful and informative to review the past disasters. Table-1 says that Kobe Water s past disasters. When we study the history of Kobe Water, the dates are going back to1870s. Kobe had just opened the port to the foreign counties and had got some waterborne diseases in return. One of them is Cholera. It is thought that the introduction of water system would be the most powerful strategy in order to prevent this disease. After introducing water system Kobe Water faced another risks, it was lack of water. The ancestors were forced to secure enough water resources to continue water services. First they constructed three reservoirs, however, it was not a final solution. In 1937 we set up Hanshin Water Supply Authority for 17 municipalities whose mutual problem is to secure enough water resources. We finally decided to get water from Yodo River system and Lake Biwa, the largest lake in Japan. In the past disaster we have experienced that transmission pipes were washed away 139

158 due to heavy rains in 1936 and We have also experienced a severe earthquake in It was the Hanshin-Awaji Great Earthquake. Many lessons are obtained from these events. They can be summarized as next three lessons. - To avoid fatal failure, the pipe system is a key facility. It is the weakest and always exposed under seismic risk. - To avoid citizen s panic, it should be stored enough amount of fresh water for emergency supply. - To avoid citizen s complaint and to support urban activity, the water system should be an easy to repairable water system. For the facility damage to water system, most of the damages are distribution pips and service connections. We can say that service reservoirs (tanks) and transmission tunnels have enough strength against an earthquake. Lessons regarding distribution system damage are as follows: - CIP damage ratio is very high and damages occurred in joint and pipe body. - Most of DIP damages are joint separation in loose ground. - A number of breaks lead to long recovery period and great loss of urban activities. - Quick recovery of water service requires less damage in distribution system. These lessons indicate that CIP and DIP in loose ground should be replaced to Seismic DIP. We will try to assess distribution system as an example and explore the upgrading strategy later. Other risks we currently have are identified as follows: - Failure of control and communication system - Lack of water, absolute amount of water and no pressure will not be able to meet various social needs including fire fighting and other urban activities. - Contamination (sewage, germ, microorganisms and so on.) Table-1. Natural Disaster Risks in Kobe Disaster Year Damage Solution Cholera 1870s, 80s Death of citizens Implementation of water system Water shortage 1900s s Limited supply in time and volume Sengari Dam Hanshin Water Authority Meter system Heavy Rain 1936, 1969 Loss of transmission pipe Transmission tunnel Earthquake 1995 Many beaks and leaks Water outage for three months Seismic upgrading of facilities Mutual aid agreement RISK IDENTIFICATION AND CONTROL PLANNING The next step is to plan how to control the risks. Basic attitude towards risks is categorized in four. The first one is to reduce the risks, the second is to avoid, the third is to 140

159 transfer and the fourth is to hold. Various practices for seismic upgrading are also categorized in accordance with the risks respectively. Identified risks and the control method are shown in Fig.-2. Among the risks to be reduced, the first priority is put on massive pipe breaks and leaks, and this would be controlled by means of pipe replacement. Structural deficiency will be also reduced by reinforcement and retrofitting. To be reduced - pipe breaks - structural failure - computer system - service pipe failure Pipe replacement New construction of LCTM Reinforcement and Retrofitting tunnels, dams and reservoirs Duplication and loop system Identified Risks To be avoided - complete outage - confusion - artificial error EWSS, LCTM Emergency manual Mutual aid exercises To be transferred - financial risks - emergency storage - Subsidy Mutual aid Insurance Public involvement To be held - customer s relations (complaints etc.) - geographical features Communication Information disclosure Accountability Fig. 2 Risk Identification and Management REDUCTION AND AVOIDANCE OF SEISMIC RISKS According to Fig. 2 seismic risks to be reduced and avoided are mostly for facility risks. New construction of Large Capacity Transmission Main (LCTM) can be expected to reduce the risks of interruption of main supply from Hanshin Water Supply Authority and it will be of help to emergency water service in down town area. LCTM will enable us to retrofit the aged concrete tunnels. Pipe replacement into seismic pipes can also reduce the risks of breaks and leaks and be of help to reduce the repair works. Emergency Water Supply 141

160 System (EWSS) can provide the disaster supply stations in case of an emergency and store the fresh water inside the service reservoirs. Currently water systems are operated by computer control system to maintain normal supply and check the water quality. In order to protect this function, it is important to duplicate the radio communication or to make loop system for signals to take another route. Risk avoidance is close to risk reduction. The difference is magnitude and frequency of exposed risks. For example, EWSS will avoid customer panics and confusions by storage of enough drinking water. LCTM is a backup system for existing tunnels, and make it redundant, which can enable us to avoid the fatal interruption of transmission supply. It will also avoid the traffic congestion within down town area. Emergency water supply activities will be interfered mostly by heavy traffic. Mutual aid exercises will deepen the understanding of Kobe Water system for mutual aid organizations and it will be strong tool for them to set up an adequate strategy to recover the damaged water system. TRANSFER AND HOLDING OF RISKS Risks to be transferred are, for example, financial risks and emergency storage. Government subsidy and insurance could solve this problem. In Japan water supply is thought to be municipality s work, it is not common to share the cost to insurances. It is expected some subsidy from central government but utilities should solve this problem at own risks. Insurance will be a powerful tool in case of emergency. However, it would be required prior agreement with customers to increase cost if we apply for insurances. Emergency storage by customers will mitigate the destructive damage of daily lives. This will be achieved by community outreach and disclosure of information. The importance of risk communication and the methods of communication in Kobe Water will be mentioned later. Risks to be held by water utilities are, for example, geographical condition, communication with customers and disclosure of information. This may induce misunderstanding of correct situation for customers and sometimes lead to panic. RISK ASSESSMENT examples: pipe replacement Successful Results of Pipe Replacement Reduction of risks should be evaluated and assessed by measurement of various indicators. Some risks could not be indicated numerically, however utilities should find the way of evaluation. Here the author picks up the example of pipe risks. Traditionally pipes have been replaced because of breaks, red water and pressure down. Fig. 3 shows the reduction of breaks and red (colored) water in accordance with the Ductile Iron Pipe (DIP) ratio. It means that risks of breaks and red water are decreasing. 142

161 % Red water Effective rate Ductile rate Breaks breaks colored w ater ductile rate effective rate Fig. 3 Break rate, colored water rate and DIP rate (%), Effective rate (%) Breaks red water /100km/year Pipe Damage Analysis of the 1995 Event and Progress of Pipe Replacement Fig.-4 shows Cast Iron pipe (CIP) and DIP damage location in the 1995 Kobe Earthquake and Fig.-5 indicates that CIP damage rate is very high. DIP is very strong the body itself, but the mechanical joints could be a weak point against the large ground motion. So the pipe replacement into DIP is effective strategy to reduce the risks. In Fig.-6 the change of total pipe length and pipe materials are illustrated. It is recognized that old materials are gradually converted into new one. And in near future pipes will be thoroughly replaced into DIP. The provability of pipe failure due to an earthquake will be greatly decreased. In Fig.-7 the replaced pipelines are indicated on GIS. Replacement is proceeding in coastal area where ground condition is loose and liquefaction likely to occur. We can recognize that seismic DIP routes are forming a pipe network, which means seismic network is expected to resist against an earthquake CIP damages DIP damages Fig.-4 CIP and DIP damages in downtown area P ipe D am ages D am age R ate Steel Pipe DIP(A,K,T) Pipe B ody Joint CIP PVC Fig.-5 Pipe damages and damage 143

162 Level3 (C IP,ACP,VP etc.) Level2(D IP) Level1(seism icdip,steel) Pipe Length (km )---- as of end of M arch Fig.-6 Pipe Materials in Kobe Water System Fig.-7 S-DIP replaced pipelines (As of the end of March) Find Out Most Vulnerable Pipelines Taking the pipe replacement progress into estimation of pipe breaks, coastal area (Low Layer) could be reduced a number of breaks on Table-3. In calculation of estimated breaks, we used JWWA method based on the results from Kobe Earthquake. On the contrary most vulnerable routes also can be visualized on Figure-8, on which pipe we should put the first priority for replacing. Table-3. Comparison of Estimated Pipe Breaks in 1995 and 2001 Low layer Middle Layer High Layer Total * Reduction Rate* 28.6% 18.7% 14.0% 26.9% 144

163 Legend Damage rate blue yellow red Fig.-8 Most vulnerable pipeline in 03/2003 Performance Evaluation for each distribution block Kobe Water implemented research program for performance evaluation of existing distribution system and finding out the best strategy of replacement. In this research four factors are considered, such as water quality, water pressure, earthquake and accident. Water quality means decrease of residual chlorine, it should be over 0.1mg/liter. Water pressure should be in range of 0.2 to 0.75 Mpa. Earthquake means estimated number of pipe breaks. Accident means accident provability calculated from pipe age and materials. These four factors are displayed by raider chart and compared with each distribution block in Fig.-9. This chart is helpful to find out the deficiency in each distribution block and to consider the priority of improvement program. Seismic 4 耐 震 面 1 水 理 面 Hydraulic aspect 老 朽 度 面 aged quality 2 水 質 面 4 耐 震 面 1 水 理 面 老 朽 度 面 2 水 質 面 4 耐 震 面 1 水 理 面 水 質 面 3 老 朽 度 面 4 耐 震 面 1 水 理 面 老 朽 度 面 2 水 質 面 1 Hydraulic Aspect 2 Water quality 3 Facility aging 4 Earthquake Fig.-9 Fig.-11 Performance Evaluation of Distribution System 145

164 Prioritization of Pipe Replacement Fig.-10 shows an example of prioritization of pipe replacement procedure in Higashinada No.2 reservoir area. In this evaluation we took hydraulic, water quality, facility aging and earthquake aspects into account. These values converted to the value of money as benefit and analyzed. Prioritization is based Cost Benefit Analysis (CBA). Actually the trunk main just downstream of reservoir tank would be difficult to replace immediately because it is large diameter, buried under heavy traffic and other reasons. In addition to that the new concept of ideal distribution system would not be considered in this prioritization. It should be one suggestion calculated from limited factors. Water system manager will have to make a realistic decision considering various situations including other social conditions. These decisions would reflect the opinion of concerned customer through risk communication mentioned next section. Pipe replacement priority (based on CBA) Higashinada No.2 Reservoir (13,700m 3 ) Fig.-10. Prioritization of pipe replacement RISK COMMUNICATION According to the above risk assessment for distribution system, we could get enough information about investment of pipe replacement. Which pipe is most vulnerable? What replacement strategy we should take? Which pipe we should put the first priority? We could get some information materials for communicating with customer about replacement strategy. Emergency exercises and community meetings will be opportunities to inform these facts. Internet HP and various media brochures, will be a tool for informing customers about exposed risks regarding on distribution system. Next stage we should do is to hear the voice of customer and to measure customers satisfaction in order to reflect their opinions. For this purpose Contingent Valuation Method (CVM) was applied to hear about the value of seismic upgrading in Kobe. CVM is a method to measure non-metric value, such as value of environment, value of feel at ease against 146

165 an earthquake and so on. This time Kobe Water provided some questionnaires to customers and got answers from them. Customers would study about seismic practices of Kobe Water through the questionnaires, consider how much he or she should pay for it and make answers. This value is called WTP, Willingness To Pay. Questionnaires were sent to Waterworks advisors around 250 customers. In result WTP was 235 yen (50% value) to 544 yen (average value). When we multiply number of household, total value for seismic upgrading is estimated 1,716 3,984 million yen. ( million USD) This is just an example of measuring value of seismic practice, which customers would feel by sense. In another means we could take these values into benefit account in Cost Benefit analysis, however, we should pay special attentions on communicating with customers. CURRENT STATUS The progress of seismic practices from the 1995 Hanshin-Awaji Earthquake recovery is checked every year. Current status as of March 2003 ca be described in various manner, by percent of replaced pipe length and other indicators. Pipe Replacement project are extensively being performed. When we compare the current status March 2003 with The total length of seismic pipes including s-dip increased from 358.8km(9.0%) in 1995 to km(24.1%). As for Large Capacity Transmission Main (LCTM) project, which is a large shield tunnel (d=2400mm, L=14 km) and can supply water of 400,000 m3/day, Phase 1 completed in March One of aims of LCTM is to realize multi-supply to all distribution blocks in downtown area and it enables us to shorten the recovery time in case of an emergency. Emergency Water Supply System (EWSS) has grown up from 21 in 1995 to 34 sites. Storage volume has increased and our options to water usage has also extended. It can be used not only for customer s daily use but also for system recovery works. Pipe Replacement, LCTM and EWSS are the three major seismic practices for Kobe Water. These projects will lead to our last goal; Disaster resistant and easily repairable water system. Above indicators would not able to describe exact situation of Kobe Water and would not inform what are the problems. It is desired that interesting Risk communication would be developed and customers would be educated to understand the situation. CONCLUSIONS Conclusions in this research are summarized as below: - Application of RM approach to water system will bring us a new insight for the evaluation of seismic practices. - Risk identification will show us the best choice of risk control strategy we should take. 147

166 - Risk reduction and avoidance should be calculated by annual progress of upgrading. - Pipe system upgrading is evaluated as an example of Risk Assessment. This leads to be a good tool for risk communication with customer. - Some risks are to be held within Kobe water system in later years, such as geographical condition, gravity distribution system, many tanks and pump stations, no water source within the city, depending on Hanshin Water Supply Authority. - Risk communication could be a method for deepening mutual understanding between water supplier and customers. It may leads to the correct evaluation of seismic practices. - It is important to build up management system using PDCA cycle. In this paper the author explored the application of RM approach to water system seismic practices. It is indicated that RM approach would be a powerful tool to control seismic risks and enable us to select a best strategy to get over a disaster. However, we cannot say that all the risks are revealed regarding an earthquake. Generally seismic risks are divided into two categories, facility risk and functional risk. Risk control methods would not be simulated completely. In particular non-metric risks are difficult for numerical evaluation. We should develop this management approach while considering control and evaluation methods. Continuing these efforts we can reveal really all the risks exposed under an earthquake. Furthermore risk communication would be more important and be a key method to decide the direction of water business management in near future. ACKNOWLEDGEMENTS The support from the Planning Division of Kobe Water is gratefully acknowledged. In particular Mr. T. Kanefuji and Mr. T. Kijima helped to prepare data of seismic upgrading. REFFERENCES [1] Matsushita, M., Restoration Process of Kobe Water System from the 1995 Hanshin-Awaji Earthquake (Proceedings of The 5 th U. S. Conference on Lifeline Earthquake Engineering (TCLEE) pp , August 1999, Seattle, USA) [2] Matsushita, M, Reconstruction Works and Future Perspective of Kobe Water System after Six Years from the 1995 Hanshin-Awaji Great Earthquake, (Proceedings of 2 nd JWWA/AWWA joint workshop for countermeasure against earthquake, August 2001, Tokyo, Japan) (Revised July 2, 2003) 148

167 Knowledge Management in Engineering A Methodology that can be applied to Seismic Risk Management Jim Woodhams ABSTRACT This paper examines the development of knowledge management in Thames Water, with particular emphasis on engineering, and discusses how the methodology might be applied to seismic risk management. Thames Water is the Water Division of the German RWE Group and is the third largest water and wastewater utility company in the world. It is also the largest water company in the UK, supplying 13 million domestic customers in London and the River Thames Valley. Worldwide, the business is run through four regional companies, with a combined customer base of 70 million people in 46 different countries. One key project is the water supply to the Izmit Municipality in Western Turkey, an area of high seismic activity that in August 1999 suffered a catastrophic earthquake. The recent acquisition of American Water Works in January 2003 has introduced new business in the USA situated in areas of known seismic sensitivity. The corporate knowledge strategy, known as K-Flow, was developed early in 2002 and identified four key knowledge building blocks. Regional development of these principles has followed. In UK, the Engineering team took an early lead in this work, establishing a knowledge management team and developing knowledge processes and technologies to provide a clear knowledge management focus that would deliver real benefit to the business. A key part of this work has been to engage the workforce in the strategy and to promote a knowledge sharing and collaborative working culture. A challenge now is to take the knowledge management philosophy, processes and technologies that have been developed and to test their relevance to improving the company s approach to seismic risk management. This practice could be adopted by the Workshop, either in part or in total, to further develop and enhance the inter-organization collaborative working and networking that has already been established around seismic engineering practices. 149

168 150

169 Emergency Restoration for Water Supply Following Earthquake Disasters in the City of Yokohama Abstract: Kiyoshi Naito Seibu Distribution Management Office Yokohama Waterworks Bureau Since water supply system has played an important role as one of basic lifeline functions for the public, the system should secure the minimum necessary water supplies even at the time of earthquake disasters by emergency restoration activities. Lessons from the damaged water supply facilities caused by the Great Kobe Earthquake (here after GKE) that occurred in 1995, the City of Yokohama has taken measures for prompt and effective emergency restoration activities without hindrance after the disaster adopting the arrangement for reciprocal support structure with other cities, development of facilities to accommodate support staffs from other cities, storage of materials and equipment of repair work, etc. The City of Yokohama gives report on emergency restoration activities at the workshop. 1. Overview of Water Supply in Yokohama Situated in the central part of the Japanese Archipelago and some 20 km to 40 km southwest of Metropolitan Tokyo, Yokohama has an area of about 434 km 2 and a population of about 3.5 million as of March After Tokyo, it is the second largest city in Japan. Since the time Yokohama established the first modern waterworks system in Japan in 1887, eight expansion projects have been undertaken to serve the growing population as well as the city s increasing social and economic activities. At present, Yokohama Waterworks operates water intake sources of 1,955,700 m 3 and has a daily water supply capacity of 1,820,000 m 3. The total length of water pipes (from 50 to 2,000 mm in diameter) as of March 2003 is approximately 8,900 km, which cover the entire city in a mesh format to provide for stable water supplies. 2. Emergency Restoration Measures As Japan sits on the Pacific-rim earthquake belt, which is one of the most active seismic zones in the world, it is a country that is especially prone to earthquakes. Even after entering the modern ages, major earthquakes have shaken Japan many times. The Great Kanto Earthquake that struck the Kanto region in 1923 inflicted major damage on the entire Kanto area. In particular, the damage sustained by Yokohama was especially serious. Waterworks facilities throughout the city suffered catastrophic damage and totally lost their water supply capabilities. Based on these experiences, we have proceeded with the construction and improvement of waterworks facilities by giving due consideration to their aseismatic 151

170 properties. With the enactment of the Law on Special Measures for Large-Scale Earthquake Countermeasures (the Major Earthquake Law) in June 1978 and the amendments to the Enforcement Ordinance of the Building Standards Law (the New Earthquake-Proof Design Law) in July 1980, Yokohama commenced full-scale earthquake countermeasure activities in 1981 by regarding waterworks facilities as important installations that constitute core lifeline facilities. Specifically, we have studied emergency restoration measures that can be applied even to vertical-type earthquakes that are expected to strike urban areas by drawing on the results of the GKE in 1995, and have clarified role sharing in emergency restoration between the Waterworks Bureau and support teams from other cities. At the same time, we have also strengthened the exchange of information concerning the management of facilities by preparing a support personnel acceptance manual that can facilitate smooth mutual support among waterworks bureaus in major cities. 3. Projected Earthquakes Yokohama projects that the following major earthquakes may occur in the future: a South Kanto Earthquake with the epicenter located in Sagami Bay (M7.9; scale of 6 7 seismic intensity in Yokohama), a Tokai Earthquake with the epicenter located in Suruga Bay (M8.0; scale of 5 seismic intensity), and a vertical-type earthquake with the epicenter located in the western part of Kanagawa (M6.5; scale of 4-6 seismic intensity). We are taking aseismatic countermeasures that would enable us to withstand even a South Kanto Earthquake, which is expected to inflict the greatest damage among the projected earthquakes. 4. Emergency Restoration 1) Basic concept The basic concept for emergency restoration of damaged waterworks facilities is to enable pipeline-based emergency water supplies as soon as possible in an extensive area in order to provide the minimum drinking and domestic water until full-scale restoration work is started. Accordingly, the priority in carrying out emergency restoration activities is placed on the restoration of transmission pipelines to distribution reservoirs that are important sites for emergency water supplies; pipelines connecting local disaster-prevention sites where emergency supply taps are pre-positioned to supply water through emergency water supply equipment that is affixed to exclusive earthquake-resistant pipes (ductile cast iron pipes with a bore of 150 mm that include anti-separation joints) from distribution trunk lines (consisting of distribution pipes with diameters of 400 mm or more) that are relatively resistant to earthquake damage; and pipelines connecting medical institutions, local first-aid sites, etc., where vehicle-delivered water supplies are provided. A focus is also given to the installation of temporary pipes and supply taps in areas where serious damage has been inflicted on distribution pipelines and considerable time is required for restoration. Our plans call for establishing an efficient structure for restoration activities at such times with cooperation provided by support teams from other cities. 2) Work Sharing in Restoration Activities Yokohama Waterworks Bureau With the priority given to the restoration of pipeline-based water deliveries to enable 152

171 emergency water supplies in an extensive area as soon as possible in order to provide the minimum drinking and domestic water to victims, Yokohama Waterworks Bureau will focus on taking emergency measures to prevent secondary disasters and the expansion of damage, as well as on restoring the following priority pipelines: transmission pipelines to distribution reservoirs that are important emergency water supply sites; pipelines connecting medical institutions and local disaster-prevention sites for which vehicle-delivered water supplies are requested by the Disaster Countermeasures Headquarters; and priority restoration lines consisting of distribution pipelines upstream of emergency supply taps. Support teams from other cities Support teams from other cities will be assigned to block groups consisting of Water Distribution Management Offices and Service Offices in each area under the supervision of the Water Section, and will assume responsibility mainly for general restoration work for pipes with diameters of 300 mm or below. Disaster Countermeasures Headquarters Water Section Water Supply Work Team Block Groups Waterworks Bureau Personnel Support Team from Other Cities Restoration of Priority Lines General Restoration General Restoration 5. Securing Restoration Sites Figure 1: Emergency Restoration Work Sharing 1) Facilities to Accept Support Personnel Lessons learned from the GKE indicate the need to enhance the facilities to accommodate support personnel from other cities throughout the country. On the basis of our experience from carrying out support activities in Kobe, Yokohama started the construction of facilities to accept support personnel in We completed eight buildings by 1998, as scheduled in the development plans. These buildings have a total floor area of m 2 and the capacity to accommodate approximately 30 persons. These facilities include bedrooms, meeting rooms, kitchens/dining rooms, toilets, bathrooms and laundry rooms, and are equipped with such necessities as portable toilets, emergency tools, kitchenware, bedding, television sets, radios and washing machines. 153

172 Warehouse/Meeting Room Toilets Washroom Kitchen/ Dining Room Bathroom Bedroom Hall/Bedroom 9. 0 m 27.0m Figure 2: Floor Plan of Support Personnel Accommodation Facility 2) Emergency Restoration Materials and Equipment Storage Sites Although Yokohama Waterworks Bureau has pre-positioned a certain volume of emergency restoration materials and equipment to be used at the time of earthquakes within the premises of purification plants and distribution reservoirs, we had concerns over the use of such materials at the time of earthquakes as degradation by aging has been accelerated due to the prolonged storage because a rotation system had not yet been fully established. Accordingly, this system was reviewed and materials on hand are managed on a dispersed basis at distribution reservoirs under each management area in accordance with the emergency restoration structure consisting of four Water Distribution Management Offices as the core. At the same time, in order to prevent the degradation of stored materials, material yards are divided into two sections, with materials at one side alternatively used in each fiscal year for ordinary work so that new materials are always on hand. 154

173 Disaster Countermeasures Sites Kawasaki City N onda Aoba ward Tuzuki ward Machida City miho Kouhoku Kouhoku ward Tsurumi wward Midori Tsurumi Yamato City Yazashi Seya ward Kanagawa ward Asahi ward Nishiya Hodogaya ward Nishi ward Takatsuka Izumi ward Minami Naka ward Kosuzume Totsuka ward Kounan ward Mine Isogo ward Isogo Tokyo Bay Electrical room of (former) kuden Inlet Sakae Kanazawa Kamakura City Zushi City Kanawaza ward Legend Reservoir Facility to accommodate support staff (7 sites, 8 buildings) Storage site for emergency water supply materials and equipment (14 sites) Storage site for emergency restoration materials and equipment (4 sites) Figure 3: Disaster Countermeasures Sites 155

174 6. Quantities of Materials Stored 1) Damage Ratio and Number of Damage Incidents The storage quantities of materials in stock were calculated on the basis of the damage ratio assumed by Yokohama (0.22 incidents/km) in the event of a South Kanto Earthquake (with the same level of seismic intensity as the Great Kanto Earthquake). Because the pipe bore, the type of pipe and geological features have a great impact on the damage to water pipes in various regions, such data as the damage ratio and the number of damage incidents by type of pipe and by surrounding geologic strata in Kobe, Ashiya and Nishinomiya at the time of the GKE were used as reference information in calculating the number of damage incidents, as no detailed data of the projected South Kanto Earthquake is available. Table 1: Number of Damage Incidents Based on Damage Projection Total Cast Iron Number of Damage Ratio Site Pipe Length Incidents (Incident/km) (km) Chubu Distribution Management Office Hokubu Distribution Management Office Seibu Distribution Management Office Nanbu Distribution Management Office Damage Ratio by Site , incidents/km Items taken into consideration: Bore, type of pipe and 1,569.2 geologic strata , Total 5,262 1, The total cast iron pipe length includes pipes with diameters of 75 mm or more at the time of a 1995 review. 2) Size of Pipes in Stock With respect to the pipes buried in the city, those with diameters of mm account for about 90 percent of the total length of cast iron pipes. As pipes with diameters of 400 mm or over are highly earthquake resistant, we have determined to stock pipes with diameters of mm (Table 2). While straight lengths, elbows and collars are stocked outdoors, such fittings as pressure rings, neoprene bands, and bolted connectors are stored in the warehouse of the support personnel facility to effectively utilize such facilities. While there may be shortages of pipes with diameters of 50 mm or less, it was determined not to stock such items, as they are widely used and can be procured even at the time of a disaster. Moreover, a fixed quantity of distribution and supply pipe materials are stored in each Service Office of each administrative area. 156

175 Table 2: Stock Quantity of Materials by Site Chubu Hokubu Nominal Distribution Distribution Diameter Item Management Management (mm) Office Office Straight pipes φ100 φ150 φ200 φ300 Total Seibu Distribution Management Office Nanbu Distribution Management Office Total Collars Bend (11 1/4) Bend (22 1/2) Fittings Straight pipes Collars Bend (11 1/4) Bend (22 1/2) Fittings ,325 Straight pipes Collars Bend (11 1/4) Bend (22 1/2) Fittings Straight pipes Collars Bend (11 1/4) Bend (22 1/2) Fittings Straight pipes ,200 Collars ,200 Bend (11 1/4) Bend (22 1/2) Fittings 835 1, , Agreements to Accept Support Teams from Other Cities As there are limits to the emergency activities that can be carried out by an affected city when a large-scale disaster occurs, reciprocal support agreements have been entered into with other cities, etc. Because the experience at the time of the GKE suggests the 157

176 danger that rigid rules and the strict operation of such rules may hinder early responses during the confusion immediately after an earthquake strikes, however, the Memorandum on Reciprocal Support at the Time of Disasters Among Waterworks Bureaus in 12 Major Cities was amended in 1996, and the Memorandum on Reciprocal Support at the Time of Disasters Among the Kanagawa Branches of the Japan Waterworks Association and other agreements were amended in The amended Memorandum on Reciprocal Support at the Time of Disasters Among Waterworks Bureaus in 12 Major Cities required the preparation of a disaster countermeasures manual and a support acceptance manual in order to facilitate prompt emergency measures by support teams. The basic action guidelines for emergency activities are shown in Regional Disaster Prevention Plans in Yokohama (Earthquake Countermeasures) and the Earthquake Countermeasures Handbook prepared by the Yokohama Waterworks Bureau. On the basis of these guidelines, Yokohama Waterworks Bureau prepared the Manual for Requesting and Accepting Support from Other Cities, the Emergency Water Supply Manual, the Emergency Restoration Manual, the Manual for Dispatching Support Personnel to Other Cities, the Earthquake Countermeasures Manual (Purification), etc. These manuals help clarify role sharing among each staff member and facilitate the establishment of a structure that enables each staff member to promptly take specific actions. After the GKE, moreover, the Agreement Concerning Reciprocal Support at the Time of Disasters Between the Yokohama Waterworks Bureau and the Chiba Waterworks Bureau was newly concluded in 1997 and the Agreement Concerning Reciprocal Support at the Time of Disasters Among the Kanagawa Branches of the Japan Waterworks Association was concluded in Emergency Restoration Training Under the recognition that carrying out constant training is important as a disaster prevention measure, such training is implemented on a regular and continuous basis throughout the year. We are striving to improve our abilities to respond to disasters by conducting disaster-response training on January 17, the Day of Disaster Prevention and Volunteers, and joint disaster prevention training among seven prefectures and cities and comprehensive disaster prevention training in Yokohama on September 1, known as Disaster Prevention Day. Moreover, based on the Memorandum on Reciprocal Support at the Time of Disasters Among Waterworks Bureaus in 12 Major Cities, Yokohama is carrying out joint training and technology exchange meetings with Nagoya, which is our main reciprocal-support city, in order to facilitate support for restoration activities in case Yokohama is struck by a major earthquake. These programs are implemented by visiting the other city in alternate years. Specifically, the joint training program includes such topics as support requests, support acceptance, the establishment of support headquarters, information gathering, emergency activities, etc. Technology exchange meetings are designed to increase understanding of the technologies necessary to carry out emergency restoration measures, such as measures to manage water operation, supply and distribution facilities. 158

177 9. Closing Some eight and half years have passed since the GKE, and it appears that the importance of disaster prevention measures is gradually being diluted. On the basis of the city s policy of facilitating the safety, peace of mind and stability of its citizens, Yokohama is implementing disaster prevention measures to build a city that can meet the challenges of natural calamities and realize a comfortable life for both residents and visitors. For these purposes, we are making efforts to enhance software aspects as well, such as regularly and continuously providing training programs and verifying various agreements, in addition to such hardware aspects as the establishment of facilities. We intend to promote these activities in the future in order to provide for prompt and effective emergency post-disaster restoration. 159

178 160

179 3 rd US-Japan Workshop on Water System Seismic Practices TECNICAL SESSION IV Los Angeles Water System History of Los Angeles Water System Seismic Improvements Presenter: Craig A. Davis (Los Angeles Department of Water and Power) 161

180 3 rd US-Japan Workshop on Water System Seismic Practices 162

181 History of Los Angeles Water System Seismic Improvements Craig A. Davis and Le Val Lund ABSTRACT The Water System of the Los Angeles Department of Water and Power (LADWP) has been implementing seismic improvements and hazard mitigation since the early 1900 s. The Water System was established in 1902 and shortly thereafter William Mulholland, the first Chief Engineer and General Manager, recognized the need for additional water supply for a growing city. The Los Angeles Aqueduct (LAA) was completed in 1913 to bring water from the eastern Sierra Nevada snowmelt. The LAA brought forth the first significant seismic improvement program; construction of large storage reservoirs south of the San Andreas fault crossing, both along the LAA and in the city. The LAA also provided the means for city expansion, and with it seismic improvements have been ongoing with the development of the Los Angeles Water System. Most seismic improvements were recognized and undertaken following the 1971 San Fernando earthquake as a result of the damage incurred to the Los Angeles Water System. Improvements implemented after 1971 performed well during the 1994 Northridge earthquake. Additional improvements were implemented after 1994 as a result of additional lessons learned from that earthquake. This paper discusses the seismic improvements and hazard mitigations implemented for the Los Angeles Water System since the early 1900 s, provides an overview of aspects that developed the system s resiliency that enabled it to provide service following the 1971 and 1994 earthquakes, and gives some perspectives on current seismic improvements being implemented in conjunction with necessary system changes for improving water quality. The modern developed LADWP water system seismic history provides insight and information useful to others who are interested in improving the earthquake resistance of water and other lifeline systems. Craig A. Davis, P.E., Ph.D., Waterworks Engineer, Geotechnical Engineering Group, Los Angeles Department of Water and Power, 111 N. Hope Street, Room 1368, Los Angles, CA, Le Val Lund, P.E., Civil Engineer, Los Angeles, CA. 163

182 INTRODUCTION The Los Angeles Department of Water and Power (LADWP) is the largest municipally owned utility in the United States. It exists as an independent proprietary department by virtue of the City Charter of the City of Los Angeles. The LADWP provides water and electric service to more than 3.8 million Los Angeles City residents and businesses in a 465-square mile (120,435 ha) area. The LADWP s operations are financed solely by the sale of water and electricity. Capital funds are raised by the sale of revenue bonds. No tax support is received. A five-member citizen Board of Water and Power Commissioners establishes LADWP policy. The Water System was created over 100 years ago, in 1902, when the city purchased the private Los Angeles City Water Company. William Mulholland, shown in Figure 1, became the first Chief Engineer and General Manager and began enlarging the capacity of the Los Angeles River System. Growth of the Water System is synonymous with the growth of Los Angeles. As the city grew in the early 19 th century, Mulholland saw the need for additional water supply and developed the Los Angeles Owens River Aqueduct system to bring snowmelt from the East side of the Sierra Nevada (mountain range). The Los Angeles Owens River Aqueduct was placed into service in As the need for water continued to grow, Mulholland and the LADWP initiated the water rights and preliminary planning for the Colorado River Aqueduct (CRA). The CRA was transferred for construction and operation to the Metropolitan Water District of Southern California (MWD), a state created wholesale water agency, in 1928 and placed into service in MWD is also a contractor to receive water from the California State Water Project, which has brought water from Northern California since The Los Angeles Aqueduct has been expanded and modified several times since it was initially completed, including an extension of the aqueduct into the Mono Basin in 1940 and completion of the Second Los Angeles Aqueduct in 1970, but further descriptions of these and other modifications are beyond the scope of this report. The entire aqueduct system will be referred to herein as the Los Angeles Aqueduct (LAA) System. Today the citizens and industries of Los Angeles receive their water from the Los Angeles River ground water and other basins, LAA, MWD, and recycled water. Figure 1. William Mulholland, first LADWP Chief Engineer and General Manager. WATER SYSTEM DEVELOPMENTS RELATED TO SEISMIC IMPROVEMENTS Development of the Los Angeles Water System has incorporated seismic concerns from the initial inception of the LAA to present day modifications. System components were designed with the common seismic considerations at the time of development, and many were developed with leading edge seismic technologies. The system layout was determined from normal operational and engineering requirements. Due to the need for daily water deliveries, seismic designs naturally followed operational concerns. However, the LADWP has instituted for over 100 years a culture of developing a redundant system when ever and where ever possible to allow 164

183 versatility under normal daily operational demands and emergency conditions. These redundancies have helped create a water system that is very versatile and resilient to strong seismic shaking. Supply Sources The LADWP has developed many water supply sources to allow redundancy for normal and emergency operations. Figure 2 shows the supply sources from the LAA, CRA, California Aqueduct, and the Los Angeles River groundwater and other basins. In addition, recycled water is used to supplement potable water for irrigation and industrial purposes, but is not considered a primary supply in the event of an earthquake. The numerous tanks and reservoirs contained in the water system store water from these various sources for distribution throughout the city. The numerous connections with the MWD throughout the City provide water from the CRA and California Aqueduct for distribution. In the event of an earthquake, the water system draws upon the many available sources to provide a reliable water supply. Following an earthquake, these supply redundancies allow water to be provided to unaffected areas of the city with little or no interruption, and also significantly aid in post-earthquake recoveries to severely damaged portions of the water system. Within the severely damaged portions of the system, the redundant supplies and in-city storage significantly reduce the outage time an average resident experiences following an earthquake. Figure 2. Map of California showing LADWP water supply sources and the San Andreas Fault. Figure 3 shows the LADWP water supply and transmission system. Terminal storage for the LAA was developed at the City s northern limits, in an area now called the Van Norman Complex (VNC). The VNC, located in Figure 3, is an important site for collecting water supply and distributing it to the city because it serves as the LAA terminus and also receives the majority of MWD water. As the main water supply hub for the entire City, about 75 percent of the City s water passes through the VNC before it is distributed throughout the City. Once water leaves the VNC it is sent directly into portions of the distribution system and transmitted to reservoirs and tanks throughout the City. 165

184 Los Angeles Aqueducts Van Norman Complex Figure 3. LADWP water supply and transmission system showing the first and second LAA entering the city from the north, major trunk lines (solid bold lines), MWD supply lines (dashed lines), reservoirs (labeled), major tanks (solid circles), other LADWP water facilities as labeled, freeways (labeled solid lines), within the Los Angeles City boundaries. Storage The Lower and Upper San Fernando Dams, which retained the Lower and Upper Van Norman Reservoirs were constructed in the VNC in 1913 and 1921, respectively, to provide terminal storage for the LAA. All other reservoirs used for normal operation along the LAA were located north of the San Andreas Fault. As shown in Figure 2, the San Andreas Fault generally runs from south to north nearly the entire length of the State of California, and the LAA intersects the fault about 25 miles north of the City. William Mulholland recognized that movement on the San Andreas Fault could sever the LAA. As a result, in addition to the normal operational storage, Mulholland identified emergency storage requirements south of the San Andreas Fault, which was met in 1926 by construction of the St. Francis Reservoir in San Francisquito Canyon, shown in Figure 4, and its associated hydroelectric power plants [1]. Unfortunately, as shown in 166

185 Figure 5, the St. Francis Dam failed catastrophically in March of After the dam failure, the need for LAA emergency storage south of the San Andreas Fault still existed and Bouquet Canyon Reservoir, shown in Figure 6, was constructed as the replacement for St. Francis Reservoir. Mulholland also recognized the need for emergency storage within the City, which resulted in the construction of several large reservoirs including Hollywood, Franklin Canyon, Stone Canyon, Silver Lake, Chatsworth, and Encino Reservoirs shown in Figure 3, and the Upper and Lower Van Norman Reservoirs on the VNC. At its peak in 1971, the Los Angeles Water System had approximately 24 large reservoirs to maximize storage and water distribution within the City for normal and emergency operations. At the same time, the City s water system contained eight smaller operating reservoirs, numerous water tanks, and six large water storage reservoirs along the Northern LAA. Figure 4. St. Francis Dam and Reservoir placed in service in 1926 to provide water storage south of the San Andreas Fault. Figure failure of St. Francis Dam. 167

186 Dam No. 1 Dam No. 2 Figure 6. Bouquet Canyon Reservoir constructed to provide replacement storage south of the San Andreas Fault after the St. Francis Dam failure. Bouquet Canyon Reservoir Dams No. 1 and 2 were the first compacted earth fill embankment dams constructed using modern compaction equipment and testing methods developed by Ralph R. Proctor. Supply Transmission Lines/Trunk Lines. Figure 3 shows the LADWP water supply system within the City limits. In order to provide a reliable water supply throughout the system, numerous large diameter [up to 120-inches (3048 mm) in diameter] supply trunk lines originate from the VNC and other supply sources and extend to various reservoirs and tanks throughout the City. Other trunk lines interconnect the main supply lines to allow for distribution of different water sources to many parts of the City. The trunk lines are also equipped with many valves that allow isolation of the water supply. The trunk line interconnectivity and isolation capability provides great versatility in the ability to supply water to many parts of the system from many sources. In many cases, damaged trunk lines can be isolated for repair without a significant water outage. At the same time, water can still be supplied to damaged portions of the distribution system from other sources in order to provide the needed water surcharge for finding and repairing pipe breaks. The supply line redundancy is a significant component in creating a system resilient to severe earthquake effects, which was proven to be effective in the 1971 San Fernando and 1994 Northridge earthquakes. Pressure Zones The Los Angeles distribution system is broken into approximately 115 water pressure service zones. Water pressure in the different zones is controlled by the reservoir elevations, pressure head created by pumping, and pressure drop created by regulating stations. A regulator station causes a desired head loss to prevent high service pressures from entering lower ground elevation regions. The numerous pressure zones are the result of natural topographic changes throughout Los Angeles. Many pressure zones in Los Angeles can be supplied by multiple sources, multiple storage facilities, and multiple supply lines. Those that have a single source of supply are generally very small zones. The numerous pressure zones aid in the ability to isolate and provide redundancies to block areas of the distribution system that might be damaged by an earthquake, which enhances the system s seismic resiliency and reliability. 168

187 SEISMIC IMPROVEMENTS PRIOR TO THE 1971 SAN FERNANDO EARTHQUAKE The most significant water system seismic improvements, outside of those described as part of the system developments, were initiated following the 1971 San Fernando Earthquake as a result of lessons learned from earthquake impacts on modern urban infrastructures. However, there are several important examples of significant seismic related activities and improvements implemented prior to the 1971 earthquake. Building Code The first edition of the Uniform Building Code was published in 1927 and contained an optional appendix prescribing a lateral force coefficient of 0.1g. Following the 1933 Long Beach, California earthquake, the City of Los Angeles, County of Los Angeles, City of Long Beach and other municipalities, primarily in southern California, adopted seismic provisions that commonly used a lateral force coefficient of 10% of gravity, following the precedent established in Japan. Since the LADWP construction was in the City of Los Angeles, it was required to obtain City building permits that complied with similar seismic codes. Since 1933, all Water System structures have been designed to meet the seismic codes in force at the time with the addition of an importance factor when appropriate, which increases the minimum code seismic coefficient. The importance factor is applied by recognizing the need for the facility to be operable after a seismic event. In more recent times, many critical facilities have been designed using site specific seismic criteria, exceeding minimum code requirements. Strong Motion Instrumentation In the 1950 s early versions of strong motion instruments, called Wilmont Seismoscopes shown in Figure 7a, were installed at a number of LADWP dams. The seismoscopes were usually placed in pairs with one instrument on the dam crest near the center and the other on the dam abutment. These instruments record motion of a pendulum with a needle that scratches a blackened hourglass located at the top of the pendulum. Several seismoscopes are maintained and presently remain in service. a. b. Figure 7. Strong motion instruments, a. Wilmont Seismoscope, b. Kinemetrics SMA-1. Strong motion instrumentation and data collection was implemented in 1965 to evaluate the performance of buildings and structures in the immediate aftermath of an earthquake. Code mandated instrumentation about that time was required for most buildings over six stories high in 169

188 certain California cities like Los Angeles, which adopted the instrumentation provisions of the Uniform Building Code [2]. The LADWP General Office Building (John Ferraro Building, JFB) was under construction at that time. Strong motion accelerographs (SMA) were installed in the JFB basement, 7 th floor, and top floor. These were the first instruments to be installed in the City and the first to be manufactured by Kinemetrics. One of the first SMA-1 instruments installed in the JFB is now preserved and placed in the front office of Kinemetrics in Pasadena, California. Many important LADWP sites have been instrumented with strong and weak motion recorders for seismological and engineering purposes. These instruments have been installed by the LADWP or by other organizations with cooperation from the LADWP, including other government organizations such as the California Geological Survey, United States Geological Survey, universities such as the California Institute of Technology, University of Southern California, and earthquake research centers such as the Southern California Earthquake Center (SCEC). Figure 7b shows a Kinemetrics SMA-1 instrument that is installed at many LADWP dam sites. Information obtained from installations on LADWP sites has made very significant contributions in seismological research and earthquake engineering developments, most notable are the LADWP recordings made during the 1994 Northridge Earthquake [3]. Figure 8. Ralph R. Proctor, inventor of modern soil compaction and testing methods. Soil Mechanics Ralph R. Proctor, shown in Figure 8, was a LADWP Water System Civil Engineer and innovator of modern geotechnical engineering and soil mechanics. The methods for soil compaction and testing developed by Proctor have proved to be very useful in improving the seismic performance of geotechnical structures, such as embankment dams and fills, and building foundations throughout the world. Many LADWP dams constructed before 1928 used hydraulic fill methods for placing soil, which left relatively loose and weak soil deposits making up the embankment dam. A few dams built in the 1920 s used newly emerging mechanical compaction methods, which provided some strength improvement over hydraulic fill, but both hydraulic fill and mechanical fill placement methods were used through the 1920 s. The LADWP recognized the need for improved dam construction techniques. With the advent of improved mechanical equipment, Proctor developed specifications to construct a heavy sheepsfoot roller and methods for placing and compacting soil for the primary purpose of constructing embankment dams [3a], but the methods are useful for any compacted earthen structure. He also developed testing methods and equipment to monitor construction of dams and to measure the in-place density of soil in the field and the shear strength of soil in the laboratory. Additionally, Proctor developed a specialized tool called the Plasticity Needle, now commonly called the Proctor Needle, that was used for measuring penetration resistance to indirectly measure field density, shear strength, and moisture content. The Plasticity Needle penetration resistance was calibrated in the laboratory 170

189 using soil test specimens compacted at different moisture contents to allow rapid identification of soil density and moisture content in the field. Figure 9a shows a partial set of needles with the penetrometer and the laboratory compaction mold and rammer developed by Proctor. The use of needles for aiding in compaction testing has been standardized by the American Society for Testing and Materials (ASTM) as standard ASTM D1558. Figure 9b shows the initial implementation of Proctor s methods during the construction of Bouquet Canyon Reservoir ( ); the two dams retaining this reservoir became the first compacted earth fill embankment dams ever constructed using rigorous quality control techniques. Bouquet Canyon Reservoir was constructed primarily for seismic improvements to the water system in order to provide emergency water storage along the LAA south of the San Andreas Fault to protect the City s water supply in the event of a major earthquake. Modifications of the procedures for testing relative soil compaction developed by Proctor have been standardized in ASTM D698 and ASTM D1557, commonly referred to as the Standard Proctor and Modified Proctor testing procedures, respectively. The Proctor tests, and the methods developed at the LADWP for placing and compacting soils, are today international standards for improving soil performance for every day working loads and seismic concerns, and may possibly be the single most significant factor for improving the seismic resistance of engineered structures worldwide. a. b. Figure 9. a. Soil compaction testing equipment: mold (top left), rammer (top right), and Proctor Penetrometer and Needles (bottom); b. Soil compaction for Bouquet Canyon Reservoir Dam Number 1. The first large scale use of modern soil compaction with sheepsfoot rollers and testing methods developed by Ralph R. Proctor at the Los Angeles Department of Water and Power. For these dams the sheepsfoot rollers followed a single pattern until the soil had been rolled a minimum of 16 times. Embankment Dams Since the mid-1930 s, all dams were constructed of well compacted earth fill with special seepage control measures; most of which were built under Proctor s supervision. Stone Canyon and Encino Dams were originally constructed in the 1920 s using combinations of hydraulic and newly emerging mechanical rolled fill methods, respectively. In the 1950 s and 1960 s, the capacities of these two reservoirs were increased to provide additional in-city water storage. As part of the reservoir improvement programs, the new embankments were designed to have stable slopes using a 0.1g lateral seismic coefficient and a significant volume of the existing embankment dams and underlying alluvium were excavated prior to placing well-controlled compacted earth fill. The purpose for removing the existing dam and alluvium materials were to 171

190 help ensure that the new embankments were able to resist strong seismic shaking. For both Encino and Stone Canyon Dams, portions of the existing loose, relatively weak dam fill and alluvium were left under the downstream slopes of the re-constructed dams. All of the dams constructed and significantly modified since 1930 have proven adequate to resist seismic forces under actual shaking and using modern state-of-the-art analyses. Therefore, the advancing dam design and construction methods significantly improved seismic stability. LADWP earth fill dams constructed between 1930 and 1971 were conservatively designed and constructed with seismic stability in mind. LADWP records indicate that dams designed at least since the early 1950 s, such as Stone Canyon Dam, were evaluated for seismic slope stability using a minimum of 0.1g horizontal coefficient. Earlier dams, such as the Bouquet Reservoir dams in the early 1930 s, were not designed using seismic stability analyses as we use today; thus, these seismic improvements are considered to be more intuitive than analytical. Figure 10 shows a cross-section of the Lower San Fernando Dam (LSFD). Stability concerns for the LSFD in the 1930 s lead to the placement of a compacted fill buttress on the downstream slope. The LSFD retained the Lower Van Norman Reservoir and was primarily constructed in 1913 using standard hydraulic fill methods. The downstream buttress fill was placed in The buttress fill material was obtained from a large storm water improvement project undertaken on the VNC to bypass storm water around the west side of the Upper and Lower Van Norman Reservoirs [4]. Upstream pre-1971 section reconstructed section Pre-1971 centerline Existing centerline Downstream exising section 1971 slide debris 1913 hydraulic fill blanket 1940 berm Figure 10. Cross-section of the Lower San Fernando Dam showing the original 1913 hydraulic fill, 1940 downstream berm, cross-section geometry in 1971 prior to upstream slope failure, and reconstructed shape following 1971 earthquake. The LSFD provides the first fully documented case in which a LADWP dam was analyzed for seismic slope stability [5]. The LSFD was evaluated in the 1960 s by LADWP engineers, under the review of an eminent board of consultants including experts such as Dr. C.F. Richter from the California Institute of Technology and C. Martin Duke from the University of California at Los Angeles, using a pseudo-static lateral force coefficient, similar to common simplified slope stability methods used today. However, liquefaction was not considered to be a concern for embankment dams, mainly because little was understood about liquefaction at that time [6]; therefore, liquefaction was not considered as part of the seismic stability analysis. Nevertheless, the stability evaluation indicated the LSFD may have slope stability problems when shaken by a significant earthquake. As a result of this evaluation, a large seismic improvement project was implemented that consisted of constructing the Lower Van Norman Bypass Pipeline and Reservoir (Bypass Pipeline and Bypass Reservoir). Once the Bypass Pipeline and Reservoir were constructed, the LSFD was planned to be removed and replaced with a new compacted earth fill dam. The 60-inch (1524 mm) diameter Bypass Pipeline was constructed in 1968 and the 240 acre-foot (296,036 m 3 ) Bypass Reservoir was completed and placed in service in November The final phase of the seismic improvement project, reconstruction of the LSFD, was never completed as a result of a large liquefaction induced slide on the LSFD upstream slope during the February 9, 1971 San Fernando Earthquake, only two months after placing the Bypass Reservoir 172

191 in service. The new Bypass Pipeline and Dam, shown in Figure 12 of this report, were not damaged by the San Fernando Earthquake SAN FERNANDO EARTHQUAKE The February 9, 1971 San Fernando Earthquake occurred on the San Fernando Fault with a moment magnitude M w of 6.7 and an epicenteral distance of approximately 11 km north-east of the VNC [7]. The VNC overlies the westerly boundary of the fault that ruptured toward the southwest. The main fault surface ruptures primarily occurred east of the VNC, with small surface ruptures in the vicinity of the LSFD east abutment [8]. Small surface ruptures were also identified in the vicinity of the LSFD and the area to the north of the dam. A limited number of seismic recordings were made in the near-fault region, including seismoscopes located on the LSFD crest and east abutment [9]. The San Fernando earthquake damaged water system facilities, especially those in the earthquake near-field. The most significant water system damage occurred on the VNC [10] with liquefaction induced slides resulting on the Upper San Fernando Dam (USFD) and LSFD. Figure 11 shows damage to the LSFD, after the reservoir was partially drained; the upstream slope slid over 200-feet (61 m) into the reservoir and the dam lost 30-feet (9.1 m) of freeboard, leaving only 5-feet (1.5 m) of freeboard above the reservoir water surface after the earthquake. Draining of the lower reservoir began soon after the earthquake and the upper reservoir was permanently lowered. Loss of the valuable VNC storage in the Upper and Lower Van Norman Reservoirs revealed the importance and timeliness of the Bypass Pipeline and Reservoir seismic improvements; without these two facilities the water system would have had extreme difficulty in supplying water to the City. Figure 11. Photograph of the Lower San Fernando Dam, looking to the west from the east abutment area, showing upstream slope slide into the reservoir. Additionally, breaks occurred on many large diameter trunk lines and several thousand distribution pipes and service connections throughout the system, primarily in the San Fernando Valley area. Damage also occurred to LAA pipes and channels, water storage tanks, older pumping and chlorination station buildings, unreinforced masonry buildings, water operating district yard buildings, and embankment dams in addition to the USFD and LSFD described above. 173

192 POST 1971 SAN FERNANDO EARTHQUAKE SEISMIC IMPROVEMENTS Many lessons were learned from water system seismic damage in As a result, numerous seismic improvements were implemented in the years following the earthquake. The LADWP also significantly aided in the data collection and documentation of water system damage [11], allowing for great advances in lifeline earthquake engineering research and developments [12] that made important contributions to LADWP and other organization system improvements worldwide. The seismic lessons, implementations, and mitigations are too numerous to describe in detail in this paper. Only some highlights will be described herein. Upper San Fernando Dam Bypass Reservoir Los Angeles Reservoir Lower San Fernando Dam Figure 12. Van Norman Complex improvements before and after the 1971 San Fernando Earthquake; Photograph identifies the Lower Van Norman Bypass Reservoir, Los Angeles Reservoir, and reconstructed Lower San Fernando Dam. As a result of damage sustained to the USFD and LSFD, an extensive research program was funded by the LADWP and California Department of Water Resources Division of Safety of Dams (DSOD) and undertaken by the University of California Berkeley under the direction of Dr. Harry Seed. Results of this study [6] identified liquefaction and slope stability as the primary cause of damage to the dams and made several recommendations for improved seismic stability evaluations of dams including advanced numerical computer modeling and dynamic laboratory testing. The DSOD required all jurisdictional dams in the State of California to have seismic stability evaluations. The LADWP implemented recommendations from the study to the greatest extent possible and helped lead the geotechnical industry in dam stability evaluations throughout the 1970 s. The LADWP obtained support from Dr. Seed and other prominent earthquake researchers, expert consultants, and review boards before developing final conclusions and implementing recommendations from the relatively advanced testing and analyses. Custom cyclic triaxial chambers were designed by LADWP engineers and fabricated in the LADWP machine shop to be used with an MTS Universal Electro-hydraulic Testing Machine to simulate earthquake loading. State of the art finite element programs were used by LADWP engineers and consultants to analyze all DSOD jurisdictional dams. Stability analysis results showed that many existing dams were seismically stable and identified several dams that were in need of improvement or replacement. All dam stability analyses were performed with site specific seismic design criteria for Maximum Credible Earthquakes (MCE). An extensive dam 174

193 improvement program was undertaken and seismic deficiencies for dams on at least 12 reservoirs were mitigated by removing and reconstructing the dams, removing the dams from service, or restricting the reservoir high water level. In 1979, Los Angeles Reservoir, shown in Figure 12, was placed into service on the VNC to replace the original Upper and Lower Van Norman Reservoirs. In the 1980 s the LADWP worked with consultants from Dames and Moore to complete the first practical non-linear effective-stress deformation analyses of Pleasant Valley and South Haiwee Dams [13][14][15], which are both located in the Owens Valley. Figure 13. Seismic vulnerability assessment for pumping stations performed after the 1971 San Fernando Earthquake. The 1971 San Fernando earthquake also caused damage to some of the older pumping stations, chlorinating stations and water operating district yard buildings. Following the 1971 earthquake, a seismic water system vulnerability assessment (SVA) was made under the direction of the second author for all pumping stations, chlorination stations, maintenance and construction yards, tanks and sumps, reservoirs, and well installations in the City [16]. Figure 13 shows a SVA example for pumping stations excerpted from [16]. Surveyors and construction inspectors made the vulnerability assessments using visual inspections, after being trained, to look for anchorage of equipment, type of building construction, roof to wall connections, hazardous materials storage, electric substation facilities, flexible pipe-to-structure connections, etc. Questionable building construction and other seismic deficiencies were referred to the appropriate engineers for evaluation. The facilities were graded as low, medium, or high for importance to 175

194 system and cost. A similar SVA report was prepared in the Owens Valley and the Mono Basin. The SVA reports were used for a multi-year budgeting program to implement the seismic upgrade of these facilities. Some of these facilities were seismically retrofitted and others were completely reconstructed. All pumping station, chlorination station, and district yard buildings were evaluated and improved as necessary. Following the 1971 earthquake, two large diameter pipes were retrofitted to help mitigate future earthquake damage. These mitigations were undertaken on the Second Los Angeles Aqueduct (SLAA) at Terminal Hill [17] and the Granada Trunk Line (GTL) on the VNC [18][19] as a direct result of damage inflicted by the 1971 earthquake. The SLAA was damaged from a unique combination of large ridge shattering and slope deformations. Special pile supported piers were installed to support the pipe traversing the slope, flexible couplings were installed at the top of the hill, and rock anchors, as shown in Figure 14, were installed to restrain future ridge movements [20]. The GTL suffered damage to many pipe joints and mechanical couplings as a result of large liquefaction induced ground movements [18][19]. Figure 15 shows examples of wrinkled bell and spigot welded steel joints that occurred on the GTL, SLAA, and other pipes. Mitigation of the GTL to resist future ground deformations included installation of 12 special unrestrained long barrel mechanical couplings, shown in Figure 16, designed and constructed by LADWP personnel. A few trunk lines constructed after the 1971 earthquake to supply water from the VNC were routed specifically to avoid known liquefaction and lateral spreading areas that were observed following the 1971 earthquake; an example is provided by the Foothill Trunk Line construction in the 1980 s around the Juvenile Hall Slide lateral spread. Figure 14. Terminal Hill cross-section showing crushed/shattered rock zone, damage and displacement locations from the 1971 San Fernando Earthquake (circled numbers), post-1971 tieback anchor installation, and 1994 damage and displacement locations (non-circled numbers) on the Second Los Angeles Aqueduct pipeline. Modified from [20]. As a result of lessons learned and damage inflicted to water storage tanks in 1971, the LADWP engineers became involved in the development of new tank design criteria and were instrumental in establishing the AWWA D100 standard seismic design criteria for tanks. The LADWP also actively investigated and evaluated expected performances of tanks constructed using different methods (e.g., reinforced concrete, various pre-stressed concrete methods, welded steel with different anchorage methods). Results of these evaluations identified techniques which are now common in LADWP tank designs such as sketch plates for welded steel tank anchorages and use of DYK type pre-stressed concrete tanks that have post-tensioned seismic cables [21]. The LADWP commonly adds special seismic details that exceed industry standards (e.g., AWWA D100) and contractor recommended details (e.g., DYK). 176

195 a. b. Figure 15. Compression wrinkling of welded bell and spigot joint; a) pipe joint in place, b) cut out of wrinkled pipe joint. Figure 16. LADWP long barrel mechanical coupling design for the Granada Trunk Line (GTL) to accommodate displacements from permanent ground movements NORTHRIDGE EARTHQUAKE The January 17, 1994 Northridge Earthquake occurred on an unmapped blind thrust fault with a M w 6.7 and an epicenter located in the Northridge/Reseda area of the San Fernando Valley [7]. The ruptured thrust fault was below much of the San Fernando Valley, which subjected many critical water system facilities to strong near-source pulses. Many seismic recordings were made in the near-fault region, including several important recordings at LADWP facilities on the VNC [3], which included some of the largest ground motions ever recorded. The Northridge earthquake caused considerable damage to water system facilities, especially those in the earthquake near-field, but not nearly as significant as the damage caused by the 1971 earthquake. Many breaks occurred on large diameter trunk lines and several thousand distribution pipe and service connection leaks resulted, primarily in the San Fernando Valley [22]. Damage also occurred to LAA pipes and channels, water storage tanks, and some buildings [23]. Figure 17 shows water flowing from the GTL in Balboa Boulevard. The leak resulted from ground deformations [24] causing damage to the pipes as shown in Figure 18. A natural gas line was also damaged, which ignited after a car stalled in the water. 177

196 Figure 17. Water flowing into Balboa Boulevard from damaged Granada Trunk Line (GTL). A fire ignited from a natural gas pipe leak. a. b. Figure 18. Damage to Granada Trunk Line (GTL) in Balboa Boulevard resulting from permanent ground deformations: a. compression wrinkling on south end of damage zone, b. tension separation on north end of damage zone. As was the case during the 1971 San Fernando Earthquake, the greatest water system damage in 1994 occurred in the northern San Fernando Valley, in and around the VNC [25]. The most significant problems were to water supply pipelines and channels. Restoration of these damages was critical for restoring water supply before the in-city storage was depleted. There was some damage to embankment dams, but not significant enough to affect operations, except for a small dike on the San Fernando Power Plant Tailrace, shown in Figure 19, which failed and affected supply from the LAA [26]. Several older tanks were damaged in the Santa Susana and Santa Monica Mountains [21]. 178

197 Figure 19. Tailrace dike failure following the 1994 Northridge Earthquake. Nearly all of the improvements made to building structures, tanks, and dams following the 1971 earthquake, and all new facilities constructed after 1971 performed well in the 1994 earthquake, proving that the lessons learned and mitigation strategies implemented were very beneficial. However, there were some additional lessons learned from the 1994 earthquake, including the identification of necessary improvements to buried pipes and the effects of nearsource ground motions on above ground and buried structures and geotechnical structures. All of the damage, lessons learned, and seismic mitigations resulting from the 1994 earthquake are too numerous to report herein. Only a few highlights are described below. POST NORTHRIDGE EARTHQUAKE SEISMIC IMPROVEMENTS Seismic improvements following the 1994 earthquake consisted of (1) implementing knowledge and design strategies that had previously been learned from the 1971 earthquake, but not fully carried out on all facilities due to priority and funding conflicts with water quality and other issues, and (2) identification of new earthquake related issues requiring further research. An example of implementing previous lessons is provided in tank performances. Many damages to older tanks, which included breakage of rigid inlet-outlet lines, roof damage, and movement of unanchored tanks, were of no great surprise after evaluating the post-earthquake damage. Much of this damage type was identified in the 1974 SVA report [16] and recommended to be reviewed further. In addition, knowledge on tank performance and design were proven effective by the fact that all newer tanks performed well [21]. As a result, an improvement program was undertaken after the 1994 earthquake to add flexible connections to tank inlet-outlet lines where needed that allows extension, compression, and rotation of the connections. Figure 20 shows one type of flexible connections used on tanks. 179

198 Figure 20. Example of tank inlet-outlet line flexible connections. Examples of mitigations and improvements resulting from new knowledge learned from the 1994 earthquake include: 1) Figure 21 shows failure of the 96-inch (2438 mm) diameter corrugated metal pipe Lower San Fernando Drain Line No. 1 after it sustained a complete lateral collapse. This pipe was replaced with a reinforced concrete pipe surrounded by a heavily reinforced concrete encasement. The LADWP and the University of Southern California performed research on this unique pipe failure and determined that the combination of dynamic pore pressure build up in the surrounding bedding soil and large near source shear strain pulses led to this damage [27][28][29]. a. b. Figure 21. Failure of Lower San Fernando Drain Line No. 1 corrugated metal pipe. a. view looking south, b. view looking north. 180

199 2) The SLAA at Terminal Hill and GTL on the VNC both sustained damages during the 1994 earthquake, even though significant mitigations were undertaken following the 1971 earthquake. Investigations were carried out to better identify causes of damages to these facilities [18][19][20]. Additional mitigation strategies were evaluated and in each case, relocation was determined the best alternative. A seismic mitigation project relocating the GTL out of the liquefaction induced ground deformation zone has been completed. Relocation of the SLAA through a 600-foot (183 m) tunnel and a 300-foot (91 m) vertical shaft at Terminal Hill is currently in design and planned to be completed by ) Figure 22a shows the 54-inch (1372 mm) diameter Van Norman Pumping Station Discharge Line, an above ground welded steel pipeline supported by concrete piers and steel H- piles. The pier supports for this pipe were originally damaged in 1971, but did not require repairs; instead they were left in their post San Fernando Earthquake tilted condition. The pipe did not leak and damage was limited to leaning of piers, which continued performing their intended function, and considered to be a minor structural problem. In 1994, the piers were further damaged; causing significant leaning and failing several ring girders, but the pipe did not leak. Follow-up investigations identified liquefaction and lateral ground movement as the mechanism damaging the piers and underlying H-pile foundations. Figure 22b shows the pipe after completion of a seismic repair and mitigation project. Figure 23 shows details of the seismic mitigation designed and constructed by LADWP engineers and construction crews, which consisted of lowering the pipe near the ground surface and placing the supports on low-friction base isolators made of Teflon and stainless steel. This is the first known use of base isolation to mitigate liquefaction induced ground deformation effects to a water pipeline. a. b. Figure 22. Van Norman Pumping Station Discharge Line: a. Earthquake damage to pier foundations and ring girders, and b. Completed base isolation pad and protective roof. 4) The High Speed and Bypass Channels are important LAA water conduits, which also flow a significant amount of California Aqueduct water from MWD connections, on the VNC that were damaged in 1971 and 1994 from permanent ground deformations in areas where the channels pass through weak, saturated natural soil deposits [30][25]. Case studies from multiple earthquakes identified the differing effects that transient and permanent ground movements had on channel damage and that there is possibly a local permanent deformation regime causing different levels of channel damage within the larger lateral ground deformations that extend well beyond the channels [31][32]. The case study results are currently being evaluated in greater 181

200 detail to determine the feasibility of performing soil improvement techniques to cost effectively mitigate channel damage for future earthquakes. Figure 23. Details of LADWP designed Teflon bearing surface base isolator for the Van Norman Pumping Station Discharge Line. As a direct result of the Northridge Earthquake, the LADWP has significantly aided the research and development of new technologies and advancement in earthquake engineering knowledge through cooperative research with the University of Southern California (USC), Southern California Earthquake Center (SCEC), Pacific Earthquake Engineering Research Center (PEER), Multidisciplinary Center for Earthquake Engineering Research (MCEER), the United States Geological Survey, and others. The numerous cooperative research undertakings have significantly helped develop knowledge and understanding in near-source ground motions [33][34][35], understanding of underground pipe behaviors [28], permanent ground deformation measurements [36], development of liquefaction and lateral spreading databases and improved models [37], development of GIS-based earthquake evaluations and modeling [38][36][37][39], water system component fragility curves [40][41] and other projects that have improved seismology and earthquake engineering practices. PRESENT AND FUTURE PERSPECTIVES The LADWP continues to provide support and leadership in advancing earthquake engineering practices. For example, LADWP provided numerous pipe samples and support to MCEER for testing and modeling of welded bell and spigot steel pipe joints to better understand the mechanism leading to buckling, as shown in Figures 14 and 17a, of these joints under axial compression [42]. The research was then applied to development of strategies that can be implemented to reduce the potential for bell and spigot joint buckling by use of welding techniques, fiber reinforced composite wrapping, changing the methods for manufacturing the joint shape, etc. Research for this project is in its final completion stages [43]. The LADWP is presently undertaking an extensive capital improvement program to meet the requirements of the United States Environmental Protection Agency (EPA) and California State Department of Health Services (DHS) requirements stipulated in the Surface Water Treatment Rule (SWTR) and Disinfection Byproducts Rule (DBR). Descriptions of these rules are beyond the scope of this report; however significant water system changes are necessary to meet their requirements. System changes include the removal of Hollywood, Encino, and Stone Canyon reservoirs from normal operating service, which places a much greater importance on the VNC and Los Angeles Reservoir for water supply throughout the City on a daily basis. These three reservoirs are planned to retain storage for emergency supply purposes. Hollywood and 182

201 Encino Reservoirs have already been removed from service and Stone Canyon Reservoir is scheduled to be removed in Silver Lake Reservoir is planned to be similarly removed from normal operating service in the near future and replaced with alternate covered storage, of reduced capacity. The Los Angeles Reservoir is planned to be divided in half with a new embankment dam to allow for greater system flexibility and each side will be covered with a floating cover. In addition, many miles of new large diameter trunk lines are being installed to allow greater system flexibility for the water quality projects and also as a part of a trunk line replacement program. The replacement program was initiated following a study of older trunk lines which identified several that were in need of repair and replacement. The system changes necessary for water quality improvements leave ambiguity concerning how the system may perform in future earthquake scenarios similar to what the City experienced in 1971 and Although in-city storage capacity is being significantly reduced the increased number of trunk lines being constructed will provide greater redundancy and flexibility. As a result, the Board of Water and Power Commissioners (Board) requested the water system to be evaluated for its ability to withstand natural disasters. In response to the Board s request, a dam seismic stability evaluation program is underway to evaluate the performance of 17 of the LADWP s large dams that are located south of the San Andreas Fault. The initial phase included a screening evaluation of the dams to determine their seismic stability using current knowledge of Southern California seismicity and state-of-the-practice simplified analysis methods [44]. All of these dams have previously been evaluated and determined to be seismically stable. This new program is to re-evaluate the dams based on the most recent knowledge. The screening evaluation identified 8 dams that needed additional information from field and laboratory tests and more detailed evaluations, which are currently underway and planned to be completed in a few years. During the more rigorous analysis of Stone Canyon Dam, LADWP engineers worked closely with consultants from URS Corporation to characterize the extent of the alluvium left in place under the downstream shell of the dam and to complete the first practical 3-dimentional non-linear effective-stress deformation analysis of an earth dam [45][46]. In 2002, the LADWP initiated a cooperative program with MCEER to perform an extensive research project for developing state-of-the-art system analysis modeling techniques that can account for non-linear hydraulic flow conditions to model pipe breaks in the system. The programs developed will be calibrated with known performances and actual recorded data from the Northridge Earthquake. The models are planned to be utilized for different seismic hazards throughout the Los Angeles area to estimate system performance under other earthquake scenarios, which can then be used to help identify potential system weaknesses. The LADWP is also initiating a seismic hazard evaluation program for the LAA. The Sepulveda Trunk Line (STL) is a new 84 to 96-inch (2134 to 2438 mm) diameter trunk line that was put in service in July 2003, to supply water from the VNC to many other parts of the City. As part of the system changes, the STL will become one of the most critical and important supply pipe lines in the City. The STL crosses the San Fernando Fault zone, a westerly extension of the fault that ruptured in As a result of this active fault hazard, the STL was aligned to cross through the fault zone as rapidly as possible, constructed with welded butt joints within the fault zone, isolation valves were installed outside of the fault zone, and additional interconnections were made to other nearby trunk lines. In addition, the STL was originally intended, in part, to replace the old 1913 vintage riveted steel City Trunk Line as part of the replacement program. After evaluating the seismic threat to the critical STL pipeline, plans for the City Trunk Line were altered to keep it in service and have it slip lined with a high density polyethylene (HDPE) pipe, of smaller diameter than the original riveted steel carrier pipe. This will provide additional redundancy to the STL as they both pass through the fault zone. The City Trunk Line is supported on piers through a portion of the fault zone. 183

202 The Pacific Pipeline is a privately owned pipeline that transports crude oil from Northern to Southern California for refinement. The pipeline crosses over the LAA channel within the active Santa Susana fault zone. As a part of approval to construct the new oil pipe over the LAA, several improvements were implemented to help resist seismic forces, including constructing an embankment berm around and covering the LAA channel in the pipe vicinity, routing the oil pipe above ground as it passes over the channel and placing it within a larger diameter pipe set within an even larger reinforced concrete box structure to aid in resisting strain from potential fault movement. Consultants from Woodward-Clyde and an expert engineering panel experienced with fault offset concerns for critical pipelines provided advice and direction in implementing the Pacific Pipeline improvements. As part of the water quality improvements and the numerous large diameter pipelines and appurtenant underground structures being constructed, the LADWP recently developed and implemented an improved method for evaluating lateral seismic stresses on rigid underground structures such as vaults and box conduits [47]. The method is much simpler to implement provides more realistic pressures than previous models, and is easy to implement into common underground structural design procedures. SEISMIC PREPAREDNESS This report primarily focused on capital improvements developed to help the Los Angeles water system physically resist earthquake effects. However, this is does not cover the spectra of the LADWP seismic preparedness measures. Other aspects including system maintenance and rehabilitation, emergency response and recovery plans, system management, risk management strategies, risk assessment, hazard assessment and evaluation, etc. are incorporated into the overall LADWP seismic preparedness and contribute to the historical water system seismic improvements, but description of these topics are beyond the scope of this work. A companion report on multi-hazard mitigation [48] summarizes some of these additional topics, for example emergency response and pipe rehabilitation. CONCLUSION The LADWP has a long and illustrious history of water system seismic improvements. Many examples have been described herein; however, there are many other examples that are not mentioned. The LADWP seismic improvement history has made significant impacts in the earthquake engineering industry by instrumenting some of the first structures in Southern California, developing the soil compaction methods that are most commonly used worldwide, developing and implementing seismic evaluation and construction methodologies for embankment dams, aiding the development of tank seismic design standards, and implementing innovative seismic improvement solutions. In addition, the LADWP has learned much from actual earthquake experiences and has provided significant efforts in documenting, preparing case studies, researching, and providing others with all available information in order to learn and improve the Los Angeles and other water systems from these experiences. The seismic improvements established throughout LADWP s history have proven beneficial to withstand strong earthquake shaking effects during the 1971 San Fernando and 1994 Northridge earthquakes by allowing the system to continue operating after suffering damage. Performance during these two earthquakes clearly shows the seismic resiliency developed within the City of Los Angeles Water System. The Los Angeles Water System is currently undergoing significant system wide changes to improve the quality of the water supplied to customers. In responding to this challenge the LADWP has embarked on a new seismic improvement program to re-evaluate the seismic stability of LADWP dams based on the current state of knowledge in seismology and geotechnical earthquake engineering. The LADWP has also undertaken a cooperative program 184

203 with MCEER to perform a systems analysis, which will help identify weaknesses in the supply and distribution systems that require improvement. The LADWP s goal is to continuously improve the seismic resiliency and performance capabilities of the Los Angeles Water System. ACKNOWLEDGEMENT The authors would like to thank and acknowledge the many LADWP employees who have aided in developing seismic improvements into the Los Angeles water system over the past 100 years; George W. Brodt for review of the manuscript; Victor Murillo, Kien Huang, and Roger Callo drafted figures; Phil Clark provided information on tank flexible couplings. REFERENCES [1] Outland, C. F., 2002, Man-Made Disaster the Story of St. Francis Dam, 2 nd printing of the revised edition, The Ventura County Museum of History and Art, Ventura, California. [2] International Conference of Building Officials (ICBO), 1964, Uniform Building Code, 1964 Ed., Whittier, California, May. [3] Bardet, J. P. and C. A. Davis, 1996, Engineering Observations on Ground Motion at the Van Norman Complex after the Northridge Earthquake, Bulletin of Seismological Society of America, Special Northridge Issue, Vol. 86, No. 1B, pp. S333-S349. [3a] Proctor, R. R., 1933, Fundamental Principals of Soil Compaction, Engineering News Record, Aug. 31, Sept. 7, Sept. 21, Sept. 28. [4] Nelson, S.B.S., 1941, Heavy Construction of Stormdrains to Protect San Fernando Reservoirs, Southwest Builder and Contractor, Feb. 7, 1941, pp [5] Los Angeles Department of Water and Power, 1966, Lower San Fernando Dam Slope Stability Analysis, Water Engineering Design Division Report AX [6] Seed, H. B, K. L. Lee, I. M. Idriss, and F. I. Makdisi, 1973, Analysis of the Slides in the San Fernando Dams During the Earthquake of February 9, 1971, Report No. UCB/EERC 73-2 University of California, Berkeley, California. [7] Wald, D.J., and T.H. Heaton, 1994, A Dislocation Model of the 1994 Northridge, California, Earthquake Determined from Strong Ground Motions, United States Department of the Interior, U.S. Geological Survey, Open-File Report [8] Weber, F. H., 1975, Surface Effects and Related Geology of the San Fernando Earthquake in the Sylmar Area, Chapter 6 in San Fernando, California, Earthquake of 9 February 1971, Oakeshott, G. B., Ed., California Division of Mines and Geology, Bulletin 196, Sacramento, CA. [9] Scott, R.F., 1973, The Calculation of Horizontal Accelerations from Seismoscope Records, Bulletin of the Seismological Society of America, Vol. 63, No. 5, pp [10] Cortright, C. J., 1975, Effects of the San Fernando Earthquake on the Van Norman Reservoir Complex, Chapter 29 in San Fernando, California, Earthquake of 9 February 1971, Oakeshott, G. B., Ed., California Division of Mines and Geology, Bulletin 196, Sacramento, CA. [11] Subcommittee on Water and Sewage Systems, 1973, Earthquake Damage to Water and Sewerage Facilities, San Fernando, California, Earthquake of February 9, 1971, N. A. Benfer and J. L. Coffman eds., U. S. Dept. of Commerce, NOAA Spec. Rpt., Vol. II, pp [12] O Rourke, T. D. and M. Hamada, Eds., 1992, Case Studies of Liquefaction and Lifeline Performance during Past Earthquakes, NCEER , Vol. 2, National Center for Earthquake Engineering Research, Buffalo, NY. [13] Dames and Moore, 1985, Evaluation of Earthquake-Induced Deformations of Pleasant Valley Dam, Report prepared for the City of Los Angeles Department of Water and Power, Dames and Moore Job No [14] Roth, W.H., Bureau, G., and Brodt, G., 1991, Pleasant Valley Dam An Approach to Quantifying the Effect of Foundation Liquefaction, Int. Conference, Commission of Large Dams, Vienna, Austria. 185

204 [15] Dames and Moore, 1991, Stability Evaluation South Haiwee Dam Inyo County, California, Report prepared for the City of Los Angeles Department of Water and Power, Dames and Moore Job No [16] Los Angeles Department of Water and Power, 1974, Report of Water System Vulnerability to Earthquakes, L. Lund and R. Triay, Water Engineering Design Division Report AX [17] Miedema, H.J., and J.B. Olson, 1974, Repair of Seismic Damage to Above Ground Pipelines, Transportation Engineering Journal, ASCE, 100, TE3, pp [18] Davis, C. A., 1999, Performance of a Large Diameter Trunk Line During Two Near-Field Earthquakes, Proc. 5th U.S. Conf. on Lifeline Earthquake Engr, ASCE, Seattle, Aug., pp [19] Davis, C.A., 2001, Retrofit of a Large Diameter Trunk Line Case Study of Seismic Performance, Proc. of 2 nd Japan-US Workshop on Seismic Measures for Water Supply, AWWARF/JWWA, Tokyo, Japan, Aug. 6-9, American Waterworks Association Research Foundation Project 2786, Session 2. [20] Davis, C. A. and S. R. Cole, 1999, Seismic Performance of the Second Los Angeles Aqueduct at Terminal Hill, Proc. 5th U.S. Conf. on Lifeline Earthquake Engr, ASCE, Seattle, Aug., [21] Brown, K., P. Rugar, C. Davis, and T. Rulla, 1995, Seismic Performance of Los Angeles Water Tanks, Proc. 4th U.S. Conf. on Lifeline Earthquake Engr, ASCE, San Francisco, Aug., pp [22] Toprak, S., 1998, Earthquake Effects on Buried Lifeline Systems, Dissertation presented to Cornell University in partial fulfillment for the degree of Doctor of Philosophy. [23] Lund, L., and E. Matsuda, 1994, Lifelines, Northridge Earthquake, January 17, 1994, Preliminary Reconnaissance Report, Chapt. 6, Earthquake Engineering Research Institute, pp [24] O Rourke, T. D. and M. J. O Rourke, 1995, Pipeline Response to Permanent Ground Deformation: A Benchmark Case, Proc. 4th U.S. Conf. on Lifeline Earthquake Engineering, ASCE, San Francisco, Aug., pp [25] Davis, C. A. and J. P. Bardet, 1995, Seismic Performance of Van Norman Water Lifelines, Proc. 4th U.S. Conf. on Lifeline Earthquake Engineering, ASCE, San Francisco, Aug., pp [26] Davis, C.A., and J.P. Bardet, 1996, Performance of Two Reservoirs During 1994 Northridge Earthquake, J. Geotech. Engrg. Div., ASCE, Vol. 122, No. 8, pp [27] Bardet, J.P. and C.A. Davis, 1995, Lower San Fernando Corrugated Metal Pipe Failure, Proc. 4th U.S. Conf. on Lifeline Earthquake Engineering, ASCE, San Francisco, Aug., pp [28] Davis, C.A. and J.P. Bardet, 1998, Seismic Analysis of Large Diameter Flexible Underground Pipes, J. Geotech. and GeoEnv. Engrg. Div., ASCE, Vol. 124, No. 10, pp [29] Davis, C. A., 2000, Study of Near-Source Earthquake Effects on Flexible Buried Pipes, Dissertation presented to the University of Southern California in partial fulfillment for the degree of Doctor of Philosophy. [30] Youd, T. L., 1971, Landsliding in the Vicinity of the Van Norman Lakes, in The San Fernando, California, Earthquake of February 9, 1971, U. S. Geological Survey Prof. Paper 733, pp [31] Los Angeles Department of Water and Power, 1997, Response of the High Speed and Bypass Channels to the 1994 Northridge Earthquake and Recommended Repairs, Water Supply Division Report AX [32] Davis, C.A., J.P. Bardet, and J. Hu, 2002, Effects of Ground Movements on Concrete Channels Proc. 8th US-Japan Workshop on Eq Resistant Des. of Lifeline Fac. and Countermeasures for Soil Liquefaction, Tokyo, T. D. O Rourke, J. P. Bardet, and M. Hamada, Eds., Tech. Rep. MCEER , pp [33] Archuleta, R. J., G. Mullendore, and L.F. Bonilla, 1998, Separating the Variability of Ground Motion over Small Distances, Proceedings from The Effects of Surface Geologyon Seismic Motion, Balkema, Rotterdam. [34] Trifunac, M. D., M. I. Todorovska, and V. W. Lee, 1998, The Rinaldi Strong Ground Motion Accelerogram of the Northridge, California Earthquake of 17 January 1994, Earthquake Spectra, EERI, Vol. 14, No. 1, pp [35] Cultrera, G., D. M. Boore, W. B. Joyner, C. M. Dietel, 1999, Nonlinear Soil Response in the Vicinity of the Van Norman Complex Following the 1994 Northridge, California, Earthquake, Bulletin of the Seismological Society of America, Vol. 89, No. 5, pp [36] Sano, Y., T. D. O Rourke, and M. Hamada, 2002, Permenant Ground Deformation due to Northridge Earthquake in the Vicinity of the Van Norman Complex, Proc. 7th US-Japan Workshop 186

205 on Earthquake Resistant Des. of Lifeline Fac. and Countermeasures for Soil Liquefaction, Seattle, T. D. O Rourke, J. P. Bardet, and M. Hamada, Eds., Tech. Rep. MCEER [37] Bardet, J.P., J. Hu, T. Tobita, and N. Mace, 2002, Large-Scale Modeling of Liquefaction-induced Ground Deformation, Earthquake Spectra, Vol. 18, No. 1, pp [38] Toprak, S., O Rourke, T.D., and Tutuncu, I., 1999, GIS Characterization Of Spatially Distributed Lifeline Damage," Proceedings, 5th US Conference on Lifeline Earthquake Engr, Seattle, WA, [39] Bardet, J.P., J. Hu, T. Tobita, and N. Mace, 1999, Database of Case Histories on liquefactioninduced Ground Deformation, A report to PEER/PG&E, Task 4 Phase 2, October, Civil Engineering Department, University of Southern California, Los Angeles, CA. [40] American Lifelines Alliance, 2001, Seismic Fragility Formulations for Water Systems, John Eidenger Principal Investigator, [41] O Rourke, T.D., Y. Wang, P. Shi, and S. Jones, 2004, Seismic Wave Effects on Water Trunk and Transmission Lines, 11 th International Conference on Soil Dynamics & Earthquake Engineering, Berkeley, CA, Jan. 7~9, Submitted. [42] Tutuncu, I., T.D. O'Rourke, J.A. Mason, and T.K. Bond, "Seismic Rehabilitation of Water Trunk Lines with Fiber Reinforced Composite Wraps", Proceedings, 7 th National Conference on Earthquake Engineering, Boston, MA, July, 2002, EERI, Oakland, CA.. [43] O Rourke, T. D., 2003, Professor, Personal communication, School of Civil and Environmental Engineering, Cornell University, Ithaca, N.Y. [44] URS Greiner Woodward Clyde, 2000, Review of Seismic Stability of Seventeen Water Service Organization Dams, Prepared for the Los Angeles Department of Water and Power. [45] URS, 2002, Seismic Stability Evaluation of Stone Canyon Dam, Report prepared for the City of Los Angeles Department of Water and Power, URS Project No , Water Services Organization Report No. AX [46] Roth,W. H., E. M. Dawson, C. A. Davis, and C. C. Plumb, 2004, Evaluating the Seismic Performance of Stone Canyon Dam with 2-D and 3-D Analyses, 13th World Conference on Earthquake Engineering, August 1-6, Vancouver, BC, Canada, in press. [47] Davis, C. A., 2003, Lateral Seismic Pressures for Design of Rigid Underground Structures, Proc. 6th U.S. Conf. on Lifeline Earthquake Engineering, ASCE, Long Beach, Aug., 10-13, pp [48] Lund, L., and C. Davis, 2004, Multihazard Mitigation Los Angeles Water System A Historical Perspective, ASCE Technical Council on Lifeline Earthquake Engineering, Multihazard Monograph, Craig Taylor editor, in preparation. Presentation at ASCE TCLEE Workshop, 6 th US Conference on Lifeline Earthquake Engineering, Long Beach, CA, August 10,

206 188

207 3 rd US-Japan Workshop on Water System Seismic Practices TECNICAL SESSION V System Performance and Management Lessons Learned from the World Trade Center Disaster for Water Supply Management Presenter: Thomas D. O Rourke (Cornell University) 189

208 3 rd US-Japan Workshop on Water System Seismic Practices 190

209 LESSONS LEARNED FROM THE WORLD TRADE CENTER DISASTER FOR WATER SUPPLY MANAGEMENT T.D. O Rourke Cornell University ABSTRACT The events of September 11, 2001 have changed profoundly our approach to critical civil infrastructure, including the engineering and management of water supplies. Professor O Rourke will examine the performance of the water distribution system in New York City during the World Trade Center Disaster (WTC), covering damage sustained by the system, collateral effects of that damage, and impact of water supply disruption on fire fighting surrounding the WTC Complex. He will compare and contrast water supply performance during September 11 with water supply performance during earthquakes, making specific comparisons with water distribution and fire fighting during the 1906 San Francisco and 1989 Loma Prieta earthquakes. Lessons learned from water supply performance during extreme events, including earthquakes, severe accidents, and acts of violence, will be summarized. The impact of September 11 on future policies and procedures for critical infrastructure management will be discussed. The paper documenting the information presented by Professor O Rourke has previously been published in Beyond September 11: An Account of Post-disaster Research. Special Publication #39. Boulder, CO: Natural Hazards Research and Applications Information Center, University of Colorado 1. A copy of the manuscript from this publication was provided as a part of the preliminary proceedings handed out at the workshop for use by each of the workshop participants. The manuscript is not reprinted herein. Please refer to the above reference for the paper presented by Professor O Rourke. 1 O Rourke, T. D., A. J. Lembo, L. K. Nozick, Lessons Learned from the World Trade Center Disaster about Critical Utility Systems, Beyond September 11: An Account of Post-disaster Research. Special Publication #39. Boulder, CO: Natural Hazards Research and Applications Information Center, University of Colorado. 191

210 192

211 3 rd US-Japan Workshop on Water System Seismic Practices TECNICAL SESSION VI Risk Assessment and Analysis II Maintenance Management System for Lowering Possible Seismic Damages onto Water Works Facilities Presenter: Jiin-Song Tsai (National Cheng Kung University, Tainan, Taiwan) An Overview of the Metropolitan Water District of Southern California s Seismic Program Presenter: Clark Sandberg (Metropolitan Water District of Southern California, Los Angeles) A Fast Simulation Method for Predicting Seismic Responses of an Extensive Water Distribution Network Presenter: Nobuhisa Suzuki (Japan Water Steel Pipe Association, Kawasaki, Japan) Multi-Hazard Risk Assessments, Elements in Common with Seismic, Security, and Other Risk Studies Presenter: Don Ballantyne (ABS, Seattle, Washington) A Study on the Development of a Backup System in a Big Urban Area Presneter: Yoshiharu Sorakuma (Japan Water Research Center, Tokyo, Japan) 193

212 3 rd US-Japan Workshop on Water System Seismic Practices 194

213 Maintenance Management System for Lowering Possible Seismic Damages onto Water Works Facilities Jiin-Song Tsai and Cheryl S.F. Chi ABSTRACT This paper presents a thought using maintenance methodologies to keep ensuring seismic safety with respect to regularly updated seismic resistant requirements against future events, especially for newly restored plant such as Fongyuan water treatment plant. It was serious damaged in the 1999 Ji-Ji earthquake and resulted in a serious consequence of water shortage to millions people in Taichung municipal area for a month. Previous maintenance considered rapid repair service the upmost desire. This is because the repair activities were used to be under requests when damages or failures occurred. To improve the maintenance operation that passively subjected to occasionally damage events to actively regular activities for seismic safety is thus of interest. In this paper, application of both RBI (Risk Based Inspection) and CMMS (Computer Maintenance Management System) to the maintenance of a water treatment plant is described. Both RBI and CMMS technology have been adopted to guard the safety of another risk threatened business, the petroleum and petrochemical industry, in which RBI identifies the critical facilities in a process while CMMS makes the maintenance works simple and cost efficient. To Fongyuan plant, identification of the concerns and contents of the maintenance for seismic safety is then the important works for developing a maintenance system that fulfills the desire of lowering damages of water works facilities under future seismic strikes. Jiin-Song Tsai, Professor, Department of Civil Engineering, National Cheng Kung University, Tainan, 701, Taiwan, R.O.C. Cheryl S. F. Chi, Research Associate, Department of Civil Engineering, National Cheng Kung University, Tainan, 701, Taiwan, R.O.C. 195

214 INTRODUCTION Taiwan is an island located at a complex juncture between the Eurasian and Philippine Sea plate, where occupies an unstable region between two subduction systems of opposite polarity. In this region, 23 millions peoples resident on a land of 36 thousands square kilometers area, in which 42 active faults are embedded. It is the nature that seismicity is extremely active. On September 21, 1999, at 1:47 a.m. local time, Ji-Ji earthquake (ML=7.3) stroke the central part of Taiwan. The casualties are 2,161 losses, 8736 injuries and $3.7 billion property loss [1]. This is the most costly natural disaster and the strongest earthquake that Taiwan has ever experienced in the past 100 years. Close to the epicenter, two near-by main facilities of water works, Shi-Gun Dam and Fongyuan water treatment plant (WTP), were serious damaged and the result-in consequence afflicted 700 thousands families in Taichung municipal area suffering water shortage for a month. This case showed the seismic disastrous would send ripples through societies and economies besides its direct harm. After the earthquake, three needs were identified for the safety of water supply systems against future seismic strikes are: (1) seismic resistant strength enhancement of water supply system, (2) rapid repair measures for emergent water supply systems under both regular and emergency cases, and (3) new water resources with good management to lower the possible risk of water shortage. To enhance the seismic resistance of water supply system, seismic design criteria or codes are either regularly revised or occasionally updated after major events. This implies that no existing design and retrofit skill shall guarantee a perfect seismic proof result. Hence a consistent maintenance plan in addition to proper design and construction is needed to have a water supply system safe from possible major events, in which a risk management system including evaluation of seismic risk to identify critical equipments is crucial. It is worthy noted that maintenance plan regarding seismic safety is to lengthen the service life of equipment, and the purpose of the evaluation is to properly invest limit resources to critical equipments of large consequence. This paper describes the concept of developing a maintenance system to prevent seismic risk threatening water supply industry, in which Fongyuan WTP that has been newly recovered from Ji-Ji earthquake is first chosen to study an efficient maintenance system to safeguard those newly installed facilities against future seismic risk. To water treatment plant, seismic threaten is unpredictable risk, while to a petrochemical plant, failure of pressure equipment in a process unit resulting in large lost is risk of similar undesirable. The difference is that the concepts of risk analysis and management have been around in petroleum and petrochemical industry for a long time. It is encouraging that the public and the media are beginning to understand and appreciate that water resource is getting precious nowadays and just as petroleum in last century. Experience of the application of the Risk-Based Inspection (RBI) technology and the Computer Maintenance Management System (CMMS) available in petroleum and petrochemical industry should be worthy and adoptable to water supply industry now. In the following, description first covers the fundamental concepts RBI and CMMS, and then the concept of their application onto Fongyuan WTP. Future goal is to introduce RBI and CMMS skill to develop a maintenance management system to water treatment plant to resolve the problem of continuously updating seismic resistant requirements. DAMAGES OF FENGYUAN WTP DURING JI-JI EARTHQUAKE As shown in Figure 1, Fengyuan water treatment plant was built around 1930 and has gradually grew to meet the increasingly water needs since then. It consists of two sub-units, named Fengyuan 1st plant and Fengyuan 2nd plant, and is located in the north of Fengyuan 196

215 County close to Chelungpu fault, the major active fault causing Ji-Ji earthquake. Taking untreated water from nearby Shi-Gun Danm, Fengyuan WTP serves the duty to supply clean water to the Taichung municipal area in mid-west of Taiwan. Fengyuan WTP is capable to serve 1.3 MCMD (Mega tons per day) or 337 MGD (million gallons per day), in which Fengyuan 1st plant produces about 0.4 MCMD and Fengyuan 2nd plant produces about 0.9 MCMD. Shi Gun Dam Fungyuan WTP Fontyuan County Taichung City Miaoli County Taichung County Tamoupu-Hsuangtung fault Chang uua County Nantou County Chelungpu fault Ji-Ji earthquake epicenter (M L =7.3, 9/21/1999) Yunlin County N W E Figure 1. Location of Chelungpu fault and epicenter of Ji-Ji earthquake S Chelungpu fault has been recognized as the main trigger of the Ji-Ji earthquake. After the quake, the resulting fault displacement formation on the ground surface is believed to be the most conspicuous phenomenon to geologists and engineers. The enormous fault length (total length about 105 km) and the tremendous thrust fracture (the largest offset around 4-10 meters in different locations) of the steep reverse fault have impressed on humans the power of nature. The induced offset deformation of the opposite sides of the faults tore up all kinds of human-made infrastructure, and of course breakage of the water supply facilities crossing the faulting fault was also inevitable. For example, Shi-Gun Dam broken by a large fault offset of 9.8 m in vertical direction and an embedded large steel pipe (2000 mmφ ) tore up by fault movement are both worldwide impressive. Fengyuan WTP is in close proximity to Shi-Gun Dam, and is about 160 meters away to the ground rupture of Chelungpu fault (Figure 2). The principal damage of Fengyuan 1st plant was associated with opening and pounding of expansion joints and collapse of the reservoir roof and wall, and leakage was as the result in consequence. Other damages were: 1. Flocculation baffles in flocculation basins destructed totally. 2. Incline plates installed in sedimentation basins collapsed and turned out entirely breakdown of the equipment. 3. Pipes attached to rapid filtration basins damaged, and most intake pipes to slow filtration were broken. 4. Pipe connection joints to the administration building are serious damaged. 5. Damages of administration building were mainly associated with cracks and fractures of wall, and the installed electrical facilities fell and damaged. 6. Structures associated to purification process damaged by different settlement and 197

216 movement of surrounding ground. 7. Water treatment facilities, including pump suction wells, inlet pipes, settling basins, conduits, channels and clarifier tanks are damaged and resulted in a complete shut down of the entire system. Shi Gun Dam Chelungpu fault Fongyuan 2nd WTP Fongyuan 1st WTP Figure 2. Chelungpu fault and Fengyuan WTP In Ji-Ji earthquake, damage to Fengyuan 2 nd plant suffered relatively less damages. It was quickly recovered by emergent repair, but water treatment process could carry on with a lower capacity mainly owing to fracture of reservoir s wall. In addition to the plant sustained only about one-half regular capacity of water treatment ability, pipelines damages, such as the main conduit (a 2000 mmφ steel pipe) to supply water to Taichung metropolitan was broken by the Chelungpu fault, resulted in a very serious consequence of water shortage to millions people in Taichung for a month. One more case of major pipeline failure was along the same fault and about 10 km south of Shigarng Dam, which was a primary water supply pipeline bridge (1000 mmφ ) passing by Ijiang Bridge at Toubianken. The pipeline bridge was also broken due to immense faulting movement and worse the water shortage condition [2][3]. MAINTENANCE OF FACILITIES IN PETROCHEMICAL INDUSTRY Fire and explosion surely threaten petrochemical plants, and the consequence would result enormous lost of finance and human lives. During the last three decades, efforts have been put to upgrade the skills to safeguard the production process. Despite the production process have been promoted in several ways and are very different from the original operation in a housekeeping way, M&M Protection Consultants has investigated 170 major losses in petrochemical industry and indicate that the losses seems irrelevant to those considerable upgrades [4]. In fact, Figure 3 shows that the frequency and cost of those 170 losses have increased significantly over the last 30 years. Frequency Frequency Cost Billions of dollars Figure 3. Large property damage losses in the hydrocarbon-chemical industries [4] 198

217 It had been believed for long that updating and standardizing design criteria regarding all details in production process should ease the risk to a certain level. In the past three decades, development of technology to safeguard petrochemical plants was mainly on passive ways of disposal, mitigation, fire protection and relief. Until the fact revealed that major events had hardly been eased, people started to look the fundamentals of the problem: (1) The activities of petrochemical are mostly in a mass production way that operates in a sequential process, and problem to any step within the process would result in a chain-effect. (2) Careless on handling hazardous materials would cause flammable events, toxic releases. Tremendous losses on business interruption and major environmental damage to surrounding are then inevitable. (3) Despite the fact that existing obsolete operation might cause dangers and induce serious consequence, some earlier installed facilities serve with other newer models in one production process. (4) Optimization and standardization should not be applied solely onto individual step, since safe of individual does not guarantee the safety of entire process. Integration of system is the issue of more importance. The above understandings give a clear picture that the safety against risky hazards does not rely on additional protection shelters, mitigation systems and other solid equipment. Furthermore, it is obviously not practical to invest unlimited sources to eliminate the risk of surely occurrence events without exact time. On the other hand, prioritizing our concerns with a systematic process that exposes real risk would make a better sense. Wisdom of previous experiences indicates that maintenance with inspection and renew/repair on the basis of evaluation using risk based analysis (RBI) shall offer a closer solution [5, 6, 7, 8]. Rasche and Woolley [9] simply defined risk as the combination of likelihood of a failure occurring, or hazard arising, and showed the consequence of failure were it to occur as shown in Figure 4. Aller et al. [6] mentioned that Risk Based Inspection (RBI) is a method for using risk as a basis for prioritizing and managing the efforts of an inspection program, and summarized the purpose of a risk based inspection as follows: (1) Screen operating unit within a plant to identify areas of higher risk (2) Calculate a risk value associated with the operation of each equipment item in a process plant based on a consistent methodology (3) Prioritize the equipment based on the calculated risk (4) Design an appropriate inspection program to lower total risk (5) Systematically manage the risk of equipment failures Likelihood Consequence Risk Likelihood is a prediction of the probaility that a given item will fall. Consequence is a measure of the effect of a give failure occuring. This may be a measure of environmental hazard, safety hazard or economic hazard. Risk = Lieklihood Consequence Figure 4. Components of risk [9] Risk based inspection methodology is both qualitative and quantitative. The qualitative risk assessment allows for a relatively, quick overview to rank plant equipment on the basis 199

218 of likelihood of failure and consequences of failure which are then presented in a risk matrix to identify clearly the priorities (Figure 5). Judgment and experience are the main basis of this type of assessment. This means that to a greater degree, the results are dependent on the knowledge and experience of the engineers carrying out the analysis. The quantitative approach involves detailed assessments of the likelihood and consequences of failure of each plant item. Likelihood of failure is evaluated from generic equipment failure frequency data when available, modified to take account of equipment age, condition, complexity, process conditions, modes of degradation and rates of deterioration etc. The consequence analysis calculates loss of containment effects based on an established set of scenarios. Aller et al. [6] showed an overview of the quantitative RBI approach as Figure 6. The risk for each equipment item is then found by summing the individual risk components from each scenario calculation. LIKELIHOOD CATEGORY Medium-High Risk Medium Risk Lower Risk Higher Risk PSM/PHA ASSESSMENTS EQUIPMENT FAILURE DATA BASE PROCESS CONDITIONS PROBABILITY OF FAILURE ANAALYSIS MANAGEMENT SYSTEMS EQYUONEBT EVALUATION EQUUIPMENT DATA FILE RISK PRIORITIZATION CONSEQUENCE ANALYSIS PROCESS SAFETY MANAGEMENT IMPROVEMENTS INSPECTION PROGRAM A B C D E CONSEQUENCE INSPECTION PROGRAM MITIGATION IMPROVEMENTS CONSEQUENCY CATEGORY Figure 5. Qualitative RBI matrix [6] Figure 6. Application of Risk-Based Inspection [6] Risk levels are prioritized in a systematic manner so that an inspection program can be planned that focuses more resources on higher risk equipment while possibly saving inspection resources that are not doing an effective job of reducing risk. It is thus anticipated that an effective risk based program will result in a reduced level of risk for a given level of inspection activity [6, 7]. It is worthy noting that RBI concept have been increasingly applied to structure targeted inspection programs in the petrochemical industry during the last decade [8]. To effectively maintain production facilities on the basis of risk evaluation using RBI, CMMS has been adopted by petrochemical industry in Taiwan to manage the inspection and corresponding repair/renewal works. With clear data of equipment operation information, the maintenance activities can be greatly simplified using computer to optimize an efficient scheme. Considering aging effect to all installed equipment, life cycle cost is the issue to evaluate the existing value of a petrochemical plant. The operation records of all installed equipment are thus essential to the evaluation regarding its safety, cost effects and extension of service life. CMMS can be used to systemically manage all equipment data from very beginning of installation till the end. Computers are surely the tools to keep and organize all kinds of equipment data including fundamental background information, plan for regular inspection and maintenance, work request and schedule, and control of spare parts inventory. Using computer management, purchase and test to new equipment, materials and parts equipment can be proceed systemically. Procedure and benefits of adopting CMMS show as Figures 7 and 8, respectively. Travis [10] and Weil [11] summarized benefits of using CMMS: (1) increase availability of equipment, (2) increase productivity of staff, (3) increase efficiency of 200

219 maintenance, (4) lower cost of inventory, (5) lower downtime and (6) gain life prolongation of equipment. The purpose of coupling RBI and CMMS is to effectively longer the life cycle duration of facilities. Conceptually, risk-based inspection is a methodology/standard, which prioritizes risk levels in a systematic manner so that the owner-user can then plan an inspection program that focuses more resources on the higher risk equipment. Meanwhile, CMMS can actively automatically process RBI strategy, which changes the concept of maintenance from repair-after-breakdown to preventive maintenance. RBI together with CMMS, when properly applied, can reduce overall costs and increase the efficiency and longevity of facilities. Facility information Decision making studies of failure cases Technical information cost of facilities Maintenance works up time analysis jbg requests job estimation and approval job planning operation control record of failure cases benefits to operation benefits to maintenance Parts/Inventory Contract management Workers data optimization of maintenance Figure 7. Procedures of Computer Maintenance Management System facility information inspection and repair accessible data base safety reduction of events early stage diagnosis and care decision making quality reduction of harmful factors reduction of parts inventory and repair cost cost increase stability of system gain reliable facility productivity Figure 8. Benefits of using computer maintenance management system SEISMIC RISK STUDIES OF WTP In this study, causes of seismic hazards or damages to a water treatment plant are different from those causing flammable and explosive events to a petrochemical plant. Most causes to petrochemical plant are internally raised and mainly in two categories: mechanical breakdown (38%) and improper operation (26%) [8]. All causes generally generate their consequences in a three-step sequence that begins with initiation, propagation and then end up with final consequence. Once any cause accidentally occurs, a chain-effect would proceed with no interruption until the final consequences shown. To a water treatment plant, the only cause leading to seismic hazards is active fault fracture outside the plant. Adopting Fengyuan WTP as an example, Chelungpu fault suddenly ruptures initiate a seismic event. Damage of Fengyuan WTP and the pipeline systems surrounding enlarge the influences of the designated Ji-Ji earthquake, and the influences propagated until shortage of water for daily and industrial use, fire extinguishments and disease prevention use. Threaten to the community with problems of recovery works, economical activities, and public health are then the final consequence. 201

220 As mentioned in above, mechanical problems inside a petrochemical plant are the cause of damages of explosions, flammable events or/and others, while irresistible fault movement outside a water treatment plant play the main role as the cause of seismic event which damages the installed facilities. From the point of view of maintenance, regular inspection to keep facility in good function is the efficient strategy to lower the possibility of occurrence of mechanical problems. Following the similar strategy, regular inspection to keep or/and upgrade the required strength of facility to fulfill the seismic design codes is proper to survive safely from major earthquake strikes. It is worthy noting that some activities, such as strengthen all structures, plan water supply pipeline with backup systems and have all in-time emergency repair measures in good condition, are important to a water treatment plant to face the challenge of possible seismic strike and should be well prepared all the time with regular inspection and test. To mitigate the hazards of possible major earthquakes in Taiwan, Water Work Associate of Republic of China (TWWA) has issued the first version of guideline and illustration for seismic design and construction of water work facilities in 2002, and the guideline is subjected to update in every three years. In the first version, importance of maintenance has been mentioned in addition to the conventional concept of strengthening or retrofitting. This study intends to forward the thoughts to preventive maintenance regarding seismic risk, and the results shall be useful for the guideline updating. Lee et al. [12] gave an example regarding the Seismic Improvement Program (SIP) implemented after the 1989 Loma Prieta Earthquake, in which the East Bay Municipal Utility District (EBMUD) invested $200 millions in 10 years to increase the reliability of its water system. Two important tasks in the SIP were: (1) evaluation of the entire water treatment system to identify the critical components and structures essential to the continued operation of the facilities and (2) upgrade design code of water treatment plant. The authors appreciate good works from this a marvelous case, especially assessment of the evaluation based on a risk and vulnerability analysis to determine which elements would require improvements to ensure satisfactory performance. These excellent works lead to a thought of preventive maintenance regarding to seismic risk, in which the evaluation of risk analysis is not only to identify the critical components with great consequence to the entire water supply system but the likelihooh of an event occurrence. Researches of engineering aspect on enhancing seismic resistance used to be in two main categories. The first category covers works before earthquake, such as seismic risk analysis, seismic mitigation planning and retrofit of existing infrastructures [13][14]. The second category covers works after earthquake, such as rapid report system, rapid repair skills and measures to ease emergency conditions. To rich the research contributions, the authors believe studies that bridges these two categories are valuable. As shown in Figure 9, maintenance works to keep updating/upgrading the seismic strength required during life cycle period are of interest. SEISMIC WORK BEFORE EARTHQUAKE OPERATION AFTER EARTHQUAKE seismic risk analysis seismic mitigation planning retrofit of existing infrastructures rapid repair skills rapid report system emergency measures Figure 9. Seismic works 202

221 Previous experiences [5,6,7,8] regarding risk analysis and maintenance management available in petrochemical industry indicates that the importance of equipment and its consequence in a possible process event should be first evaluated by risk analysis. Based on assessment of the evaluation, measures with proper activities, such as regular inspection, are then executed. In addition, management information system using computer is introduced to control the activities effectively by integrating all necessary information including function index of facility, data of inspection and analysis results. Cases using the practice mentioned above showed that the feasibility of major events is considerably lowered [15]. SEISMIC MAINTENANCE MANAGEMENT OF FENGYUAN WTP The business goal of Fengyuan WTP is to supply water in good quality for long. To meet the goal, threaten from the near by Chelungpu fault must be serious cared. Since the seismic factor affects the water treatment facilities, retention of enough seismic resistance ability is then important to keep or longer the designated service life of Fengyuan WTP. The life cycle of WPT begins with planning, design, contracting, construction until revoking, and then demolishing (Figure 10). Capacity of the WTP is founded during the period from planning till construction completion, and the following maintenance during operation is to keep the same capacity lasting. Yanev (2001) [16] gave a picture showing the variation of maintenance cost with time as Figure 11, while Frangopol et al. [17] indicated the trade-off function of facility regarding the inevitable decay of servibility (Figure 12). Previous studies [18,19, 20] showed Figure 13 as a simple guideline to find the optimistic life extension for aging facilities. To Fengyuan WTP, seismic factor considerably influence the maintenance cost, servibility, and optimistic life extension of all installed facilities. It is believed that maintenance management is equivalent important to other considerations for strengthening seismic resistance. Operational maintenance Plan Design Contract Construction Performance initial performance beginning debilitating appearing debilitating accelerating debiltating Rebuild Demolish regular maintenance restore and strenghten Time Figure 10. Life cycle of water treatment plant performance Cost maintenance cost initial investment Servibility minimum requirement of performance (suitability, stability, activity, endurance) Time (year) Figure 11. Maintenance cost versus time Time Figure 12. Performance versus time 203

222 Cost performance maintenance cost minimum requirement of performance initial investment Servibility match point Time Figure 13. Optimum performance age to facility The 1999 Ji-Ji earthquake (ML=7.3) is now taken as the scenario event for future operation of Fengyuan WTP regarding seismic safety requirements. To find a proper investment strategy covering the cost of maintenance, RBI together with CMMS is the combination to fulfill the seismic safety requirements. Oshumi and Yamamoto [21] adopted RBI on the study of seismic retrofit planning for sewage facilities. Ohtsu [14] performed cost-benefit-analysis using the similar methodology to evaluate various countermeasures for problems of seismic slope stability. It is worthy noting that most RBI applications emphasized on: (1) prioritizing risk levels of facilities and focusing proper resources on those of high risk level during the planning stage for seismic safety, and (2) investing enough resources to maintain those high risk facilities during the stage of operation. It is known that RBI application has to cope with many detail determinations including (1) determination of risk levels, (2) determination of damage levels and consequence levels in either qualitative or quantitative ways, and (3) the analyses of result-in cost effects and consequence effects. In real practice, CMMS is introduced to implement RBI works and put those works in a regular schedule. Applying CMMS for seismic safety, two important questions are raised before its application: (1) what facilities should be subjected to regular inspection? (2) how often is the inspection for each facility? All these questions and the detail determinations mentioned in above require more contribution of lasting efforts. SUMMARY Previous experiences have shown that no matter how good the earthquake resistant design, people has recognized that no guarantee is able to be assured regarding absolutely safe against seismic threaten. In real practices, it has been known as well that the most important thing is the planning of effective measures and enhancing the necessity of long-term projects from the viewpoint of risk management. To better safeguard the seismic safety of WTP, the authors regard the management of inspection and maintenance as a new research subject. In the present stage, experiences available in petrochemical industry using RBI and CMMS against unpredictable explosion and flammable events are worthwhile for WTP to face inevitable seismic events. In an ongoing research project, Fengyuan WTP that recovered from serious damage in the 1999 Ji-Ji earthquake is first chosen to practice RBI and CMMS methodology to safeguard those newly installed facilities against future seismic risk. It is believed that a long-term preventive maintenance regarding updating seismic requirements is useful, especially for judging proper resources to deal with the immense risk. 204

223 REFERENCES [1]. National Science Foundation (NSF/ROC) Summarized Report on Damaged Bridges de to Chi-Chi Earthquake, 1999/9/2, Taiwan, the Republic of China, December [2]. Hsu, Y.T. and C.C. Fu Study of Damaged Wushi Bridge in Taiwan Earthquake, Practice Periodical on Structural Design and Construction, ASCE Reston VA USA, 5(4), pp [3]. Yeh, C.S Emergency Water Supply Measures in Taichung Area After 921Earthquake, Proceedings of City Disaster Precaution and Relief Conference, Taipei, pp [4]. Marsh and McLennan Large Property Damage Losses in the Hydrocarbon-Chemical Industries. A Thirty Year Review, M&M Protection Consultants, 14th Edition, Mahoney. [5]. Chang, I. C Chemical Process Safety Management, Yang-Chih Book Co., Taipei. [6]. Aller, J.E., Reynolds, J.T., Horowitz, N.C. and B.J. Weber Risk Based Inspection for The Petrochemical Industry, Risk and Safety Assessments: Where Is the Balance? Proceedings of the 1995 ASME/JSME Pressure Vessels and Piping Conference, v 296, pp , Jul 23-27, [7]. Renolds, J.T The Application of Risk-Based Inspection Methodology in The Petroleum and Petrochemical Industry, Risk and Material Performance for Petroleum, Process and Power Proceedings of the 1996 ASME Pressure Vessels and Piping Conference, v 336, pp , Jul 21-26, [8]. Renolds, J.T Api Methodology for Risk-Based Inspection (RBI) Analysis for the Petroleum and Petrochemical Industry, Codes and Standards Proceedings of the 1998 ASME/JSME Joint Pressure Vessels and Piping Conference, v 360, pp , Jul 26-30, [9]. Rasche, T. and K. Woolley Importance of Risk Based Integrity Management in Your Safety Management System-Advanced Methodologies and Practical Examples, Presented at 2000 Queensland Mining Industry Health and Safety Conference, August 27-30, [10]. Travis, D CMMS : Five Steps to Ensure Workability, Process Safety Progress, 15(2) pp [11]. Weil, M Raising the Bar for Maintenance Apps, Manufacturing Systems, 16(11). [12]. Lee, D., Mo, B. and T. Zimmerman Water Treatment Plant Seismic Upgrades, Water Supply Anti-Seismic Measures on Water Supply, 18(3), pp [13]. Federal Emergency Management Agency Direct Damage to Lifelines - Utility Systems, Earthquake Loss Estimation Methodology: HAZUS 99 Technical Manual. [14]. Ohtsu, H., Ohnishi, Y., Mizutani, M. and M. Itou The Proposal of The Methodology Associated with Decision-Making of Reinforcement of Slopes Based on Cost-Benefit-Analysis, Journal of Environmental Systems and Engineering, Japan Society of Civil Engineers, 51(679), pp (in Japanese) [15]. Holland, M.L Cost savings achievable through application of risk based inspection philosophies, Risk, Economy and Safety, Failure Minimisation and Analysis, Balkema, Rofterdam, c1998. [16]. Yanev, B.S Bridge Maintenance Life Cycle Cost Assessment. [17]. Frangopol, D.M., Kong, J.S. and E.S. Gharaibeh Reliability-Based Life-Cycle Management of Highway Bridges, Journal of Computing in Civil Engineering, 15(1), pp [18]. Sire, R.A. and S.W. Hopkins Analytical Modeling for Life Extension of Aging Equipment, International Journal of Fatigue, v 19, pp. S261-S266. [19]. Boyd-Lee, A.D., Harrison, G.F. and M.B. Henderson Evaluation of Standard Life Assessment Procedures and Life Extension Methodologies for Fracture-Critical Components, International Journal Fatigue, v 23, pp. S11-S19. [20]. Christian, J. and A. Pandeya Cost Predictions of Facilities, Journal of Management in Engineering, pp [21]. Ohsumi, T. and Yamamoto, K Study on seismic retrofit planning method for sewerage facilities on the basis of seismic risk management, Public Report, v 10, pp

224 206

225 An Overview of The Metropolitan Water District of Southern California s Seismic Program Clark Sandberg S.E. and Ray DeWinter ABSTRACT The Metropolitan Water District of Southern California (Metropolitan) is one of the world s largest water purveyors, providing supplemental water to nearly 18 million residents along Southern California s coastal plain. As the primary importer of supplemental water, Metropolitan faces an interesting dichotomy in providing adequate, reliable and high-quality water within one of the world s most active seismic regions. As such, this paper will address Metropolitan s seismic program, in particular its Evaluation and Seismic Strengthening Program, performance of Periodic Earthquake inspections, and Emergency Response procedures. 207

226 Earthquakes are a major concern in Southern California, and an interruption in a key community service such as water delivery can be devastating to a community s recovery after an earthquake. In 1971 the Metropolitan Water District of Southern California (Metropolitan) had a wake-up call. The magnitude 6.5 San Fernando Earthquake caused about $3,000,000 damage to its Jensen Water Filtration Plant, which was under construction at the time (see Figure 1). Figure 1 Damage to Jensen Filtration Plant while under construction in 1971 The likelihood of a large earthquake in Southern California is high. The Southern California Earthquake Data Center has analyzed the probability of large earthquakes that may affect Southern California and their analysis predicts that the region should experience a magnitude 7.0 or greater earthquake about seven times each century. About half of these will be on the San Andreas "system" (i.e., the San Andreas, San Jacinto, Imperial, and Whittier-Elsinore faults see Figure 2) and half will be on other faults. 1 For a chronology of major earthquakes in Southern California over the past century see Table

227 209 Figure 2 Southern California Earthquake Faults

228 Table 1 Chronology of Major Earthquakes in Southern California Location Date Magnitude Imperial Valley February Long Beach March Kern County July San Fernando February Northridge January Hector Mines October As a public agency that supplies southern California with over 60% of its drinking water, Metropolitan is determined to minimize the impact of these expected earthquakes on its employees and the people and businesses within its service area. Metropolitan has seismically upgraded 42 of its more than 320 facilities and efforts have been made to brace or anchor all of the essential or critical non-structural systems such as equipment and piping within all of its facilities. The 42 upgraded facilities are a result of evaluations of 158, or approximately half, of the total facilities. A program is underway to seismically evaluate the remaining facilities. General information regarding the Seismic Program is provided below. General Information Introduction The Seismic Program focuses on five areas: 1. Evaluation of all facilities 2. Study of facilities with seismic deficiencies 3. Design of seismic upgrades for facilities with deficiencies 4. Periodic inspection of all facilities to verify that all essential non-structural systems and equipment are anchored or braced 5. Organization and training of damage assessment teams for use after an earthquake 210

229 Program Elements and Description Seismic evaluation phase The key element of the Seismic Program is the seismic evaluation. This is a rapid evaluation of the lateral load resisting elements of each facility that is performed to determine if there are any structural deficiencies that need to be corrected by strengthening. Each evaluation takes between two and ten days, depending upon the facility or structure being evaluated. In order to evaluate the most critical facilities first, the more than 300 facilities have been prioritized based on: Service area reliance on facility for water Importance to Metropolitan operations Earthquake fault proximity Population served Once an evaluation has been completed, a report is prepared that describes the evaluation, summarizes the results, and dispositions the facility or structure as either OK As-Is or Needs Strengthening. No additional action is performed on approved as-is structures, and facilities or structures that need strengthening move into the next phase of the program. Study phase Once a deficiency is noted during the seismic evaluation phase, management authorization is obtained to perform a study to determine how the deficiency might be corrected. The study phase involves a more in-depth review of the deficient lateral load resisting system, sufficient to determine what corrective action needs to be taken. The analysis methodology and results are then summarized and used as the basis for a recommendation, which includes: The proposed seismic strengthening system A rough estimate of construction costs. Design phase The recommendation prepared in the study phase will then be used to obtain project approval from Metropolitan s Board of Directors to do preliminary and final design for a seismic upgrade and to issue a bid for construction. 211

230 Periodic inspections of non-structural items Periodic inspections of the non-structural items at the facilities are made to verify that essential systems and equipment are anchored or braced. These are intended to document and evaluate any nonstructural changes to the facility since the last inspection. These changes might be design modifications or equipment additions or changes that could present a hazard to employees or to water deliveries because of inadequate anchoring or bracing. The periodic inspections are also documented with a written report. Damage assessment Damage assessment focuses on the after-effects of an earthquake. A Damage Assessment Team (DAT) is assigned to each branch of the distribution system. Each of the seven DATs is comprised of two engineers from the Structural, Civil, Mechanical, Electrical, and Field Inspection disciplines. Once an earthquake occurs, DATs assist the emergency response personnel in their assessment of the damage caused by the earthquake. The DAT uses this information to: 1. Make recommendations to Metropolitan s Water System Operations Group to assist them in their responsibility to maintain or restore service to the water distribution system 2. Prevent or minimize the same damage from occurring in other facilities and in future designs Summary The mission of the Metropolitan Water District of Southern California is to provide its service area with adequate and reliable supplies of high-quality water to meet present and future needs in an environmentally and economically responsible way. Since its inception, Metropolitan s Seismic Program has been consistent with this mission and has resulted in the evaluation of nearly 50% of its facilities. The seismic upgrade of Metropolitan s essential facilities, particularly its Colorado River Aqueduct and associated pumping plants, has helped Metropolitan maintain uninterrupted flow of imported water into Southern California providing a major portion of the imported water demand of its 18 million customers. Metropolitan will continue its Seismic Program in the face of southern California s ongoing seismic vulnerability, with the ultimate intent of providing 100% dependability of the structures and water conveyance systems which constitute the backbone of Metropolitan s system. 212

231 A Fast Simulation Method for Predicting Seismic Responses of An Extensive Water Distribution Network Nobuhisa Suzuki JFE R&D Corporation, Kawasaki, Japan ABSTRACT An extremely fast simulation method applicable to nonlinear deformation analysis of a buried extensive water distribution network is proposed in this paper, which network consists of continuous pipelines and is subjected to such seismic loadings as temporary ground deformation and permanent ground deformation. It is well known that finite element analysis is versatile and effective to deal with the nonlinear deformation of the water distribution network subjected to various kinds of seismic loadings, however, that requires sophisticated techniques of engineers and sometimes results in unnecessary expenditure. The water distribution network system is so essential for our daily life that the network integrity should be ensured against the temporary ground deformation and the permanent ground deformation. In order to predict damage to the distribution network appropriately, a versatile and effective fast simulation method are required. In order to predict the nonlinear deformation, the water distribution network should be divided into segments each of which consists of a straight element and two boundary elements connected to both ends of the straight element. A nonlinear fundamental equation has been derived representing the nonlinear deformation of the network segment under the effect of the seismic loadings. Databases for the boundary elements should be installed their load versus displacement relationships obtained by finite element analysis. The derived nonlinear fundamental equation regarding the segment can be solved promptly applying an iterative procedure. The proposed simulation method provides both simplicity of design formulas and versatility of computer codes preserving required accuracy for practical use. Therefore the simulation method can deal with the extensive water distribution network system and yield very accurate solutions shortly. Key words: buried pipeline, pipeline network, water distribution, seismic design Nobuhisa Suzuki, Proncipal Researcher, JFE R&D Corporation, 1-1 Minami-Watarida, Kawasaki, JAPAN Tel: , Fax: , [email protected] 213

232 INTRODUCTION It may be easy for pipeline engineers to take a part from a high-pressure pipeline as an analytical model to predict their nonlinear behaviors induced by temporary ground deformation because the geometry of the high-pressure pipeline is recognized to be straight [1]. While the other pipelines such as middle-pressure and low-pressure pipelines have essentially a complicated geometry, where pipe bends and tee branches and other fittings have been built in, and it may be cumbersome for the pipeline engineers to idealize them to predict their deformation taking into account the same ground deformation mentioned above [2]. In order to predict the deformations, several design formulas and finite element analysis are applicable to seismic design of buried pipelines. The design formulas have been derived to be simple and conservative based on simply idealized models [1][2]. On the other hand the finite element analysis can deal with complicated pipeline models, however, it requires sophisticated skills and sometimes results in a large expenses. A very fast simulation method is proposed in this paper, which is effective to solve nonlinear deformation of a buried extensive pipeline network subjected to seismic loadings such as temporary ground deformation and permanent ground deformation [3][4][5]. The extensive pipeline networks should be divided into network segments each of which consists of a straight element and two boundary elements. A nonlinear fundamental equation is derived to solve deformation of the network segments under the effect of the seismic loadings. The derived nonlinear fundamental equation can be solved successfully applying an iterative procedure [3][5]. The proposed fast analytical method provides both the simplicity of design formulas and the versatility of computer codes preserving required accuracy for practical use. The solution obtained by the proposed fast simulation method presents very good agreement with that obtained by the finite element analysis whose discrepancy is recognized to be less than several percent. Also the proposed method is extremely swift to deal with the network models and the number of the models can be approximately from 5000 to times as large number as a finite element analysis code can handle. The proposed fast simulation method enables us to perform seismic diagnosis of the buried extensive water distribution network successfully keeping the accuracy of the solution as same as that of the finite element analysis. SEGMENTATION OF PIPELINE NETWORK Deformation of Pipeline Network to Temporary Ground Deformation Figure 1 represents an example of a part of a continuous pipeline network, where the nominal pipe diameters of 150 and 200 mm and the length of the straight pipes of 100 m and the wavelength of 200 m are assumed for finite element analysis. As shown in the figure, the geometry is different from that of high-pressure pipelines, in which pipe bends and tee branches are built in and the length of the straight pipes is shorter than the wavelength. The example network represented in Fig. 1 is provided to explain responses of the continuous buried pipeline network to the temporary ground deformation, which are the theoretical basis to idealize the pipeline network. The results will also be used later for comparison of the results obtained by the finite element analysis and the proposed fast simulation method. Figures 2, 3 and 4 show the results obtained by the finite element analysis regarding axial force and bending moment induced in Segment No.1, which consists of a straight pipe and two pipe bends and straight pipes connected to the pipe bends. Figure 2 shows a surface wave traveling along the 214

233 straight pipe of Segment No.1, in which the surface wave has a wavelength of 200 m and the Y φ200mm φ200mm φ200mm φ200mm Seg.2 φ150mm Seg.1 φ200mm φ150mm φ200mm φ200mm 100m 100m 50m X Seg.3 φ200mm 50m 100m 100m Fig. 1 Example of a part of a buried pipeline network 5cm Fig. 2 Surface wave propagating along Segment No.1 0.2% 0.2% Fig. 3 Longitudinal strain distribution of Segment No.1 Fig. 4 Bending strain distribution of Segment No.1 215

234 amplitude of ground displacement of 4 cm. As shown in the figure, the node of the wave is located at a quarter point of the straight pipe. Figure 3 represents longitudinal strain distribution of Segment No.1, due to which it is recognized that longitudinal strains are mainly induced in the straight pipe and the maximum strain is generated near the node of the wave, which strain is estimated to be approximately 0.1 %. Figure 4 represents bending moment distribution, in which the maximum bending strain of 0.1 % is observed in the left hand side straight pipe connected to the pipe bend. The longitudinal deformation and the bending deformation of Segment No.1 are not discussed in this paper as they can easily be estimated based on the longitudinal strain distribution and the bending moment distribution. Among the results obtained by the finite element analysis, we would like to underline the specific tendency of strain distribution that the longitudinal strain distributes along the straight pipe and the bending strain distributes in the straight pipe connected to the pipe bend and the bending strain vanishes at a points 10 m from the pipe bend. Definition of Segments to Idealize the Distribution Network The results obtained by the finite element analysis, in which the entire network was idealized, revealed that the deformation induced in a pipeline network could be almost the same as that obtained by such segments as shown in Fig. 2. In other words, the results derived by the network segment can be recognized to be independent of each other. This basic property of deformation gives us a hint to apply the segment mentioned above when we idealize the distribution network. Based on the basic property of the deformation, we will be able to propose a new simulation method associated with a very fast solution procedure. When the surface wave propagates over a pipeline network as shown in Fig. 1, we have to define three types of the segments represented in Fig. 5 for idealization of the pipeline network. The component of Segment No.1 was already mentioned above. Segment No.2 consists of two tee branches connected to a straight pipe in the center, and Segment No.3 has a pipe bend and a tee branch at both ends of the straight pipe. Seg.1 Seg.2 Seg.3 Fig. 5 Definition of network segments to idealize the pipeline network 216

235 SIMULATION PROCEDURE OF THE NETWORK SEGMENTS Description of the Network Segment The simulation procedure applied to deformation analysis of the network segments is explained as follows taking Segment No.1 represented in Fig. 1 for example. The top figure in Fig. 6 represents Segment No.1 along which a seismic wave is traveling. The amplitude of the ground displacement in the longitudinal direction is written as U h in the figure. The middle figure represents an analytical model of the segment, where the pipe bends and two straight pipes connected to the bends are represented by nonlinear springs. The components of the nonlinear springs are recognized as boundary elements of the straight element, to which the nonlinear springs are connected. And the bottom figure represents external forces acting on the straight element in the longitudinal direction, which are friction forces produced by the pipe-soil interaction and reaction forces associated with the boundary elements. 0 y u g Uh x u g 0 δbl L 0 L R τ δbr F BL τ F BR L L Fig.6 Analytical model of a network segment Assumptions for Analysis of the Network Segment The following assumptions are taken into account in order to derive a fundamental solution of the network segments connected to various boundary elements. Stress-strain relationship of pipe The stress-strain curve is a Round-House model and can be represented by the following three parameter representation, the Ramberg-Osgood formula [6]. N σ ασ 0 σ ε = + E E (1) σ 0 Where E represents Young s modulus and σ 0 represents yield stress correspond to a specified stress at 0.5 % strain. The parameters α and N are the R-O constants depend on work-hardening property of materials. 217

236 Pipe-soil interaction The soil spring of a rigid-perfectly plastic model is assumed in the longitudinal direction and another soil spring of an elastic-perfectly plastic model is taken into consideration in the transverse direction. Temporary ground deformation Temporary ground deformation to be considered in the derivation of the fundamental equation shall be characterized by surface waves, in which the direction of ground motion coincides with that of wave propagation. The temporary ground deformation is defined by a design spectrum in the Seismic Design Codes for High-Pressure Gas Pipelines in Japan (2000) and the surface wave is assumed to propagate along a buried pipeline. Solution for the Network Segment (The First Step of the Proposed Simulation) Deformation of the network segment represented in Fig. 6 can be derived as follows when the value of x is positive. Relationship between the axial force and the friction force has to satisfy equation (2) and if we substitute equation (2) into equation (1), longitudinal stress can be expressed as equation (3). R = f τ ( LR x ) FBR (2) F ( x ) + FR ( x ) fτ FBR ( x ) = = ( LR x ) A A A σ (3) R + Where F R (x) represents longitudinal stress when x-axis is positive and F BR expresses a reaction force from the right hand side boundary element. And the friction force per unit length acting on the pipe is f τ =πdτ. Longitudinal strain of the pipeline can easily be obtained as equation (4) by putting equation (3) into equation (1) and longitudinal displacement can be represented as equation (5) after integrating equation (4). Equation (6) gives the solution regarding the longitudinal displacement which was derived from equations (5) and (4). If we substitute X = 0 into equation (6), we can obtain equation (7) which represents longitudinal displacement at the origin. And if we substitute X = LR into the same equation, we can derive equation (8) which represents the longitudinal displacement at the right hand side end of the pipe. N σ R( x ) ασ R( x ) ε R( x ) = + (4) E N 1 Eσ 0 u pr ( x ) 1 f ( L EA = τ R x ) + F 2 pr = x 0 R u ( x ) ε ( x ) dx (5) BR N + 1 N + 1 [{ } { } ] α fτ LR + FBR fτ ( L R x ) + FBR x + (6) N N 1 ( N + 1) f EA σ τ 0 u pr (0 ) = 0 (7) 218

237 u pr ( L R N + 1 N + 1 { } LR LR α ( f L + FB ) F ) = fτ + FBR + N N EA 2 ( N + 1) f EA σ τ (8) Furthermore, equation (9) represents reaction of the right hand side boundary element applied to the straight pipe. The parameter K BR used in the equation represents a nonlinear spring coefficient of the boundary element that can be obtained considering the combined effect of the nonlinear pipe-soil interaction and the nonlinear stress-strain relationship of the pipe. F BR { u ( L ) u ( L )} τ B 1 0 = K (9) BR gr Relative displacement in the longitudinal direction between the pipeline and the ground can be written as equation (10), if we rewrite the equation as equation (11) and substitute that into equation (8), we can deduce a nonlinear solution regarding the relative displacement δ BR as equation (12). R pr R u gr ( LR ) u pr( LR ) = δ (10) BR u pr ( LR ) ugr( LR ) = δ (11) N + 1 N fτ 2 α{ ( f L } R + K BRδ BR ) ( K BRδ BR ) δ BR = EAu g ( LR ) LR N 1 EA + K BR LR 2 ( N + 1) fτ ( Aσ 0 ) BR τ (12) The unknown parameters involved in equation (12) are δ BR and K BR, however, we can solve the nonlinear equation as K BR is dependent on δ BR as shown in Fig. 7. Moreover, similar reduction regarding the left hand side boundary element to that mentioned above can be performed. And we can derive another nonlinear solution as equation (13) represented in terms of the left hand side relative displacement. N + 1 N fτ 2 α{ ( f L } L + K BLδ BL ) ( K BLδ BL ) δ BL = EAu g ( LL ) LL N 1 EA + K BL LL 2 ( N + 1) fτ ( Aσ 0 ) τ (13) F B K B Fig.7 Force-displacement relationship of a boundary element δ B 219

238 Solution for the Network Segment (The Second Step of the Proposed Simulation) Equations (12) and (13) are derived referring to the bottom figure in Fig. 6, due to which equations we can obtain the relative displacements, δ BR and δ BL. And the corresponding reaction forces, F BR and F BL, can be expressed if we know δ BR and δ BL. Then we can derive an equation that satisfies static equilibrium of the network segment in the longitudinal direction, which can be written in terms of the friction force and the reaction forces. Total amount of the external forces can be written as equation (14) and the equation gives the location of the point as equation (15) where the sign of the friction changes as shown in Fig. 8, which means the change of direction of the friction force. The variable s represented by equation (15) is a correction parameter for L R and L L, which shall be updated successively in accordance with equations (16) and (17). F BR + f L F f L f s (14) τ R L τ L = τ F s = R F f τ L + L R L L (15) L R( i+ 1 ) = L R( i ) s ( i ) (16) L + L( i+ 1 ) = LL( i ) s( i ) (17) After updating the neutral point where the direction of the friction forces change, the first step simulation procedure mentioned above should be conducted again in order to investigate the static equilibrium of the external forces. This iterative procedure from the first step to the second step shall be done until the parameter s converges when the solution is recognized to have a required accuracy. The converged value of s gives the final solution of δ BR and δ BL. L 0 L R τ F BL τ L L u p S τ F BR F BL(i) τ F BR(i) Fig. 8 Updating friction zones considering static equilibrium of external forces 220

239 VERIFICATION OF THE PROPOSED SIMULATION METHOD Distribution Network Model In order to verify the accuracy of the results obtained by the proposed method, the results are compared to those obtained by the finite element analysis. The distribution network model to be discussed in this section is represented in Fig. 9 whose geometry is the same as the network model presented in Fig. 1 provided for the finite element analysis. Figure 9 shows additional information regarding node numbers of the network model in order to compare the results with respect to displacement at every node. And the wavelengths of 200, 300 and 400 m are considered for the comparison and the location of the surface waves are also explained in the figure. Therefore the lengths of network segments will vary in accordance with the wavelength. Assumptions for the Solution Procedure The stress-strain curve shown in Fig. 10 was applied to the analysis and the soil springs in the longitudinal and the transverse directions were defined as Fig. 11 in accordance with the Seismic Design Codes for High-Pressure Gas Pipelines in Japan (2000). The relationship of the reaction force versus displacement of the boundary elements, the database, were obtained by the finite element analysis, whose relationship is represented as Fig. 12 applying the data represented by Figs. 10 and 11. The notations explained in Fig. 12 regarding the database correspond to the fittings included in the boundary elements. Verification of the Proposed Simulation Method Tables 1 through 3 compare the results obtained by the proposed method and the finite element analysis, in which longitudinal displacements at the both ends of the straight element are compared and the numbers written in the tables coincide with those in Fig. 9. And the relative discrepancies represented in the tables are calculated by dividing the difference between the two methods by the finite element solution. And the signs of the calculated results in the tables represent the direction of the nodal displacement. We can observe very good agreements between the results of the proposed method and the finite element analysis in the tables, in which the errors vary from 0.1 % to 11.1 %. Case 2-2 gives the maximum error of 11.1 %, however, the calculated longitudinal displacement is 0.1 cm which is small enough for practical use. Besides the comparison mentioned above we can observe that some data have the error of 5 to 7 %, however, their longitudinal displacements are approximately 0.3 cm and also enough small to neglect. Moreover, other than the network model shown in Figs. 1 and 9, an extensive distribution network spreading over a rectangular area of 1x2 km was analyzed to compare with the efficiency of the proposed method and the finite element analysis. Fujitsu-VX and DELL Dimension-XPS R400 were used for the finite element analysis and the proposed, respectively. The computing time of the proposed method is 1/8000 times as short as that of the finite element analysis. Taking into account the other results, the computing time of the proposed method is very short compare with the finite element analysis, which can be estimated to be from 1/5000 to 1/ Based on the comparison regarding the computing time, we can conclude that the proposed method is extremely swift compare with the finite element analysis. 221

240 σ (MPa) E = 206GPa α = 0.5 N = 50 σ 0 = 290MPa ε (%) Fig. 10 Stress strain relationship of pipes 6 7 Friction 15kPa φ150 and φ200 Reaction 510kPa 480kPa φ150 φ200 Case cm 2.6cm Case1-2 Displacement Displacement Case2-1 Case2-2 Fig. 11 Soil springs for 150mm and 200mm diameter pipes Case3-1 Case φ200tee Fig. 9 Definition of nodes of the network and locations of propagating surface waves F B (kn) φ150bend φ200bend δ B (cm) Fig. 12 Load displacement relationship of boundary elements 222

241 Table 1 Examples with wavelength of 200 m Case 1-1 Case 1-2 δ B (cm) δ B (cm) Proposed Method FEA Error (%) Proposed Method FEA Error (%) Table 2 Examples with wavelength of 300 m Case 2-1 Case 2-2 δ B (cm) δ B (cm) Proposed Method FEM Error (%) Proposed Method FEM Error (%) Table 3 Examples with wavelength of 400 m Case 3-1 Case 3-2 δ B (cm) δ B (cm) Proposed Method FEM Error (%) Proposed Method FEM Error (%) CONCLUSION A fast simulation method is proposed in this paper, which is applicable to deformation analysis of a pipeline network under the effect of seismic excitations. The fast simulation method proposed in this paper has been developed to analyze responses of continuous pipeline networks, however, the method is also applicable to the networks consisted of segmented pipelines providing they behave as a continuous pipeline only in a compression zone. Therefore, the proposed simulation method is effective to conduct seismic design and seismic diagnosis of various buried pipeline networks. As for the database regarding the deformation property of the boundary element, the accuracy 223

242 and the quality of the database has been confirmed because the database is constructed based on the results obtained by the finite element analyses. The proposed fast simulation method enables us to conduct deformation analysis of an extensive distribution network very shortly keeping the accuracy of the finite element analysis without paying additional efforts to conduct the finite element analysis. Besides the boundary elements discussed in this paper, the pipe bend and the tee branch, other boundary elements can be taken into account such as a pipe bend with an arbitrary bending angle and a single crank loop and a double crank loop. We can storage other useful information regarding deformation of the straight pipes connected to the pipe bend and the tee branch for example in the database. Then we can afford to refer other information to be required for the fast simulation analysis. ACKNOWLEDGMENTS The proposed fast simulation method in this paper was originally developed in order to conduct seismic diagnosis of an extensive gas distribution network system in Tokyo metropolitan area. The author would like to thank Mr. A. Fiujita, Dr. Y. Shimizu, Mr. K. Koganemaru, Mr. Y. Hatsuda, Dr. K. Yoshizaki and Mr. N. Hosokawa of Tokyo Gas Co. Ltd., Mr. T. Iwamatsu of Tokyo Gas Engineering Co. Ltd. and Mr. T. Mayumi, Mr. T. Mori, Mr. H. Horikawa and Mr. N. Hara of JFE Engineering Corporation for their invaluable comments and advices. The author also wishes to thank Mr. I. Kubo of Japan Industrial Testing Co. Ltd. for his support regarding program development and the finite element analyses. REFERENCES [1] JGA, 2000: Recommended Practices for Seismic Design of High-Pressure Gas Pipelines. [2] JGA, 1983: Recommended Practices for Seismic Design of Middle- and Low-Pressure Gas Pipelines. [3] Shimizu, Y., Koganemaru and Suzuki, N., 2002: Development of Earthquake Resistance Evaluation Method for Buried Pipeline Networks. [4] Koganemaru., K., Shimizu, Y., Yoshizaki, K. and Hatsuta, Y., Suzuki, N., Horikawa, H., Mori T. and Mayumi, T., 2002: Development of Evaluation Technique of Pipeline Network Integrity during Seismic Excitations. [5] Suzuki, N., Horikawa, H., Mori, T., Mayumi, T., Shimizu, Y., Koganemaru., K., Yoshizaki, K. and Hatsuta, Y., 2002: A Fast Solver for Responses of Buried Pipeline Networks, 11 th Japan Earthquake Engineering Symposium. [6] Ramberg, W. and Osgood, W. R., 1943: Description of Stress-Strain Curves by Three Parameters, NACA, TN

243 Multi-Hazard Risk Assessments of Water Systems, Elements In Common with Seismic, Security, and Other Risk Studies Donald Ballantyne 1 ABSTRACT This risk assessment methodology is used to screen mutli-hazard risks for water systems. Multihazard risk assessment is becoming more popular as utilities try to balance the risk across multiple hazards. The Federal Government has instituted the Disaster Mitigation Act of 2000 (DMA 2000) that requires to public agencies to conduct multi-hazard assessments if they are to receive future hazard mitigation grant funding. The DMA 2000 methodology employs a similar analytical approach as that developed for water system assessment.. The methodology also has many common components with the Risk Assessment Methodology for Water (RAM-W SM ) required for evaluation of he security risk to water utilities. The methodology uses a basic risk equation and quantifies value ranges for each term hazard, vulnerability, consequences, and correlation factor. The terms are multiplied together resulting in a number representing the risk from a given hazard. Hazard risk is ranked for all hazards and/or for each facility within the system. The ranking can be used to select hazards for further more detailed analysis. More detailed evaluations can be conducted using fault tree analyses and/or hydraulic network analyses. Benefit/cost analyses are used to justify mitigation measures. INTRODUCTION This paper presents the risk assessment methodology used to screen hazard risk for water systems. The methodology was developed and refined over five multi-hazard risk assessment projects of water systems in the United States and Canada. The evaluated systems range in size from those serving 50,000 to over one million people. There is increasing attention paid to evaluating risk from an all-hazards perspective. Owners are subjected to risks from multiple hazards, and are interested in mitigating them using a balanced approach. Mitigation alternatives can be developed that can be used to address multiple hazards. Many components of the analysis are common and can be applied at a small additional cost compared to analysis for one hazard. The United States Congress passed the Disaster Mitigation Act in 2000 (DMA 2000). This law requires public sector organizations to prepare multi-hazard mitigation plans in order to be eligible for future hazard mitigation grant funding from the Federal Government. The same general approach applied to the water systems is required, to comply with DMA One of the more recent multi- 1 Donald Ballantyne PE, VP Lifeline Services, ABS Consulting (formerly EQE International), Seattle, WA [email protected] 225

244 hazard risk assessment projects referenced in this paper conducted the evaluation in accordance with the Federal requirements. Following the September 11, 2001 strike on the World Trade Center, the Environmental Protection Agency was given the responsibility to develop a program to protect the nation s water supply against terrorist attacks. The Risk Assessment Methodology for Water (RAM-W SM ) used a methodology with the same elements used in the previous water system risk assessment studies. Risk can be calculated as: Where: RISK EQUATION FORMULATION Risk = Hazard x Vulnerability x Consequence x Correlation Factor Hazard is the probability of exceedance of a given intensity over a given time period Vulnerability is the probability that the given facility will fail when subjected to the given hazard intensity. Consequence is a measure of loss in terms of dollars, loss of life, or other comparable parameters. Correlation Factor is a measure of the likelihood that the hazard will impact multiple facilities in a single event. The hazard is often defined as probability of the given intensity being exceeded within the facilities expected 50 year life. A 50 year life is taken as being representative of civil engineering facilities where mechanical equipment may have a 20 year life, buildings a 50 year life, and buried pipelines a 100-year life. A probability of 50 percent in 50 years represents a 72-year return period, and a probability of 10 percent in 50 years represents a 475-year return period, the basis of many building codes. For system failures would result in catastrophic consequences such as dam failure, the probabilities of exceedance would be increased to for example having a return period of 10,000 years. Vulnerability can be considered using fragility curves, that relate the hazard intensity parameter (ground motion) to the probability of failure. The stronger ground motion, the higher the probability of failure. Consequence can be quantified using various parameters such as economic impact in dollars, or in catastrophic hazard events, loss of life. Lifeline systems are sometimes described as interconnected facilities (networks) of facilities (nodes) with pipelines or wires making up the network links. Analysis of these systems can become very complex. Simplified methods considering the consequence of failure have been used to take into account the importance (criticality) of each facility within the system. Consequence parameters representing dollar losses are sometimes put in terms of percent of system outage times outage duration. The consequence units can be customer outage days that can then be converted to dollars. The consequence term takes into account system redundancy. System facilities that are redundant may have a low consequence of failure. A correlation factor has not typically been used in earthquake analysis of lifeline systems within a limited area because there is a high probability that all of the facilities would be exposed to the hazard all at one time. In this situation, the correlation factor would approach 100 percent. For lifeline 226

245 systems covering a larger area such a regional power systems, the correlation factor may be smaller as many of the facilities would be outside the earthquake impact area. This approach has the advantage that the numerical basis and relative magnitude of each term compared to each other term is correct. That is for example, probabilities and percentage of service areas served are combined using the correct units so the results approach realistic values of risk. Other approaches employing high/medium/low values without a numerical basis only provide approximations of the relative risk. More detailed analyses using fault tree analysis or hydraulic network modeling can be used provide more thorough evaluation of the systems once the hazard screening focuses the effort. PROJECT/UTILITY BACKGROUNDS The general multi-hazard risk assessment methodology has been applied to five water systems, two of which are described below. The other situations are generally described without providing specific system information. The Portland water system provides retail water service to the City of Portland, Oregon and wholesale water to a number of suburban utilities. The primary source is the Bull Run surface supply (BRS), with a secondary groundwater supply (GWS) from a well field just south of the Columbia River. The Bull Run supply consists of dams, chlorination/treatment facilities, and three parallel conduits in different corridors, each with bridges and trestles. Segments of the conduit system are vulnerable to landslides. The groundwater supply includes wells, a collection piping system, and a major pump station and provides almost 100 percent redundancy with the BRS at winter flow demands. The GWS facilities and Willamette River crossings are particularly vulnerable to liquefaction. Refer to Figures 1 and 2. Mt. Hood Columbia River Groundwater System Bull Run Watershed Powell Butte Reservoir Sandy River Willamette River City Center Figure 1. Portland water system geographic overview N 227

246 Figure 2. Portland Oregon supply system schematic. Portland is in a region with moderate seismicity, and had evaluated the earthquake vulnerability of a number of their critical facilities. They were ready to proceed with mitigation projects when in the winter of , they nearly lost service to at least some customers as a result of a series of storms causing landslides, increased turbidity, and flooding. As a result, they decided to conduct a multihazard assessment of the overall system. The Greater Vancouver Regional District (GVRD) is located in British Columbia, Canada. Their supply comes from three lakes that independently feed south into the distribution system (Figure 3). The pipelines crossing the Burrard Inlet and the Frazier River are vulnerable to liquefaction. Seymour Supply Capilano Supply Coquitlam Supply Burrard Inlet Vancouver Frazier River N Figure 3. Greater Vancouver Regional District water supply. 228

247 In 1993, the GVRD conducted an earthquake vulnerability assessment. In 2000, the GVRD evaluated the status of implementation of the 1993 report recommendations, and conducted a multi-hazard assessment of the system driven in part by flooding and other regional hazard events. In 2002, a water utility serving a suburban Portland, Oregon community elected to conduct a multihazard assessment (including a detailed seismic evaluation) in parallel with a security vulnerability assessment. The system has a single primary supply but has interties that would allow it to operate should it s main supply be disrupted. For this project, many of the evaluation elements were common between the multi-hazard and security risk assessment. The resulting recommendations were able to integrate and prioritize the results across hazards. A suburban Seattle, Washington water utility integrated a multi-hazard risk assessment, a security risk assessment, and seismic upgrade project. The utility is highly dependent of the Seattle Public Utility supply. The multi-hazard assessment was conducted in accordance with the DMA-2000 regulations. The consequence analysis developed in the security vulnerability assessment was integrated into the multi-hazard assessment so that both studies had a common basis. The fifth system where the methodology is applied provides wholesale water to water utilities in a densely populated region of Northern California. The system has multiple supplies and multiple treatment facilities. It is subjected to earthquakes, flooding, and other hazards common to utilities across the United States. HAZARD IDENTIFICATION AND QUANTIFICATION In each of the five projects, the first step was to identify and quantify hazards. A list of the hazards considered for the Portland project is shown in Table 1. Hazards are identified and quantified through a literature search, discussions with regional emergency management personnel, and discussions with utility personnel. The probability of occurrence of a hazard event is taken for one that will produce an intensity that will cause failure of at least one significant water system component. Table 1. Hazards Screened Hazard Category/Type Hazard Category/Type Hazard Category/Type Natural Human/Technological Transportation River Flooding, Dike Failure Staff Unavailable (Public Health, Labor Airplane Dispute) Snow Melt/Rain on Snow Intentional Act Airplane Fuel Dump Land/Rock Slides/Debris Flow Bureau Building Piping Failure/Flood Truck/Car Structural Impact Tree Fall Bureau Building/Facility Fire/Explosion Marine River Crossing Winter Snow/Ice Storms Chemical Release Light Rail Prolonged Freezing Computer Disruption Lifeline Service Loss Earthquake Groundwater Contamination Electrical Seiche Dams Failure Wire Communications Volcanic Activity Mechanical Failure Wireless Communications Fire in Watershed Third-Party Damage Sewer Urban Firestorms Operational Error Natural Gas/Propane Turbidity Liquid Fuel Microbial Contamination Treatment Chemical Supply/Delivery 229

248 The probability of occurrence can be estimated to be in one of three categories as for example: High Probability - Damaging intensity to occur in less than 50 years (mid point of 25 years). The probability of occurrence in 50 years is 100 percent. Moderate Probability Damaging intensity to occur once every 50 to 250 years (mid point 72 years). The probability of occurrence in 50 years is 50 percent. Low Probability Damaging intensity to occur once every 250 to 1,000 years (mid point 475 years). The probability of occurrence in 50 years is 10 percent. Hazards with recurrence intervals greater than 1,000 years are beyond the planning horizon for each of the five projects. VULNERABILITY ASSESSMENT Once the hazard intensity is established, a representative scenario event that could produce the hazard intensity is applied to each system facility, and the probability of failure determined. For each hazard (other than an intentional act (security), the vulnerability can categorized to be in one of three groups as for example: High Vulnerability likely, greater than 50% probability (mid point 70%), to result in loss of service for extended period when subjected to associated hazard intensity. Moderately Vulnerability possibly, 10 % to 50% probability (mid point 25%), result in loss of service for extended period when subjected to hazard intensity. Low Vulnerability unlikely, less than 10 % probability (mid point 7%) to result in loss of service for extended period when subjected to hazard intensity. CONSEQUENCE ANALYSIS The consequences of failure of the facilities are also categorized into one of four groups: Very High Consequences - result in loss of service to entire system (100%). This could occur if all water sources became unavailable or something occurred in the transmission system that would make it entirely inoperable. No VH consequences were identified any of the systems due to system redundancy. High Consequence result in loss of service to a significant part of system (30% to 50%, depending on the proportion of customers affected). This could occur if there were failures distribution lines or failure of a key pump station that resulted in loss of service to a significant area of the system. Moderate Consequence result in loss of service to localized area within service area (10%). This could occur for example as the result of failure of one pump station. Low Consequence would not result in loss of service but may require aggressive response actions to maintain service. 1% loss of service used for analysis. 230

249 CORRELATION OF HAZARDS The correlation of hazards is applicable to losses that could also cause the unavailability of system interties A factor of 1 indicates no correlation, that is to say, the hazard in question would not be expected to affect the availability of system interties. Other correlation factors are 5 (low correlation), 25 (moderate correlation), and 75 (high correlation). For example, flood event could be expected to impact multiple watersheds (depending on the regional topography). Such impacts may cause the system interties to be unavailable, and thus a high correlation factor was used. HAZARD RISK TABLE EXAMPLE Table 2 combines the hazard, vulnerability consequences, and correlation factor in a table format. Table 2, Example Hazard Risk Table HAZARD Probability Rank Occurrence in 50 years Return (years) Range H 100% years M 50% years L 10% 475 > 250 years VULNERABILITY Probability of Failure H 70% Likely >50% M 25% Possibly 10-50% L 7% Unlikely < 10% CONSEQUENCES System Impact VH 100% Entire system H 40% 1 or 2 pressure zones M 10% Limited area L 1% No outage CORRELATION VH 75 Very high correlation H 25 High correlation M 5 Moderate correlation L 1 No correlation SYSTEM VULNERABILITY Consequence of Failure (% people without water) SECURITY NATURAL H SYSTEM / COMPONENT Hazard Probability (% in 50 years) 50% 50% Correlation Factor (See note 1) 1 75 North Side Watershed 1% 70% 0% 7% 3% Intake Structure 1% 70% 0% 7% 3% Lowlift Pump Station 1% 70% 0% 70% 26% Treatment Plant 1% 70% 0% 7% 3% Security Floods - Clackamas River HAZARD RISK ANALYSIS RESULTS After each facility and its corresponding consequence of failure, is evaluated for each hazard and its corresponding probability of occurrence, the results can be sorted by hazard (Table 3) or facility (Table 4). The Average risk shown in the tables is the average for each facility exposed to the hazard. For example, a facility not located in a flood plain would not be exposed to flood. The Maximum risk is for the facility with the highest risk for the selected hazard (Table 3), or the hazard that presents the highest risk to the selected facility (Table 4). The results of the analysis are used to prioritize more detailed analysis and/or mitigation. Table 3, Hazards Ranked by Risk Levels Hazard Average Maximum Risk Floods - 3.6% 26.3% VH Earthquake 3.1% 18.8% VH Security (malevolent act) 1.9% 10.5% VH 231

250 Hazard Average Maximum Risk Building / Facility Fire / Explosion 1.3% 7.5% H Winter Snow / Ice Storms 1.0% 7.5% H Land / Rock Slides / Debris Flow 2.5% 6.3% H Microbial Contamination 3.8% 3.8% M Urban Firestorms 1.1% 3.8% M Building Piping Failure / Flood 0.7% 3.8% M Treatment Chemical Supply / Delivery 0.7% 2.6% L Truck / Car Structural Impact 1.0% 2.1% L Mechanical Failure 0.5% 2.1% L Operational Error 0.4% 2.1% L Electrical 0.4% 2.1% L Tree Fall 0.4% 2.1% L Third-party Damage 1.1% 1.1% L Natural Gas / Propane 0.2% 0.4% L Liquid Fuel 0.1% 0.4% L Turbidity 0.4% 0.4% L Prolonged Freezing 0.3% 0.3% L Volcanic Activity 0.0% 0.2% L Chemical Release 0.1% 0.1% L Wire Communications 0.1% 0.1% L Sewer 0.1% 0.1% L Fire in Watershed 0.1% 0.1% L Staff Unavailable 0.1% 0.1% L Airplane Crash 0.0% 0.0% L Airplane Fuel Dump 0.0% 0.0% L Wireless Communications 0.0% 0.0% L Table 4, Facilities Ranked by Hazard Risk Levels Facility Average Maximum Risk Lowlift Pump Station 3.7% 26.3% VH Treatment Plant 2.0% 18.8% VH Reservoir A 6.5% 18.8% VH Reservoir B 6.5% 18.8% VH Distribution System 5.0% 10.5% VH Highlift / Clearwell 1.4% 9.4% H Watershed 0.7% 2.6% L Intake Structure 1.8% 2.6% L Disinfection System 0.7% 2.6% L Pump Station 1 0.5% 2.6% L Pump Station 2 0.5% 2.6% L Computer System 0.1% 0.4% L SCADA System 0.1% 0.4% L Administration Building 0.1% 0.4% L Operations Office 0.1% 0.4% L Operations Shop 0.1% 0.4% L APPLICATION TO SECURITY VULNERABILITY ASSESSMENTS OF LIFELINE SYSTEMS September 11, 2001 changed the world we live in. The Environmental Protection Agency was given the responsibility to develop a plan to protect our nations drinking water supplies from terrorists. Working with the American Water Works Research Foundation and Sandia National Laboratories, a 232

251 Risk Assessment Methodology for Water (RAM-W SM ) was developed to evaluate the security risk of water systems. The risk equation formulation is the same as we have been using for earthquake risk assessments of water systems. The only difference was that there was no detailed information on the hazard. The RAM-W SM approach set the hazard probability at 1. The associated hazard intensity is taken to be the design basis threat, a threat that the owner would like to design to resist. For example, the design basis threat might be defined to be three people with hand tools, small explosives, hand guns, and a motor vehicle. Defining this threat keeps the evaluator on track, and keeps them from considering for example a weapon of mass destruction. This approach is directly comparable to using the design basis earthquake. The vulnerability parameter was estimated using the same general approach, with a given hazard, what is the probability that the attackers will be successful. The consequence analysis is exactly the same as that used for the earthquake and multi-hazard analysis. The correlation factor is applicable as well. The correlation factor is low because there is a very low probability that a terrorist cell would attack more than one or two system facilities at any one time. FAULT TREE ANALYSIS For the Portland system, the system reliability was evaluated for the two highest-risk hazards, earthquake and intense rain. Based on Figure 2, a logical model was built connecting the components by two possible means: OR or AND gates. This logical model assumes that all components are independent of each other. The term reliability is used for the probability that a component, subsystem or system to be functional. An OR gate indicates that a component is functional if at least one of the connecting components is functional. For example, water to Powell Butte Reservoir is supplied if either the GWS OR the Bull Run supply is functional. Assuming that these events are independent, the only case when the reservoir does not receive water is when both supplies are nonfunctional. A system formed by OR gates is known as a parallel system. If we define PGWS and PBRS as the probabilities that the GWS and Bull Run supplies are functional, then the probability of supplying water to Powell Butte Reservoir, PPBR, is (Devore, J, 1995): P PBR = 1 - [(1 - P GWS )(1 - P BRS )] (2) It should be noted that (1 - PGWS) and (1 - PBRS) are the probabilities of the GWS and BRS of not being functional. AND gates indicate that the top component is functional if all of the sub-components are functional. Assume that the BRS subsystem from Figure 2 is only formed by the watershed and headworks chlorine. These two components need to both be functional in order for the BRS to be functional. Failure of any of these components will cause failure of the entire Bull Run supply. These components are said to be in series and the reliability of a system PS with m independent components with functional probabilities Pi, i = 1,, m is (Devore, J, 1995): m P S = i= 1 P i (3) Equation 3 shows that (1) failure of any component causes the system to fail, and (2) system reliability is smaller than any individual reliability. For example, if three components in series each have a 90 percent reliability, the system (probability of all three components remaining function is only 73 percent (0.90 x 0.90 x 0.90). 233

252 Based on these characteristics of series and parallel systems, a connectivity analysis of the Portland supply system was performed by assembling all of the components from the GWS and BRS. Some of the components in these subsystems are made of sub-components. Subsets of the system shown in Figure 2 were evaluated to consider different system capacities (e.g. with different combinations of the three conduits). The results of the reliabilities of each of those configurations were combined for earthquake with the outcome displayed in Figure 4 showing flow versus probability of supplying at least x flow. Figure 4 also shows the performance goals set by Portland for 100-year-return, and 500-year-return earthquakes. The existing system met the performance goals for the 100-year event, but mitigation was required to meet the 500-year performance goal. 1.0 Probability of supplying at least x mgd Probability of supplying at least "x" mgd yr return performance goal 95 near 100% reliability 100-yr return performance goal 145 near 100% reliability Flow (mgd) Flow (mgd) Return Periods 50 years 100 years 500 years 1000 years Figure 4. Earthquake reliability of the Portland Water Supply System. The results of this assessment were used to select the combinations of mitigation measures that would provide the best reliability. For example, it became clear that Portland should focus mitigation on one subsystem (e.g., one conduit/corridor including bridges, trestles, and pipeline) rather than on a reduced level of mitigation for multiple subsystems (e.g., one bridge on each conduit/corridor). HYDRAULIC ANALYSIS OF DAMAGED SYSTEM The third evaluation methodology to be employed was the use of a hydraulic network analysis program to analyze the system in the undamaged state and the post-earthquake damaged state. HAZUS 99 incorporates both the hydraulic network analysis and water system component earthquake loss estimation modules required to perform this analysis. Analysis software has only been developed for earthquake hazards as earthquakes tend to damage many system components during one event whereas other hazards tend to damage only one component during any single event. A schematic of the methodology is shown in Figure 5. The results of distribution analysis show the expected reliability of the overall system following a given earthquake scenario. Specifically, the damage state of each system component and pipeline segment are calculated in accordance with the fragility relationships in the software). The hydraulic analysis then determines the reliability of the system to move water from point A to point B comparing pre- and post-earthquake conditions. The importance of each pipeline segment is shown in 234

253 terms of the flow through the pipe compared to flow in other pipes. This information is then used to compare the expected system performance, to performance goals, and to prioritize system improvements to effectively reach those goals. Hydraulic Model Hydraulic analysis UNDAMAGED System Define seismic event Components Fragility PGA, PGD Apply system demands Simulate damage state Mitigation Hydraulic analysis DAMAGED System Monte Carlo Simulation System Serviceability Figure 5. Hydraulic network analysis schematic of undamaged and damaged system. Boxes inside the dashed line are probabilistic based used the Monte Carlo simulation. BENEFIT-COST ANALYSIS A benefit-cost analysis was performed in the Portland study by developing a benefit-cost ratio (BCR). The BCR is the annualized avoided loss cost, A, divided by the annualized cost of mitigation, M, so: BCR = A/M (4) The annualized cost of mitigation, M, includes the capital cost, C, of any construction or acquisition, interest on the invested money, plus the cost of operating and maintaining the new facilities. The annualized avoided loss cost, A, includes: 1) repair in post-disaster conditions, R, 2) loss of income, I, and 3) societal loss, S, or: A = R+ I + S (5) 235

254 For the capital cost of mitigation, C, the cost of repair after a 500 year-return earthquake was assumed. The loss of income, I to the City of Portland occurs when no water is delivered, and is valued at $53 million annually/365 days/yr = $0.15 million/day. Societal losses, S, are the loss of income to individuals and businesses that cannot function as a result of loss of water service. Societal losses are difficult to estimate, but can be bounded. An upper estimate would be based on the assumption that the productive society would cease without water. Assuming an annual average income of $25,000 per household for the 280,000 households served by Portland (assumes 3 people/household), the daily loss would be about $70/household or about $20 million per day. Income loss estimates for the shutdown of large electronics plants served by Portland are on the order of $1.5 million/day each. For all of industry, plus commercial and residential economic impacts, this could total the same estimate of $20 million/day. However, shutdown of industries following an earthquake (one of the 2 most significant hazards) may not be attributable entirely to loss of water supply. Other contributing factors could include damage to the industrial facilities and loss of other lifeline services such as electrical power or natural gas. For purposes of the study, $10 million/day was used as the estimate for the societal cost for the loss of water. The project report shows calculations of the BCR both with and without societal losses. It should be noted that repair costs and loss of income are small compared to societal losses. Societal losses are not a direct cost to the City of Portland water utility but are a cost to City property owners, and thus owners of the water system. Use of the BCR with societal costs was thus justified. CONCLUSIONS Five water systems were evaluated with up to four risk evaluation methods. The hazard/risk screening and risk quantification method is effective in quantifying the risk associated with individual components for multiple hazards in a lifeline system. The fault tree analysis is effective in quantifying the supply system reliability, and was employed for the highest risk hazards. The hydraulic analysis of damaged distribution system is effective for scenarios where the system is damaged in multiple locations such as earthquakes. The benefit-cost analysis is useful to justify proceeding with mitigation. ACKNOWLEDGMENTS Project Managers and staff with the Portland Oregon Bureau of Waterworks, the Greater Vancouver Regional District, Burnaby, British Columbia, Canada, and the other three unnamed utilities are gratefully acknowledged for their project support. REFERENCES ABS Consulting (Seattle, WA), 2002, Emergency Management Plan, prepared for the Greater Vancouver Regional District, August. Devore, J., Probability and Statistics for Engineering and the Sciences, 4th Edition, Duxbury Press, New York, EQE International (Seattle WA), 2000, System Vulnerability Assessment, prepared for Portland Bureau of Water Works, September. Kennedy/Jenks Consultants (Federal Way, WA) in Association with EQE Engineering and Design, 1993, A Lifeline Study of the Regional Water Distribution System, prepared for Greater Vancouver Regional District. National Institute of Building Sciences (NIBS), 1999, HAZUS 99, Federal Emergency Management Agency, Washington DC. Sandia National Laboratories, 2002, Risk Assessment Methodology for Water Systems (RAM-W SM ), Albuquerque, NM. 236

255 A Study on the Development of a Backup System in a Big Urban Area Fukuhisa Iwasaki, Yoshiharu Sorakuma, and Kyoji Wakamatsu At present, measures against earthquakes such as the construction of aseismic facilities and backup systems are mainly taken by each waterworks plant. However, during the 1995 Hanshin-Awaji Earthquake, waterworks facilities in large cities were seriously damaged, and it was difficult for other waterworks to provide assistance. In order to minimize damage to water systems by earthquakes, a comprehensive disaster-prevention water supply system that covers an entire city needs to be built while individual waterworks can take only limited measures. Thus, some case studies on a wide-area backup system for water supply were conducted, taking the case of Tokyo and three prefectures and assuming that one of the four is affected and the other three provide assistance in the Tokyo metropolitan area. The cost and benefits of building the wide-area backup system that enables water supply to be controlled and managed beyond the service area of each waterworks utilities in case of such an emergency were analyzed. 1. Background Large urban areas today are characterized by intensive utilization of land and underground. Should a large earthquake occur, the damage could be enormous and the functions of government might be lost. Indeed, in the event of the Hanshin-Awaji Earthquake in 1995, the waterworks facilities of large cities suffered serious damage and other waterworks found it difficult to assist. In order to minimize such earthquake damage to water systems, it is necessary to build a comprehensive disaster-prevention water supply system that covers the entire city. Under consignment from the Ministry of Health, Labor and Welfare, Japan Water Research Center studied a wide-area backup system that enables water supply to be controlled and managed beyond the service area of each waterworks utilities in such an emergency. This paper introduces the results of studying wide-area backup systems in large urban areas, and discusses the effects and roles of the backup system. 237

256 2. Concept of Wide-area Backup (1) Necessity Wide-area cooperation is required to ensure the supply of large volumes of water in an emergency such as an earthquake, drought, or water quality incident. For example, the operation of water systems in an earthquake is different from that during ordinary times, as a large quantity of water must be supplied for recovery work, and water leakage during supply is increased. Although individual water utilities cannot easily store a large quantity of surplus water due to limitations on volume and cost, the necessary investment is small and the surplus water can be stored more easily if multiple utilities share surplus water. During the Hanshin-Awaji Earthquake, a seismic intensity of 7 was recorded in the limited Disaster Band area measuring 1 km 25 km. If a wide-area cooperation system had been established, cooperation could have been obtained from the non-affected area. (2) Definition Wide-area backup is defined in this paper as water exchange among water utilities. This is done among water utilities within a prefecture whereas water accommodation in this paper means that spanning prefectural boundaries. The concept of wide-area backup is shown in Fig. 1. Water utility Water utility Waterworks Waterworks Waterworks Waterworks Waterworks Trunk facility Trunk facility Trunk facility Trunk facility Trunk facility :Water pipe :Wide-area backup connection :Prefectural connecting pipe Prefectural boundary Fig. 1 Concept of Wide-area Backup 238

257 3. Expected Effects The expected effects of a wide-area backup system are described below. (1) Effects of Securing Water in a Disaster People feel safer if domestic water is secured. The water is also used for emergency purposes and recovery work, assisting the recovery effort and emergency water supply operations. It is also used for extinguishing fires. (2) Effects of Quicker Recovery The availability of recovery water enhances the efficiency of the initial recovery work, thus reducing the recovery time by 5 to 10 days and curbing socioeconomic losses. One way to quantify socioeconomic losses is to assign a value to water at differing points during the recovery work. Those values are used to calculate the dissatisfaction or inconvenience that is caused by loosing its purchasing opportunity and it is converted into a cost. (3) Effects of Wide-Area Construction Constructing an efficient system that eliminates prefectural boundaries assures availability over a vast area, reduces differences in water service levels among localities, and raises waterworks systems in the entire Metropolitan area to a consistent level. (4) Effects of Effective Utilization of Water Resources By using water that is stored as surplus during ordinary times, existing water can be used more efficiently, thus reducing the need to develop new water resources. (5) Other Effects In situations other than earthquakes where backup is necessary such as in drought or problems with water sources or quality, additional water can be supplied from different water systems. 4. Analysis of Plan for Constructing a Wide-area Backup System (1) Case Study for the Metropolitan Area A case study under certain assumptions was performed by questionnaire survey of water utilities for the Metropolitan area (one metropolis and three prefectures). One prefecture was assumed to be struck by a disaster and the other three supported the affected prefecture. 1 Assumptions a. Both inland and ocean-trench earthquakes may hit each prefecture. We assumed that an inland earthquake strikes areas with many trunk facilities, and that trunk facilities lose 50% of their capacity. 239

258 b. The prefectures estimate that the suspension rate of water supply will be 10% to 30% for an inland earthquake and 30 to 50% for an ocean-trench earthquake. It is set that 20% (average suspension rate of water supply for inland earthquake) as the lower limit and 50% (maximum suspension rate of water supply for ocean-trench earthquake) as the upper limit, and calculated the quantity of water required for these lower and upper limits in a disaster. c. Purified water will be accommodated to the other purification plant because the receiving party must process the water if raw or industrial water is supplied. Connection between purification plants is assumed. d. Quantity of aid water The quantity of aid water is calculated by the following equation: Maximum quantity of aid water = Plant capacity α daily maximum water supply From this calculation, each prefecture can supply 10 to 15% of its daily maximum water supply. Coefficient α is 0.9 to take account of loss during water purification. 2 Required Backup Water The required backup water is the difference between the required amount and the capacity of the plant, which has been reduced by the damage. The required water in a disaster is the sum of the ordinary supply water to non-affected areas and emergency and recovery water supply to the affected areas. Required backup water = Required water in a disaster plant capacity in a disaster = (Recovery water + emergency water + ordinary water supply) (existing plant capacity reduced plant capacity due to damage caused by the disaster) The relationship between water supply and plant capacity is shown in Fig. 2 for both ordinary and disaster times. The followings are based on actual data for the Hanshin-Awaji Earthquake: Recovery water requirement: 16 times the ordinary leakage (8% of daily average water supply) Emergency water requirement: 20 liters/day/person is set as the target water requirement for 10 days from the disaster. 3 Backup Availability The result of the survey revealed that it is possible to support approximately maximum daily water supply when 20% of water supply is suspended if the margin of existing facilities is effectively used in the Metropolitan area. In case of 50% water supply is suspended, up to 85% of the daily maximum water supply can be provided. 240

259 Ordinary times V olum e Water Supply Volume Plant capacity Margin In a disaster Emergency water Volume( by use) Recovery water Ordinary water Required backup water Volume (backup) Self-supplied water Plant capacity Plant capacity in a disaster Decrease in capacity Fig. 2 Required Backup Water (2) Wide-area Backup System Construction Plan Two draft plans are proposed: Plan A: Prefectures are connected by a single route supplying backup water. Plan B: Prefectures are connected by two routes supplying backup water. For Plan A, a two-stage phased construction is used to minimize the construction cost: Stage 1: The four prefectures are interconnected by the shortest route. Stage 2: Facilities that are unlikely to be simultaneously damaged in an earthquake are connected. (100% of the required backup water in each of both stages) For Plan B, a three-stage phased construction is envisaged to minimize the construction cost: Stage 1: The four prefectures are interconnected by the shortest route. Stage 2: Facilities that are unlikely to be simultaneously damaged in an earthquake are connected. (50% of the required backup water in each of both stages) Stage 3: The second route is constructed to exchange 100% of the required backup water using multiple routes. (3) Estimation of Construction Cost The estimated construction cost for both plans is shown in Table 2. To most effectively use the surplus water held by each prefecture, water supply on the receiving side is the daily maximum water supply per person and the suspension rate of water supply is set at 20%. The connecting pipe flow rate for all conceivable earthquake cases was calculated and the pipe size was determined assuming the in-pipe flow velocity to be 241

260 1.5 m/s at the maximum flow rate. The necessity of pumps and their specifications (head) are set based on the water level of clean water reservoirs in the purification plant. Table 1 Case Setting Stages of Constructing Connecting Routes for Wide-Area Backup System Draft plan Exchange water Tokyo Kanagawa Tokyo Saitama Tokyo Chiba Saitama Chiba A 1 100% of requirement Nagasawa (To) Nishi-Nagasawa Asaka Okubo Misato (Shin-misato) Kita-chiba Shin-misato Kita-chiba B 2 100% of requirement Nagasawa(To) (Nishi-Nagasawa) Sagamihara Higashi-murayama (Asaka) Okubo Misato (Shin-misato) Showa Misono (Misato) (Shin-misato) Kita-chiba 1 50% of requirement Same as A % of requirement Same as A % of requirement Nagasawa (To) Higashi-murayama Misono (Nishi-Nagasawa) (Asaka) Okubo (Misato) Sagamihara Kanamachi (Shin-misato) Sakai (Misato) Kita-chiba Nishi-Nagasawa (Shin-misato) Showa Kanamachi (Kuriyama) Kashiwai Showa (Shin-misato) Kita-chiba Showa (Shin-Misato) Kita-chiba Showa (Shin-misato) (Misato) (Kanamachi) (Kuriyama) Kashiwai Gyoda Purification Plant Saitama Prefecture Ozaku Purification Plant Okubo Purification Plant (Saitama) Higashimurayama Purification Plant ) Metropolis of Tokyo Shin-misato Asaka Purification Plant Purification Plant (Saitama) (Tokyo) Sakai Purification Plant (Tokyo) Nishi-Nagasawa Purification Plant (Kanagawa) Showa Purification Plant (Saitama) Misono Purification Plant (Tokyo) Nagasawa Purification Plant (Tokyo) Kanamachi Purification Plant (Tokyo) Kita-chiba Purification Plant (Chiba) Misato Purification Plant (Tokyo) Kuriyama Purification Plant (Chiba) Chiba Prefecture Kashiwai Purification Plant (Chiba) Kanagawa Prefecture Isehara Purification Plant Sagamihara Purification Plant (Kanagawa) Samukawa Purification Plant Shiomidai Purification Plant Nishitani Purification Plant Kosuzume Purification Plant Legend: : Major purification plant : Wide-area backup connecting pipe (stage 1) : Wide-area backup connecting pipe (stage 2) : Wide-area backup connecting pipe (stage 3) Fig. 3 Prefectural Connecting Routes in Draft Plans 242

261 Table 2 Estimated Construction Cost for Each Case (unit: million yen) Case Plan A Plan B Section Stage 1 (total quantity) Stage 2 (total quantity) Stage 1 (half quantity) Stage 2 (half quantity) Stage 3 (total quantity) Tokyo Kanagawa 10,960 94,560 7,140 60,640 47,320 Tokyo Saitama 41,260 98,390 20,060 47,650 18,350 Tokyo Chiba 8,890 38,700 5,790 23,860 23,020 Saitama Chiba Total for each construction stage 61, ,650 32, ,150 88,690 Ratio to B Total for each plan 292, ,830 (4) Reduction of Construction Cost Plan B is better than Plan A because of its low construction cost and flexible operation due to the connections among prefectures with multiple routes. However, the immense construction cost for Plan B, which is 250 billion, was tried to be reduced while maintaining its effectiveness. The construction cost may be reduced by setting backup water requirements as shown in Table 3 (first reduction measure) and by using in common the planned pipes of water utilities for some sections of the pipe route (second reduction measure). Table 3 Reduction of Construction Cost Original plan First reduction measure Second reduction measure Plan and reduction measures Optimal plan (Plan B) Set backup water requirements Clarification of range of construction of wide-area backup system Description of the plan Connect prefectures with two routes using purification plants as the starting point Of the two routes in the original plan, use the shorter one to connect prefectures One half of the above connecting pipe length is assumed to be used in common with the planned pipes in the existing water system projects of water utilities responsible for the purification plants that are designated as the starting point of connection. Percentage of required backup water Estimated construction cost 100% 253,830 million 50% 165,140 million (corresponds to Plan B-2) 65.1% of the original plan 50% 123,860 million (165,140 ( /2)) 48.8% of the original plan 243

262 The first reduction measure reduces the required backup water to 50% and part of the water supplied to the ordinary water supply area is sent to the affected area in order to fully meet its requirement. This means that water supply restriction is applied in the ordinary water supply area. The allowable limit of water supply restriction was set at 75% of the daily average water supply for the non-affected areas. If a load factor is 0.85, the ordinary water supply rate is 0.88 (= 0.75/0.85). Seventy-five percent of the daily average water supply is achieved even when the required backup water is reduced to 50% in this study, as shown in Fig. 4. Fig. 4 Relationship between Backup Water Supply Achievement Rate and Ordinary Water Supply 1 Ordinary water supply rate 通 常 給 水 率 ハ ックアッフ Backup 水 量 確 保 率 water supply achievement rate Case 東 京 都 1 神 Case 奈 川 1 県 ケース1 Tokyo Kanagawa ケース1 埼 Case 玉 県 1 ケース1 Saitama 千 Case 葉 県 1 ケース1 Chiba 東 Case 京 都 2 神 Case 奈 川 2 県 ケース2 Tokyo Kanagawa ケース2 埼 Case 玉 県 2 ケース2 Saitama 千 Case 葉 県 2 ケース2 Chiba Case ケース1: 1: Ordinary 通 常 water 給 水 supply を 一 日 at 最 the 大 給 daily 水 量 maximum 断 水 率 water 20%の supply, 場 合 and 20% water stoppage Case ケース2: 2: Ordinary 通 常 water 給 水 supply を 一 日 at 平 the 均 給 daily 水 量 average 断 水 率 water 50%の supply, 場 合 and 50% water stoppage As the second reduction measure, the planned pipes of water utilities are used in common with part of the pipe connections. Although it is impossible to determine the appropriate ratio of common-use pipes, the total cost of constructing connecting pipes is certainly reduced. Using one half of the total length of connecting pipes in common with water utilities, the cost of common-use pipes is halved between both parties so that each party benefits from lower construction costs. 244

263 (5) Effects and Evaluation of Draft Plans Figure 5 shows the relationship between construction level and effect. Point 1 indicates that the construction level is satisfied but the construction effect is low. Point 2 is where both construction level and effect are satisfied. Point 3 is where neither construction level nor effect is satisfied. Point 4 shows the construction level is low but construction effect is high. By moving from points 3 to 4 to 2 in this order in stages, the construction level can be gradually enhanced while confirming the construction effect, thereby avoiding over-investment. Two indicators to gauge the effect of the wide-area backup system are 1. Construction Effect and 2 Construction Level. Value of [Construction Level] 整 備 水 準 の 値 1 3 一 定 水 準 の Reasonable 効 果 あり effects 4 Value of 整 [Construction 備 効 果 の 値 Effect] 2 Reasonable 一 定 の construction 整 備 水 準 level 1 Construction Effect Fig. 5 Relationship between Construction Level and Construction Effect Construction effect is evaluated by comparing differences in an amount of loss between executing and not executing the wide-area backup construction project with the estimated project cost. An indicator to evaluate the level of construction contents must be established to facilitate evaluation of the construction plans as well as evaluation of the construction effect. This indicator is given by; Construction effects = [Differences in an amount of loss between executing and not executing the wide-area backup construction project] ([service period] [risk probability])/[total project cost] Equation A where, Service period: 40 years, assuming that the pipes are the major facility Risk probability: A risk event is hypothesized to occur once every 100 years. 2 Construction Level The construction level is evaluated by the suspension rate of water supply in the case of an earthquake, the rate of successfully acquiring the required backup water, and the rate of securing ordinary supply water. The product of these factors indicates the construction level: Construction level = [Estimated size of risk] [Quantified satisfaction] = [Suspension rate of water supply in an earthquake] [rate of successfully acquiring the required backup water] [rate of securing ordinary supply water] Equation B 245

264 By using two factors above, the metropolitan construction plan and an alternative plan with better cost performance were evaluated. The results are summarized in Table 4. Note that this equation represents only part of the construction effects (effects that are difficult to quantify are excluded). Table 4 Construction Effects and Construction Level of Metropolitan Area Construction Plan Original plan Downsizing of construction plan (backup water requirements are reduced to 50% of the original plan) Same as above (water utilities share the cost of pipes that can be used for emergency) Construction effect (by equation A) Construction level (by equation B) The original plan is the base ( = 0.425: standard value) 300 billion (40 years/100 years)/ 250 billion = billion (40 years/100 years)/ 170 billion = billion (40 years/100 years)/ 120 billion = % of the original plan (0.5 x = 0.213) Same as above 5. Problems of Wide-area Backup System Project The problems of the wide-area backup system projects are summarized below based on the case studies for the Metropolitan area considering the current situation in other large urban areas. (1) The Meaning of the Wide-Area Backup System and Consensus for its Necessity Waterworks utilities, prefectural waterworks administrations, central government, and users need to be able to clarify the meaning of the plan at their own level. It is important that all these parties agree that a wide-area backup system is necessary. Also, the role of each party needs to be clarified. (2) Validity of Conditions such as Backup Water Requirement It is necessary to hypothesize scenarios in which the system will be necessary, clarify the basis of estimating casualties, state the assumptions, and discuss and agree on the validity of requirements including the quantity of water to be supplied. 246

265 (3) Development of Rational Construction Plan The present and future construction programs must be clarified. The plan should allow for intermediate reviews. The plan must be compatible with the construction plan of the water utilities (plan for wide-area and aseismic design). It is necessary to agree on cost sharing, disclose expected revenue sources, and clarify the effects on financing and costs. The plan must also take into account the operation, management, and other intangible aspects. (4) Appropriate Evaluation of Construction Plans In addition to a general evaluation, appropriate evaluation methods must be established for individual localities, and evaluation criteria must be defined. (5) Establishment of Project Execution Procedures Project execution procedures must be established as well as basic and execution plans based on those procedures. (6) Consolidation of Regulatory Laws and Systems The possibility of using water rights for other than the designated purpose and subsidy systems for transboundary wide-area projects should be studied. 6. Greater Efficiency The results of our case study revealed various problems in building a wide-area backup system, but significant effects would be gained. The system should be constructed not by individual entities, and the project should be compatible with the construction (wide-area and quake-resistant) plans of water utilities. It is also important to clarify cost sharing and available funds for the project, and to carefully study financing and cost aspects. If the necessary conditions are met, a wide-area backup system in the Metropolitan area would produce certain effects. Similar case studies are currently being conducted in the Kinki area and further detailed investigations are under way. Water utilities will need to improve the efficiency and security of their operations in the future. It is hoped that this case study not only suggests wide-area backup possibilities in a disaster, but also encourages practical, effective use of water during ordinary times. It is believed that water utilities will be able to manage their water supply businesses more efficiently and securely in the future. 247

266 248

267 3 rd US-Japan Workshop on Water System Seismic Practices TECNICAL SESSION VII Seismic Resistant Design Reinforcement Work of the Embankment of the Yamaguchi Reservoir Presenter: Hiroshi Yamada (Waterworks Bureau Tokyo Metropolitan Government, Japan) Seismic Upgrade of East Bay Municipal Utility District's Mokelumne No. 3 Aqueduct Presenter: Bruce Maison (East Bay Municipal Utility District, Oakland, California) Seismic-Proof Design for the Structures of the Sinanogawa Water Treatment Plant in Niigata Presenter: Hitoshi Hasegawa (Niigata Waterworks Bureau, Japan) A Practical Approach to Mitigation of Earthquake Pipeline Damage Presenter: William Heubach (Seattle Public Utilities, Washington) 249

268 3 rd US-Japan Workshop on Water System Seismic Practices 250

269 The Reinforcement Work of the Embankment of the Yamaguchi Reservoir Hiroshi Yamada and Isao Tahara ABSTRACT The Yamaguchi Reservoir, which the Tokyo Waterworks Bureau manages, is one of the greatest earth-fill dams in Japan completed in The available storage capacity, with the adjoining the Murayama Reservoir, is equivalent to the amount of water used in Tokyo in one week. Moreover, since water conveyance from the Tama River and raw water transmission from the reservoir to water purification plants can be performed by using gravitational force, it is also a very important facility as a measure against earthquake disasters. After The Great Hanshin Earthquake (1995), the design standards of various structures were re-examined, and higher reliability of the important facilities came to be searched for. As a result of seismic resistant analysis, it became clear that the embankment of the Yamaguchi Reservoir will be damaged when a level 2 earthquake occurs. We decided to carry out the reinforcement work of the embankment in consideration of the importance of the Yamaguchi Reservoir and the present condition that urbanization has advanced to the embankment. Based on the present condition of the Yamaguchi Reservoir, five methods of reinforcement construction of the embankment were devised in order to acquire the ability to respond to level 2 earthquake motions by slight repair. After performing the comparative examinations, it was judged that the holding banking + inclined and level drains at down-stream method of construction is comprehensively excellent, and the target seismic resistant ability would also be satisfied. Therefore, this method of construction was adopted. The construction was done in the following procedures. (1) Preparatory Construction, (2) Removal of Established Toe Weight, (3) Basic Excavation, (4) Reinforcement by Banking and Making the Drains, (5) Slope Protection. Moreover, the local procurement of banking material, effective use of removed material, protection of rare wild animals, etc. were realized. Hiroshi Yamada, Director of Water Supply Section, Chuo Branch Office, Bureau of Waterworks, Tokyo Metropolitan Government, Uchi-Kanda, Chiyoda, Tokyo, , Japan Isao Tahara, Chief of No.2 Design Team, Design Section, Construction Division, Bureau of Waterworks, Tokyo Metropolitan Government, Nishi-Shinjuku, Shinjuku, Tokyo, , Japan 251

270 INTRODUCTION The Yamaguchi Reservoir (commonly known as; Lake Sayama) is a reservoir only for water supply located in the Sayama Hills, which spreads in Tokorozawa City and Iruma City in Saitama prefecture. It is a man-made lake built during the period of 1927 to The available storage capacity is 19,530,000m³. The crest length of the embankment is 691m, the height is 35m, and the width of the top is 7.3m. As these data show, the Yamaguchi Reservoir is one of the greatest earth-fill dams in Japan. In the Taisho era, when the Yamaguchi Reservoir was planned, Tokyo depended on the Tama River for its water supply. However, the running water from the Tama River was not enough at the period of water shortage. For this reason, the Murayama Reservoir (commonly known as; Lake Tama) was built. Nevertheless, the subsequent increase of water demand made it difficult to obtain enough water only from the Murayama Reservoir (when the demand for water was high). As a result, we decided to construct the Yamaguchi Reservoir in The water of the Yamaguchi Reservoir is taken at the Hamura Intake Weir and the Ozaku Intake Weir in the Tama River, and is conveyed to the reservoir through the driving channel. Moreover, the Yamaguchi Reservoir is connected to the adjoining Murayama Reservoir through the water conveyance pipe (caliber; 2,600mm), and both reservoirs are operated as one. The available storage capacity of both reservoirs in all is 35 million m³, which is equivalent to the amount of water used in Tokyo in one week. Therefore, these reservoirs are the precious water jars to citizens in Tokyo. Both reservoirs are the facilities to supply raw water of the Tama River to the Higashi-Murayama Purification Plant and the Sakai Purification Plant. And, the raw water conveyance pipes were built that connect the Higashi-Murayama Purification Plant to the Asaka Purification Plant, which usually use water from the Tone River. We have been making good use of raw water of both the Tama River and the Tone River through these pipes. Usually, we take abundant raw water from the Tone River as much as possible, and store raw water from the Tama River in the reservoirs as much as possible. Then, we use raw water from the Tama River during summer, when water demand is high, and when accidents or water shortage of the Tone River occurs. TABLE Ⅰ. SPEC OF DAMS AND RESERVOIRS the Yamaguchi Reservoir reference the Murayama-kami Reservoir the Murayama-shimo Reservoir location Tokorozawa, Saitama Higashi-Yamato, Tokyo type earth-fill dam earth-fill dam earth-fill dam dam height above ground level 35m 24.2m 32.6m (embankment) crest length 691m 318m 587m volume 1,400,000m³ 333,000m³ 836,000m³ drainage area 7.2km2 1.3km2 2.6km2 reservoir reservoir area 1.89km2 0.41km2 1.11km2 available storage capacity 19,528,000m³ 2,983,000m³ 11,843,000m³ 252

271 Yamaguchi Reservoir Hamura Intake Weir Murayama Reservoir Asaka P. P. Tama River Ozaku Intake Weir Higasi-Murayama Purification Plant Sakai P. P. Figure 1. The information map around the Murayama and Yamaguchi Reservoir As for water conveyance from the Tama River to both reservoirs and raw water transmission from the reservoirs to each water purification plant (Higashi-Murayama, Sakai, Asaka and Misono), and purified water transmission from the Higashi-Murayama or the Sakai Purification Plant to the Tokyo metropolitan areas, all can be performed by using gravitational force. Therefore, they are very important facilities to secure the water supply at the time of an earthquake and a power failure. reinforcement for seismic resistance of the facilities reinforcement construction of the facilities reinforcement of the water supply systems facilities of water storage,intake, and conveyance purification facilities water transmission and distribution facilities assesment of seismic resistance and repair water supply equipment water transmission and distribution facilities purification plant water transmission and distribution pipe water supplying station service pipe water transmission pipe reinforcement of embankments of reservoirs preparation of private power plants reconstruction of chlorine feeding facilities replacement of transmission and distribution pipes expansion of water supplying station seismic diagnosis of facilities repair for seismic resistance of facilities reinforcement for seismic resistance of service pipe replacement of service pipes by stainless steel pipes establishment of network of transmission pipes Figure 2. The systems of the project plan about measures against an earthquake disaster 253

272 Figure 3. The overview of the Yamaguchi Reservoir before the reinforcement construction DETAILS ABOUT THE PROJECT In Japan, after the Great Hanshin Earthquake occurred in Kinki district in 1995, the design standards of various structures were re-examined, and various reinforcement constructions against earthquakes have been carried out. Similarly, the design standards of water supply facilities were re-examined. Higher reliability of the important facilities came to be searched for. Tokyo Waterworks Bureau has done the investigations on the seismic resistance of all the facilities, and re-examined the plans to reduce damages by earthquakes. Moreover, reinforcements for seismic resistance of facilities and water supply network have been carried out. Reinforcement for seismic resistance of water supply facilities Tokyo Waterworks Bureau drew up the project plan about measures against an earthquake disaster, which is a three to four year long plan, in order to minimize the damage of the water supply facilities by earthquakes and to secure the water supply to citizens in Tokyo as much as possible. And reinforcement constructions for seismic resistance of water supply facilities are being carried out according to that plan. According to this project plan, the measures against an earthquake disaster are divided into two systems. One is the reinforcement constructions of the facilities to reduce damages in case of an earthquake disaster, the other is the reinforcement of the water supply systems to reduce the shutdown area and shutdown time. As for the 254

273 reinforcement constructions of the facilities, the water supply facilities are classified according to their functions. And the reinforcement works of the facilities for seismic resistance are carried out accordingly. On the other hand, the reinforcement of the water supply systems is to reinforce the function of the whole water supply systems. To put it concretely, establishing the water transmission and distribution network and establishing block distribution system are underway. Enforcement of the reinforcement work of the embankment of the Yamaguchi Reservoir Because Tokyo Waterworks Bureau considers the Yamaguchi Reservoir as one of the most important facilities, a seismic resistant analysis was performed in Consequently, it became clear that the embankment of the reservoir would be deformed and damaged if a level 2 earthquake (for example, the Minami-Kanto strong local earthquake, which registered magnitude 7) occurs. Moreover, water conveyance from the Tama River and raw water transmission to each water purification plant (Higashi-Murayama, Sakai, Asaka and Misono) can be performed by using gravitational force as mentioned before. Therefore, the position of the Yamaguchi Reservoir is very important in terms of reinforcement of the water transmission and distribution network. After all, reinforcement of the embankment of the Yamaguchi Reservoir is considered as an extremely important measure against earthquake disasters. It is because the measure is related to not only reinforcement of the water storage facilities to reduce the damage but also securance of function of the water supply system in case of an earthquake disaster. In consideration of the position as the measure against earthquake disasters and the present situation of the advance of urbanization to the embankment during about 70 years after completion, we decided to carry out a reinforcement work of the embankment in order to raise seismic resistant ability. DESIGN The features of the established embankment of the Yamaguchi Reservoir The established embankment is divided into three zones; core, shell, and toe weight at down-stream. According to the record at the time of construction, cohesive soil and gravel were used for the core, loam for the shell, and loam and grit for toe weight at down-stream. Although we can divide the established embankment into three zones according to the materials, permeability coefficients of each zone are equal. So, the embankment may be categorized as a homogeneous earth-fill dam in terms of its function. In addition, N value is about ten at the core, about five at the shell. Besides, bulletproof stratum to protect the embankment from bombing was built on the top of the embankment during World WarⅡ. However, they didn t record the processes of the construction at all reflecting the social situation of those days. 255

274 上 流 upper stream 上 流 コンクリート ブロック concrete blocks C It Yg Yc 耐 弾 層 bulletproof stratum H.W.L T.P :3.0 上 流 サヤ shell 土 T.P コンクリート watertight concrete 止 水 wall 壁 embankment s 既 設 堤 体 ダム axis 軸 core 心 壁 T.P.+119~ 下 流 サヤ 土 下 流 下 流 既 設 抑 え 盛 土 35m toe weight at down-stream T.P T.P F Ig Is It 182m Yg Figure 4. The standard sectional view of the established embankment of the Yamaguchi Reservoir shell 1:2.5 down-stream 1:3.0 Study of the reinforcement methods of the embankment In designing, we considered the present condition of the Yamaguchi Reservoir. We aimed at keeping the embankment safe against level 2 earthquakes. Also, we tried to secure human lives and properties. And, we intended to maintain the function of the embankment as a waterworks facility even if a big earthquake occurs. Target seismic resistant ability The seismic resistant design was based on the Seismic Design and Construction Guidelines for Water Supply Facilities [1], the Cabinet Order concerning Structural Standards for River Administration Facilities etc. [2], the Manual for River Works in Japan by Ministry of Construction [3]. As a result, we decided to design the embankment by seismic coefficient method, and check the motion of the embankment in case of earthquake by FEM dynamic analysis. As for design safety factor, we decided that safety factor for slide by the circular arc method should be over 1.2, the design seismic intensity 0.2, in consideration of the type of the dam, the condition of the basic foundation, and the figures on other established dams. Moreover, we decided the safety factor for slide of the embankment in dynamic analysis should be over 1.0. And, we decided residual deformation should be deformation which does not need repair against level 1 earthquakes, and deformation which needs slight repair against level 2 earthquakes. Selection of the reinforcement method of the embankment Through seismic resistant analyses, it became clear that there was no problem about the strength and watertightness of the base of the established embankment. But there were problems about the stability caused by the small of sectional area and high wet surface. So, about the reinforcement method of the embankment, five methods of construction were devised to reinforce the stability by increasing the sectional area and lowering the wet surface. And comparative examinations were performed. We paid our attentions to the following factors. 1. The material in the reservoir can cover the volume of banking. 256

275 2. The embankment does not lose the function even if it is heavily deformed. 3. It should be free from periodical maintenance that requires falling of water level. 4. Special methods of construction should not be used. 5. The construction period should be comparatively short. 6. The construction should be economical As a result of comparative examinations, it was judged that the holding banking + inclined and level drains at down-stream method (Proposal D) is comprehensively excellent and suitable as the reinforcement work. Proposal A : simple holding banking A 案 : 単 純 抑 え 盛 土 reinforcement banking established embankment reinforcement banking B 案 : 表 面 土 質 遮 水 壁 型 Proposal B : water insulation wall of soil reinforcement banking drain established embankment reinforcement banking C 案 : 表 面 人 工 遮 水 壁 型 Proposal C : artificial water insulation wall asphalt water insulation wall drain reinforcement banking established embankment reinforcement banking D 案 : 単 純 抑 え 盛 土 + 傾 斜 及 び 水 平 ドレーン Proposal D : holding banking + inclined and level drains at down-stream Inclined drain reinforcement banking estalished embankment reinforcement banking level drain E 案 : 単 純 抑 え 盛 土 + 鉛 直 ドレーン Proposal E : holding banking + vertical drain vertical drain reinforcement banking established embankment reinforcement banking Figure 5. The examined reinforcement methods of the embankment 257

276 TABLE Ⅱ. COMPARISON OF THE REINFORCEMENT METHODS OF THE EMBANKMENT compared factors reinforcement method of the embankment Proposal A Proposal B Proposal C Proposal D Proposal E securance of banking material following ability against deformation easiness of inspection and repair reliability of construction construction period economical comprehensive estimation Examination by reference of the seismic resistant ability in dynamic analysis Input earthquake motion According to the Seismic Design and Construction Guidelines for Water Supply Facilities [1], the followings are defined. Earthquake motion which occurs once or twice during the period of the use of the facility is level 1. And, big earthquake motion which has low probability of occurrence is level 2. Taking into account the Guidelines, past literatures, results of earthquake hazard analyses, and the existence of the Tachikawa Dislocation (The Quaternary dislocation) near the reservoir (about 5km southwest), the data of three kinds of earthquake motions were set up as follows. 1The Ansei Edo Earthquake (M6.9, reappearance probability; about 1/30 years, local type) is level 1 earthquake motion, 2 The Minami Kanto Earthquake (M7.9, reappearance probability; about 1/200 years, trench type), 3 The Tachikawa Dislocational Earthquake (M7.1, reappearance probability; about 1/5000 years, near local type) are level 2. The results of analyses Two kinds of analyses (slide analysis and calculation of residual deformation based on the accumulated damage index theory) were performed. As a result of the slide analyses, the minimum values of the safety factor for slide of the embankment were more than 1.0 in all cases of the hypothetical sliding surfaces based on Proposal D. So, we judged that any sliding failure would not occur. Moreover, the amount of settlement at the top of the embankment calculated by residual deformation analysis based on the accumulated damage index theory was about 60mm at the maximum in case of a level 1 earthquake motion, and about 100mm at the maximum in case of level 2. Therefore, it was judged that these deformations are within the grade that does not need any repair in level 1, and that we can respond to with a slight repair in level 2. From these results, we judged that the holding banking + inclined and level drains at down-stream method (Proposal D) satisfies the requirements on seismic resistant ability. 258

277 上 流 耐 (removal) 弾 層 ( 撤 去 ) コンクリートブロック( concrete block (removal) 撤 去 ) upper-stream 上 流 補 強 盛 土 reinforcement banking 1:3.8 bulletproof stratum 新 ダム 軸 既 設 ダム 軸 T.P T.P T.P T.P :3.0 shell 心 core 壁 上 流 サヤ 土 下 shell 流 サヤ 土 下 流 補 強 盛 土 inclined drain 傾 斜 ドレーン 1:3.9 T.P 既 toe 設 weight 抑 え 盛 (removal) 土 ( 撤 去 ) 1:3.0 下 流 T.P It It Ig Is 45m( 上 流 補 強 盛 土 ) Yg 182m( 既 設 堤 体 ) Yg It Yg 57m( 下 流 補 強 盛 土 ) 35m 既 設 堤 体 established embankment reinforcement banking 12m 1:2.5 down-stream 水 level 平 ドレーン drain Figure 6. The standard sectional view of the reinforcement work of the embankment CONSTRUCTION The procedures of the constructions Preparatory construction Before the reinforcement work of the embankment, a double cofferdam was established at upper-stream of the embankment in order to secure the work yard for banking in the reservoir. Moreover, in consideration of animals and plants, such as waterfowls depending on the reservoir, the temporary cofferdams in order to secure the water surface were established at two places of upper-stream of the reservoir. About 70 years had passed since the completion of the Yamaguchi reservoir. So, after draining water within the cofferdam, it became clear that soft earth and sand (water content 150% - 250%) were piled up. It was two meters deep at the bottom of the lake. And any man could not walk on. In order to use this part as a work yard, Soil stabilization of the soft deposits was carried out in the original position so that heavy equipment could move around on that place. Removal of established toe weight at down-stream and bulletproof stratum Before banking, the toe weight at down-stream (which was tightened loosely) and bulletproof stratum (which was built in order to reduce damages from bombing during the war) were removed. Basic excavation As the basic foundation of banking, excavation was carried out to the level of the basis of the established embankment. 259

278 Reinforcement by banking and Making the drains Reinforcement by banking at upper-stream and down-stream (upper-stream side; 530,000m³, down-stream side; 440,000m³) was performed. Banking material was carried by damp trucks (10ton burden) out of a temporary place, made even by 21ton-class bulldozers, and tightened by 10ton oscillating rollers. Moreover, the inclined drain and the level drain were constructed between the established embankment at down-stream and the new embankment. The recycling material removed from the bulletproof stratum was used effectively for drain material. Slope Protection of the embankment Slope Protection was carried out through setting up concrete blocks at upper-stream and laying turf at down-stream. The features of the construction plan Avoiding deforestation and securing the construction space In order to preserve the surrounding residential area and the rich natural environment around the reservoir, we decided to supply banking material below the full water level without new deforestation, in accordance with the principle of local procurement. So, all the construction spaces were secured in the place which used to be underwater. And, we temporarily placed the banking material at the places where we had gathered and threw away materials. Effective use of the removed material from the established structure The grain size of a lot of removed material from the established structure was adjusted in the aggregate plant which is set up in the working area. And the removed material was made good use as the material for the level and inclined drains and for the roads within the working area. In this way, we completed the construction without discarding soil. limit 森 林 of 境 forest 界 high 常 時 water 満 水 位 level (H.W.L) T.P m 付 近 1:1.2 the site to gather banking material 土 取 場 during construction 1:1.2 施 flood 工 中 洪 season 水 時 水 control 位 T.P+96.1m level 付 近 施 工 中 ordinary 運 用 水 位 water T.P+94m level 付 近 Figure 7. The relation between water storage level and the site to gather banking material. 260

279 CONSIDERATION FOR NATURAL ENVIRONMENT The measures of preservation of goshawks In the Sayama Hills around the Yamaguchi Reservoir, rich nature is kept and commonly called the Woods of Totoro. And, the circumference of the reservoir is designated as the only special wildlife sanctuary in Saitama Prefecture. So, an environmental assessment was carried out in consideration of the natural environment and the scale of this reinforcement work. As a result, nest building of goshawks, designated as the domestic rare wild animal by the law about preservation of wild animals and plants with fear of extinction, was confirmed around the reservoir. Therefore, subsequent surveillance and measures of preservation as follows were performed continuously. Since we have taken these measures, the goshawks continue to inhabit in the forest around the reservoir in the same way as before construction, and breeding is also confirmed every year. 1. Making a change in the time zone of construction 2. It became clear that the construction schedule of the temporary cofferdams at upper-stream felled on the nest building period when they become most cautious. Therefore, the construction of the temporary cofferdams was carried out not during the day but at night because goshawks are diurnal. 3. Using low noise and low vibration machines 4. Consideration of the direction of lighting 5. It devised so that the lighting under construction might not turn toward the nesting place of goshawks. 6. The reduction of the noise made by metals In order to reduce the noise made by metals, the shock absorbing material made of rubber was inserted between the parts of heavy equipment, and between iron plates laid out to secure trafficability, and so on. Securance of water surface with the temporary cofferdams As mentioned before, two temporary cofferdams at upper-stream were established so that activities of animals and plants, such as waterfowls depending on the waterside or the water surface of the Yamaguchi Reservoir, might not be obstructed as much as possible. Thereby, we increased the water surface by 150,000 m2, and the water surface became 300,000 m2. It is twice as large as the original plan. And, the number of the waterfowls which come to the Murayama Reservoir and the Yamaguchi Reservoir was about 1,200-2,000, which was unchanged as before. Transplantation of animals and plants In the preparatory stage of construction, monitoring investigation to grasp the numbers, the kinds, etc. of animals and plants which inhabit surrounding area was 261

280 performed. And according to the results, transplantation of the Genji Firefly, the Tokyo Salamander, etc. were carried out. CONCLUSION As explained before, enough investigations and detailed construction management were performed. Consequently, during the reinforcement construction of the embankment of the Yamaguchi Reservoir, various problems that we were concerned about were conquered, and neither a situation that was not expected nor a big accident occurred. And, the construction was completed in November Moreover, from the point of view of coexistence with natural environment, we have obtained variable data in carrying out the construction. From a viewpoint of effective use of the existing stocks, it is expected that renewal constructions of the existing facilities will increase as a part of facility maintenance. We will be pleased if the knowledge that we acquired through the reinforcement work of the embankment of the Yamaguchi Reservoir can help those who plan constructions of the same kind which are expected in the future. REFERENCES [1] Japan Water Works Association Seismic Design and Construction Guidelines for Water Supply Facilities [2] Ministry of Construction. "Cabinet Order concerning Structural Standards for River Administration Facilities etc." [3] Ministry of Construction. "Manual for River Works in Japan by Ministry of Construction". [4] Taguchi,Y. and Tahara, I. et al. July The reinforcement construction of the embankment of the Yamaguchi Reservoir, The Dam Digest No.681, The Japan Dam Foundation, pp [5] Nagaoka, T. and Takada, T. et al. March The plan and result of the reinforcement construction for seismic reistance of an earth-fill dam, presented at the Symposium of Construction Technique, JSCE. [6] Fujisaki, K. and Kanbe, T. et al. March The measures for the technical problems and the protection of the environment of the reinforcement construction for seismic reistance of an earth-fill dam, presented at the Symposium of Construction Technique, JSCE. 262

281 SEISMIC UPGRADE OF EAST BAY MUNICIPAL UTILITY DISTRICT S MOKELUMNE NO. 3 AQUEDUCT Bruce F. Maison East Bay Municipal Utility District ABSTRACT The Mokelumne aqueducts are a key supply system conveying water from the Sierra foothills to the San Francisco Bay Area and are vulnerable to damage from earthquakes in the 24-km (15- mile) reach located in the Sacramento-San Joaquin River Delta. Recognizing this danger, East Bay Municipal Utility District (EBMUD) is undertaking a $39 million project to upgrade Aqueduct No. 3. Included are strengthening of levees at three pipeline river crossings, reinforcing of pipe joints for 8 km (5 miles) of buried aqueduct, and upgrading pipe supports for 14 km (9 miles) of elevated aqueduct. The elevated aqueduct retrofit involves the innovative use of elastomeric seismic isolation bearings to mitigate ground shaking effects. This paper describes the project. INTRODUCTION EBMUD is located in the east portion of the San Francisco Bay Area, serving water to over 1.3 million people. The water system includes a network of reservoirs, aqueducts, treatment plants, and other distribution facilities stretching from the Sierra foothills to the Bay Area (Figure 1). The service area includes 20 incorporated cities and 15 unincorporated communities. The main source of water is the Mokelumne River watershed in the Sierra foothills located about 145 km (90 miles) northeast of the Bay Area. Mokelumne River water is collected in the Pardee Reservoir and flows from Pardee through the Mokelumne aqueduct pipeline system to the Bay Area. The Bay Area has high seismic risk, with numerous earthquake faults throughout the region (Figure 2). Studies revealed that the aqueduct system is vulnerable to damage in the San Joaquin-Sacramento River Delta Area (located to the east of the EBMUD service area), possibly causing unacceptable outage duration. Recognizing this danger, EBMUD is undertaking a project to enhance seismic reliability, that is the subject of this paper. AQUEDUCT PIPELINES The aqueducts consist of three parallel large-diameter steel pipelines, constructed at different times in response to the growth of EBMUD s customer base. The newest is Aqueduct No. 3, built in 1963, having a 2.2-meter (87-inch) diameter. It is the primary aqueduct, typically conveying 4.6 cubic meters per second (100 million gallons per day) under gravity flow that is roughly one-half of EBMUD s demand. The aqueduct is capable of larger flows when operating in a pumped mode. Aqueduct No. 3 s recent construction and large capacity make it the best component to upgrade. Aqueduct No. 1, built in 1929, is typically 1.5 meters (61 inches) in diameter, and Aqueduct No. 2, built in 1949, is about 1.7 meters (67 inches) in diameter. 263

282 In the event of an earthquake causing an aqueduct system outage (failure of all three pipelines), the EBMUD water system master plan calls for reliance on terminal water storage reservoirs in the service area, in conjunction with customer water rationing, for up to six months while Aqueduct No. 3 is repaired. Once repaired, then supply would be provided by Aqueduct No. 3, operating in a pumped mode, while Aqueducts Nos. 1 and 2 are repaired. The project area consists of a 24-km (15-mile) segment. In this reach, there are three river crossings, 9 km (6 miles) where the aqueducts are buried, and 15 km (9 miles) where the aqueducts are elevated above grade. At the river crossings (San Joaquin, Middle, and Old Rivers; Figure 2), there are levees on each side of the river, and the aqueducts cross below the levees and under the river. Where buried, the aqueducts were built in a standard trench and cover construction mode. The aqueducts are elevated when traversing low-lying islands and are supported by pipe supports having pile foundations at regular intervals (Figure 3). Figure 1: EBMUD Water Supply System Figure 2: Bay Area Earthquake Faults 264

283 HAZARDS AND VULNERABILITIES Figure 3: Aqueduct Conditions In River Delta Potential seismic sources lie to the west of the Delta and include the well-known San Andreas, Hayward, and Calaveras Faults, as well as a lesser-known Coast Range-Central Valley (CRCV) Fault. The CRCV is a blind thrust fault located about 10 km (6 miles) west of the project area and plays the dominant role on the site seismicity because of its close proximity. Seismic intensity in terms of peak ground acceleration progressively decreases with position from west to east in the project area. The soils within the upper 15 meters (50 feet) of ground surface generally consist of peat-organic soils, Holocene alluvium, and Pleistocene alluvium. Thicknesses of the various deposits vary significantly over the project area. Peat-organic soils have developed from dense, decaying tule and reedy plants, and are poor engineering materials. Holocene deposits are relatively soft and compressible, whereas the Pleistocene deposits are generally very dense and stiff. Some Holocene deposits contain sand layers and these occur in the immediate vicinity of the rivers, especially near the Old and Middle Rivers. The geotechnical data does not indicate extensive Holocene sand deposits away from the waterways. Analysis suggests that an earthquake with a peak ground acceleration of 0.20 g could cause sand and silty sand deposits in the Holocene alluvium to liquefy, resulting in a loss of soil strength. The dense sands and silty sands in the Pleistocene alluvium are not susceptible to liquefaction. A seismic assessment of Aqueduct No. 3 in the pre-upgrade condition indicated a high probability of widespread damage under a 500-year return period earthquake scenario. Possible damage include slumping of levee embankments leading to failure of the pipeline at the river crossings (as well as flooding of low lying islands), ground settlements underneath the buried pipeline causing separation of pipe segments at the bell-and-spigot joints, and the toppling over of the elevated pipeline due to failure of pipe supports. The aqueduct system outage time duration required to enact repairs or construct a replacement was estimated to be in the range of 1-1/2 to 3 years. This has dire implications since EBMUD s terminal water storage reservoirs in 265

284 the service area are sufficient for only about a 6-month aqueduct system outage (the exact duration depends on season and possible mandated customer water rationing). ALTERNATIVES Planning studies were conducted to determine possible alternative actions and associated costs to serve as a basis of informed decision-making about the seismic mitigation solution. Four alternatives emerged as follows. 1. No seismic mitigation, thereby accepting risk of an earthquake possibly causing unacceptable aqueduct outage duration. Cost to essentially replace an extensively damaged aqueduct under post-earthquake emergency conditions was estimated at $210 to $300 million (1995 U.S. dollars). 2. Upgrade by retrofitting the existing structure for an estimated cost of $39 million. 3. Construct a new replacement aqueduct for an estimated cost of $140 million. 4. Various combinations of alternatives 2 and 3 in which major parts of the existing system are either upgraded or replaced. Combinations evaluated included new tunnels for the three river crossings, and replacing the buried pipe. Estimated costs were $30 million for three new tunneled crossings compared to $19 million for upgrading the existing crossings. Estimated costs were $24 million for replacing the buried pipe compared to $8 million for retrofitting the buried pipe joints. Alternative 2 was selected because it was the least expensive and yet satisfies EBMUD s needs. The project was executed in two phases. PHASE 1 UPGRADES These were completed in 1999 and consisted of upgrading the three river crossings and buried pipeline. Figure 4 illustrates a typical river crossing upgrade having the following features: Lengthening the existing sheet pile wall along the longitudinal axis of the levee as denoted by Label 1. Constructing a shoring and bracing system so that the land-side portion of the pipe can be removed and the pipe foundation conditions improved by installing a pile foundation under the pipe (Label 3). Replacing the pipe with strengthened joints and select backfill around and above the pipe (Labels 2, 5). Flattening the levee land-side slope and adding a toe berm for about 75 meters (250 feet) along each side of the pipe. 266

285 Strengthening the pipe joints for the pipe located below the river and beneath the levee by welding internal saw-tooth patterned butt straps across existing bell and spigot joints (Label 4). The function of the sheet pile wall is to prevent water-side slope failures. It acts as a levee protection barrier that can undergo large plastic deformations and yet prevent breaching of the levee at the aqueduct location. It is expected that the levee water-side slope will be reconstructed in the event of slope movement. The pipe foundation is improved on the land-side of the levee to provide uniform support for the pipe, thereby reducing the potential for long-term total and differential settlements caused by either compression of the Holocene peat-organic soil or by liquefaction of the Holocene alluvium. The addition of internal butt straps increase the strength of the joints, thereby enhancing the ability of the pipe to withstand differential soil settlements. Stability of the land-side levee slopes is significantly improved by constructing a levee toe berm and flattening the levee back slope. Regarding buried pipelines, those portions having high risk of seismic failure due to soil settlements were selected for upgrading. The upgrade consists of welding an internal butt strap across about 700 bell-and-spigot pipe joints. The original joints utilize a single internal fillet weld. The intent is to increase the pipe joint strength to better resist possible differential soil settlement conditions. Figure 5 depicts the buried pipeline upgrade. Figure 4: Phase 1 River Crossing Upgrade Section View Through Levee Figure 5: Phase 1 Pipe Joint Reinforcement Section View at Bell-and-Spigot Joint 267

286 PHASE 2 UPGRADES These are currently under construction and deal with the 14-km (9-mile) elevated portion of the aqueduct. Figure 6 depicts key features of the original (pre-upgrade) structure having basically two types of pipe supports denoted here as temperature blocks and bents. Blocks are spaced about every 300 meters (1,000 feet) along the pipeline. These rigidly hold the pipe above ground and resist seismic forces in all three directions (vertical, longitudinal parallel the pipe run, and lateral perpendicular to the pipe run). Blocks are squat massive concrete structures about 4 meters (13 feet) tall having concrete pile foundations. Concrete was cast directly around the pipeline when originally constructed. Bents are spaced about every 18 meters (60 feet) along the pipeline. These hold the pipe up and resist seismic forces in the vertical downward and pipe lateral directions. They are designed to provide no restraint in the pipe longitudinal direction, thereby allowing pipe thermal expansion/contraction movements. Bents consist of two steel columns (one on each side of the pipe) and a single steel diagonal brace. The columns have ball joints at both their top and bottom connections. The foundation consists of a concrete pile cap situated on top of concrete piles. There are 16 bents located between each pair of blocks. Pipeline expansion joints are located midway between blocks to allow for free axial movements of the pipeline to accommodate thermal effects. Figure 6: Typical Elevated Pipeline Geometry The upgrade design process considered both a Design Basis Earthquake (DBE) having a 500- year return period, and Maximum Considered Earthquake (MCE) having a greater than 2,000- year return period. The blocks in their original condition were found to be capable of undergoing the DBE with only minimal damage to some foundations (some pile cracking and spalling due to bending deformations). For the MCE, about one-half (27) of the total number of blocks would suffer significant foundation damage (more severe pile concrete section deterioration due to bending effects). These are generally located in the western part of the project area experiencing the greatest seismic intensity. Implication of this damage on the stability of individual blocks was difficult to quantify due to uncertainties in actual foundation conditions and analytical techniques (reliability of nonlinear analysis to estimate actual collapse). However, it was clear that the MCE would weaken many blocks and post-earthquake foundation repairs would be necessary. Mitigation would be to upgrade the 27 blocks by the addition of 4 new piles located about the perimeter of the block. The new piles are connected to the existing pile caps by dowels and extension of the concrete pile cap over the top of the new piles (Figure 7). Construction cost was a factor in the decision to upgrade the blocks, and they were 268

287 bid as a separate schedule. After review of bid prices, the decision was made to go forward with the block upgrades. The diagonal brace is the weak link in the original bent structures. DBE shaking intensities are sufficient to fail braces resulting with no lateral support for the pipeline. Subsequent lateral sway-type pipeline collapse is possible. A strategy involved two different types of bent upgrades was developed. Both involve the replacement of the existing braces with new bracing systems (Figure 8). New Pile Cap Extension & Piles Figure 7: Temperature Block Foundation Upgrade By Addition of Piles New Bracing System Elastomeric Isolation Bearing Type A Figure 8: Bent Type A Strong Bracing System and Type B Elastomeric Isolation Bearing System Upgrades Type A upgrades are used in 454 bents located in the eastern part of the project area. The upgrade involves the use of two strong diagonal steel braces connected between the existing columns and the pile cap. The existing columns are also reinforced with steel parts. The intent 269 Type B

288 is to provide a strong rigid steel bracing system capable of resisting lateral earthquake forces. For the DBE, the upgrade would result with no damage to the bracing system and slight inelastic movements in the foundation (soil-pile axial slippage). Under the MCE, similar behavior is expected but with more inelastic action in the foundation (pile axial slippage and some cracking of concrete section). For the western part of the project area having relatively greater seismic shaking intensity, use of Type A supports alone is not viable. A strengthened bracing system would lead to excessive foundation pile damage especially under the MCE (the piles become the new weak link). Bent stability could not be assured without upgrading foundations with more piles. To avoid foundation pile work, a different approach was used for 351 bents located toward the west. Type B upgrades were designed to adequately restrain the pipeline in the lateral direction and yet limit the seismic force transmitted to the foundation. This was achieved through the use of elastomeric seismic isolation bearings. Seismic isolation is a relatively new technology that has been applied to a variety of structures such as buildings and bridges, but the use here is one of the first applications to the retrofit of an existing pipeline. The bearings introduce flexibility and damping to the bents. Increased flexibility alters the pipeline dynamic properties by lengthening the natural period causing reduced seismic force in the system. The dynamic properties are changed such that they are less coincident with the ground shaking characteristics. Increased damping via energy dissipation in the bearings leads to a reduction in seismic forces as well. The foundation is, therefore, protected through the reduction in seismic force. The bearings are rectangular in plan (about 76 cm [30 inches] by 38 cm [15 inches]) and about 30 cm (12 inches) tall, consisting of layers of rubber sandwiched between steel plates. They are rigidly attached to the center of the pile cap and are connected to the existing pipe support columns with new diagonal bracing and horizontal beam. The connections are articulated (can translate and rotate freely) forming a mechanism so that when the pipe displaces laterally, the deformations are taken by the bearing without binding in the steel members. The original pipe support columns have ball joints at both the top and bottom making them well suited for the upgrade. A consequence of seismic isolation (Type B) is that the seismic displacements are typically larger than those from a non-isolated situation. Since the blocks adjacent to the Type B upgraded bents are not isolated, careful consideration was given to the effect of differential seismic support displacements on stresses in the pipeline that spans between the supports. Pipe peak seismic lateral displacements occur near the pipe expansion joint where bents support the pipe. Under the MCE, the peak displacement is about 30 cm (12 inches). The peak MCE displacement at the blocks is much less, about 6 cm (2.4 inches). Even under this differential movement, the pipe remains elastic and should experience no compromise in the pressure boundary (leaks or breaks). Hence, seismic isolation proved to be a very effective seismic upgrade solution. CONCLUSION The Mokelumne aqueduct system was identified as vulnerable to earthquake damage leading to possibly unacceptable outage durations necessary for repair. EBMUD is undertaking a two- 270

289 phase project to protect this key asset. Phase 1, completed in 1999, upgraded river crossings and buried pipelines. Phase 2, currently under construction, will upgrade the elevated pipeline portion of the aqueduct. This second phase will be complete in ACKNOWLEDGEMENTS CH2M Hill, Inc. (Oakland, California) was the design engineer-of-record for this project. Forell/Elsesser Engineers, Inc. (San Francisco, California) assisted in the design of the elastomeric seismic isolation bearing system. 271

290 272

291 SEISMIC-PROOF DESIGN FOR THE STRUCTURES OF THE WATER TREATMENT PLANT IN NIIGATA Hitoshi Hasegawa, Isao Hokari, and Kouei Ito ABSTRACT The Niigata Waterworks Bureau is presently constructing a new water treatment plant (WTP) to replace the old existing WTP which has a capacity of 80,000m 3 /day. The key structures of the new WTP have been designed to be safe in the event of Level 2 ground motion. This paper discusses three concepts on seismic-proof design with particular emphasis on the foundation and the structure in the new WTP. Concept 1: A study concerning the selection of the foundation of the structure for the new WTP based on research regarding the quality of the ground considering anti-liquefaction factors. Concept 2: A study concerning a method to restrain the occurrence of cracks in the concrete wall of the reservoir without setting expansion joints which have possible structural weaknesses in the event of an earthquake. Concept 3: A study concerning a process for checking the safety of the 52m high elevated tank using dynamic analysis. These three concepts were developed after ground assessment. The seismic bedrock was determined to be at the depth of 134m through a geo-technical survey boring and a geophysical logging. Hitoshi Hasegawa, Isao Hokari, and Kouei Ito, Waterworks Engineer, Planning Section, Niigata Waterworks Bureau, Sekiya-shimokawahara Niigata

292 1. INTRODUCTION The Niigata Waterworks Bureau (NWW) is a municipally owned and operated water utility serving approximately 520,000 residents using three water treatment plants with a combined total capacity of 360,000m 3 /day. Raw water is drawn from the Shinano River and the Agano River. The City of Niigata is presently constructing a new plant named the Shinanogawa WTP to replace the old existing Toyano WTP which has a capacity of 80,000m 3 /day. The objective is increased operational reliability and efficiency. The new WTP is located 2 kilometers upstream from the old one. The construction of the plant started in 2000 and is expected to be complete by autumn On June 16, 1964, the Niigata earthquake, magnitude 7.5, was the event that presented the phenomenon of liquefaction to engineers and this earthquake caused huge damage to our pipelines and facilities. The basic concept of the measures to protect against disasters for the new WTP is that the facilities must have the following seismic-proof abilities and redundant capacity: The key structures are designed to be safe in the event of Level 2 ground motion (L2 GM), which meets the Seismic Design and Construction Guidelines for Water Supply Facilities issued by JWWA (JWWA Guidelines). The treatment system has dual lines and its facilities are designed to have an additional 25 percent more capacity needed. This allows the water treatment system to remain stable in case of an accident, a disaster or during construction for improvements. The power supply system for the plant is figured for both normal and backup power to ensure a stable power supply. In addition, a generator is installed to provide emergency power. Drying Beds Elevated Tank Concentrator Finished Water Reservoirs Dual-media Filter BAC Filter Pump Station Operation Building Sedimentation basins Figure 1. Shinanogawa water treatment plant 2. GROUND ASSESSMENT The site of construction is in a lowland area along the Shinano River. Since NWW planned to build a 52m high elevated tank at the new WTP, a geo-technical survey boring at 15 points and analysis of the findings were conducted to study the stability of the ground to support the structure in L2 GM. One of the borings was at the depth of 150m, and a geophysical logging and a micro-tremor measurement were also performed to determine the dynamic characteristics of the ground. We will discus these dynamic characteristics in detail later. 274

293 2.1 Soil Conditions The soil conditions of the construction site are shown in Figure 2. The surface layer at -9.5m in depth consists of a layer of soft cohesive soil containing peat and fine-grained sand. Because the fine sand layer at -3.3 to -5.3m (As1) is a confined aquifer, it is expected to liquefy during an earthquake. The layer at -9.5 to -18m in depth (As2) is a tight sand layer with an N-value of 40 to 50, therefore the As2 has enough strength to support a foundation. The layer at -18 to -134m in depth is alluvial soil and the N-value of the layers changes at random from 20 to above 50. The groundwater level is near ground surface and varies according to the water level of the Shinano River. Since the surface layer is cohesive soil, its groundwater is confounded water. Elevated Tank Sand Filter Pump Station BAC Filter N-value Ac1 As1 Ac1 As2 (Foundation ground) As2 LEGEND As1, As2: Alluvial Sandy Soil Ac1: Alluvial Clayey Soil : Estimated Liquefied Layer Figure 2. Soil Conditions Seismic bedrock and Ground type The seismic bedrock was established by a geo-technical survey boring 150m and a geophysical logging. The boring reached to a buried terrace gravel layer (BTGL) at the depth of -134m through an alluvial soil layer. An average of sharing wave velocity of BTGL is 555m per second as the result of the geophysical logging. Therefore we confirmed that BTGL will be the seismic bedrock and L2 GM should be input for BTGL during dynamic analysis. Horizontal seismic intensity for design depends on the ground type that is classified by the ground predominant period in the JWWA Guidelines. By analyzing and comparing with one-dimensional ground response, the ground type at this construction site was judged to be Type 3 because both predominant periods shown in Table 1 were more than 0.6 seconds. The predominant period was calculated by an acceleration Fourier spectrum ratio of a surface wave profile against a base wave profile as a result of the ground response analysis which was a nonlinear one-dimensional finite element method (FEM) analysis. Seismic waves for the FEM analysis will be mentioned in detail later. Table 1. Ground predominant period SEISMIC WAVE PORT ISLAND E KOBE BRG Predominant frequency (Hz) 0.2~1.0(Max 0.3) 0.2~1.0 Predominant period (T G ) (s) 1.0~5.0(Max 3.33) 1.0~5.0 Type 1: T G <0.2s, Type 2: 0.2s T G <0.6s, Type 3: 0.6s T G 275

294 3. STUDY CONCERNING A PROPER FOUNDATION Most structures of the existing WTP in Niigata which were designed to be safe in the event of conventional Level 1 ground motion (L1 GM) adopted pile foundation in the past. To secure seismicproof ability in the event of L2 GM, a proper foundation for the new WTP was selected based on research regarding the quality of the ground considering liquefaction, strength and subsidence factors. 3.1 Design Considerations for Foundation We examined how to improve the foundation ground considering the following factors: (1) Liquefaction: The As1 layer was judged to be a liquefied layer after calculation using JWWA Guidelines in the case of not only L2 GM but also L1 GM. There are two requirements to prevent liquefaction of As1. One is that the N-value of As1 should be increased to above 12, and the other is that the estimated liquefied layer itself should be enclosed and fixed by a wall with the water contained there. The amount of subsidence of As1 by liquefaction is estimated 10 to 30cm (equal to 10% of the thickness of the liquefied layer). (2) Ground strength: For pile foundation, the ground strength of the Ac1 layer is needed to increase deformation coefficient to be above 50 kg/cm 2 around the depth of 5m upper part of the pile. For direct foundation, the ground strength of Ac1 and As1 is required to satisfy the load conditions of the structure. (3) Subsidence: Pile foundation and direct foundation should be set in and on the As2 layer that has N-value above 40, because the surface layer at the depth of -9.5m consisted of soft cohesive soil does not have enough bearing capacity. predominant period Table 2. Design Considerations for Foundation Pile Foundation: PHC Piles with Sand Compaction Piles Liquefaction Increase N-value of As1 above 12. Ground Strength Increase Deformation Coefficient of Ac1 above 50 kg/cm 2 around depth 5m upper part of pile. Subsidence Set in the As2 ground (N-value is above 40 ). Direct Foundation: Improved by Deep Mixing Method Liquefaction Arrange pillars in the lattice-shaped. Ground Strength Improve strength enough to the load. Subsidence Set on the As2 ground. 3.2 Selection of a proper foundation Foundation for the reservoir and similar structures (e.g., sedimentation basin, filter) A selection of the foundation for the reservoir and similar structures was carried out considering requirements of ground improvement and results of construction performance. The screening process narrowed down the practical foundation options to the following three: (a) Direct foundation: Deep Mixing Method (DMM) of soil stabilization (b) Direct foundation: Replacement Method (c) Pile foundation: Pile Installation of PHC by inner excavation after ground improvement by sand compaction pile (SCP) 276

295 The final selection was made according to financial considerations as follows: Method (a) is about 60% of the total costs of method (c). Method (b) costs too much because the replaced ground area goes to the depth of 10m. Therefore, a direct foundation with the DMM was chosen for the reservoir and similar structures in the new plant. A layout of pillars of DMM is lattice-shaped as shown in Figure 3 that is an effective measure for anti-liquefaction. In addition, slurry type of cement was used for pillars of DMM because it is more reliable in the lapped part of pillars than dry type. The specifications of DMM are shown in Table 3. Table 3. Improvement object specification of Deep Mixing Method Sedimentation Items BAC Filter Dual-media Filter Reservoirs Basins Improved length (m) ~ ~ ~4.50 Improved rate (%) Design strength (kgf/cm 2 ) Table test strength (kgf/cm 2 ) Figure 3. Deep Mixing Method in the lattice-shape at a reservoir Foundation of the elevated tank Pile foundation was adopted for the elevated tank because direct foundation may not provide needed security from large seismic movement and the tank may topple down. A cast-in-place pile of 1500mm caliber and 15m length was chosen comparing the costs of the type and the caliber of the piles. The ground improvement around the piles was performed with sand compaction pile. (Figure 4) Improved area by SCP Figure 4. Cast-in-place piles of 1500mm of the elevated tank 277

296 4. STUDY CONCERNING CONCRETE CRACK CONTROL As we all know, reservoirs need a good waterproof design. This study concerns methods to restrain the occurrence of cracks in the concrete wall of the reservoirs without setting expansion joints because expansion joints can be a source of structural weakness in the event of an earthquake. 4.1 Profile of the finished water reservoirs Capacity: 44,000m 3 11,000m 3 /one unit Structure type: RC flat slab Basement type: Direct foundation Inside dimensions for storage: W=45.4m, L=54.6m, D=5.8m Thickness of concrete Basement: 80cm, Wall: 60cm, Slab: 50cm Figure 5. Sketch of the reservoir The reservoirs have been constructed under the ground. Each facility in the new WTP including the reservoirs must be able to treat water by gravity flow. 4.2 Exclusion of crack causing joint Expansion joints proved to be weak points in the structures of the WTPs when the Kobe Earthquake occurred. Thus, studies regarding thermal cracking control were conducted to find a method to exclude crack causing joints even when expansion joints were not being used. At first, we tried to use low-heat Portland cement to help to restrain the occurrence of thermal cracks in the concrete wall of the reservoir. However, moderate-heat Portland cement was available in Niigata but low-heat Portland cement was not. Thereby, moderate-heat Portland cement was used and mixed with fly ash as substitute for 20% of the cement. Also the concrete was placed after the middle of October when the temperature goes down to 22 degrees centigrade. The results of the study of the concrete compressive strength and the thermal cracking analysis are presented below. Concrete mix design Three different conditions of using moderate-heat Portland cement partly mixed with fly ash to decrease concrete calorific value were examined. Concrete mix design and the substituted fly ash ratio were as follows: Table 4. Concrete mix design Cement Additive Nominal strength Coarse aggregate Slump Material age Moderate-heat Portland cement Fly ash 20% 21 N/mm 2 25 mm 12 cm 56 days 278

297 Moderate-heat Portland cement only (M) Moderate-heat Portland cement + fly ash 20% (M+F20) Moderate-heat Portland cement + fly ash 30% (M+F30) At 56 days material age, the compressive strength of M was almost the same as M+F20 but the strength of M+F30 was lower as shown in Figure 6. Compressive strength (N/mm 2 ) Age (day) Figure 6. Material age and compressive strength M M+F20 M+F30 Thermal craking index Temperature Figure 7. Concrete temperature and crack index M M+F20 Thermal cracking analysis The analysis of the thermal cracking index of concrete was performed using a two-dimensional FEM. The goal for thermal cracking control is to keep the thermal cracking index above 1.45 that means below 25% probability of crack generation. The number of the cracking index of M+F20 is higher than that of M at all temperatures in Figure 7. This shows the effects of mixing fly ash into the cement. In the case of M, the index 1.45 intersects at 18.3 degrees centigrade which is the temperature in the beginning of November. In the case of M+F20, it intersects at 22.0 degrees centigrade which is the temperature in the middle of October. 4.3 Actual occurrence of cracks The temperature of the concrete when it was placed in the wall was a bit lower than temperature used during analysis. The actual initial temperature was 18 degrees centigrade and the maximum temperature was 26 degrees centigrade. Nine months later when the reservoirs were filled with water. The number of cracks actually occurred only at eight points in all four reservoirs. All of those were within the allowable 0.04 to 0.15mm widths. We were satisfied with the results of this performance. 5. DYNAMIC ANALYSIS OF THE ELEVETED TANK The service area of the new WTP is flat land of 45km 2. The tank was made to be elevated so that water can be distributed through gravity in the event of a power interruption. The gravity flow system utilizing the elevated tank simplifies pumping and instrumentation equipments for distribution. The height of the elevated tank is 52m and its capacity is 6,000m 3. The structure of the tank was the subject of a dynamic analysis check because the behavior of the tank is greatly affected by the ground 279

298 characteristics in the event of an earthquake. 5.1 Profile of the elevated tank Capacity for distribution: 6,600m 3 Capacity for back washing: 1,340m 3 Structure type: PC for the tank RC for the legs Dimensions: H=52m, f26m Basement type: Pile foundation f1500 x 15m x 136 piles Figure 8. Sketch of the elevated tank 5.2 Dynamic analysis procedure The dynamic analysis involved four steps that took into consideration the ground characteristics, the structure and the influence of retained water. It was evaluated from the results of four analyses confirming the adequacy of each modeling. Dynamic analysis procedure was as follows: (a) One-dimensional response analysis for the ground: It is carried out in order to get a wave input for the two- dimensional analysis considering nonlinear response of the ground characteristics. (b) Sloshing analysis for the stored water. (c) Two-dimensional FEM analysis for the structure unified with the ground: It is carried out in order to get a wave input for the three-dimensional FEM analysis considering the foundation response. The structure is modeled on the beam and mass system and the ground is modeled on a depth 0 to -64m. (d) Three-dimensional FEM analysis for the structure: It is carried out in order to check the detailed seismic-proof ability of the parts of the structure. One-dimensional response analysis for the ground Two-dimensional FEM analysis for the structure unified with the ground Sloshing analysis for the water Surface: 0m GL. 64m Three-dimensional FEM analysis for the structure GL. 134m Figure 9. Dynamic analysis procedure 280

299 5.3 Modified waves for dynamic analysis Port Island GL-83m N12E (PI) wave and East Kobe Bridge GL-33m N12W (EKB) wave which are waves recorded in the 1995 Kobe Earthquake were chosen as the seismic motion for the dynamic analysis and pulse frequency characteristics of those waves were modified to correspond to the response velocity design spectrum in JWWA Guidelines (Figure 10). The graph of the Port Island wave shows its amplitudes and the graph of the E. Kobe bridge wave shows its periods. Figure 10. Modified waves and adjusted response velocity 5.4 The results of analyses Ground characteristics The result of the one-dimensional ground response analysis showed the maximum acceleration on the ground surface to be 324gal for the PI wave and 349gal for the EKB wave (Figure 11). These values of acceleration were smaller than 588gal which is the value of acceleration input in the case of static analysis. The decline of acceleration was caused by a strong non-linear response of the ground which has a 130 m thick alluvial soil layer coupled with the rigidity of mixed silt and sand layer as well as caused by the input of large seismic ground motion. The shearing strain of the surface layer Ac l reached 5% in some parts. Depth (m) Acceleration (Gal) Figure 11. Distribution of maximum acceleration in the ground 281

300 Sloshing The result of sloshing analysis showed the first degree natural period of the water in the upper part of the elevated tank used distribution was 6.97 sec in the outside tank and 5.17 sec in the inside tank (Table 5). These values were remarkably longer than the natural period of the structure in the case of the fixed foundation and the ground spring foundation which were 0.37 sec and 0.59 sec respectively. Therefore, it was concluded that a heavy sloshing will not occur and stored water will not have a negative effect on the structure in the event of an earthquake. So the water is viewed as fixed water on the two- dimensional dynamic FEM analysis and the three-dimensional dynamic FEM analysis. Response acceleration Table 5. Natural period of the water in the tank Outside tank Inside tank Degree Frequency (Hz) Natural period (sec) Frequency (Hz) Natural period (sec) First Second Third Figure 12 shows the results of the two-dimensional dynamic FEM analysis and the three-dimensional dynamic FEM analysis. The difference of the average acceleration between the PI wave and the EKB wave was only a small percentage, and the average acceleration in dynamic analysis was smaller than acceleration in static analysis because the acceleration of dynamic analysis was input considering the ground characteristics. Height (m) Acceleration (Gal) Port I. 3-D E Kobe Brg. 3-D Port I. 2-D E Kobe Brg 2-D Static Analysis Figure 12 Distribution of maximum acceleration in the structure Check of design strength The dynamic analysis of a structure as defined in JWWA Guidelines is an analysis to check the reliability of the static analysis which uses the seismic coefficient method. The maximum bending moments and axial forces at the bottom wall on the basement are as shown in Table 6. The numerical values of the static analysis are bigger than the ones of the dynamic analysis in 282

301 all check points. The result of the dynamic analysis showed the reliability of the static analysis. The safety of the elevated tank was checked from the result of the four analyses. Check point Outside wall on the basement Center wall on the basement Inside wall on the basement Table 6. Maximum bending moment Dynamic analysis Static analysis Port I. wave E. Kobe Brg. wave Ny (tf/m) My (tfm/m) Ny (tf/m) My (tfm/m) Ny (tf/m) My (tfm/m) ±18.13 ±2.78 ± ±20.65 ±3.08 ± CONCLUSIONS This paper discussed three seismic-proof design schemes for the foundation and the structure of the Shinanogawa water treatment plant in Niigata. 1. To secure seismic-proof ability against the Level 2 ground motion, the Deep Mixing Method was chosen to improve the ground where the foundation of the reservoir and similar structures were to be constructed. The layout of the pillars was lattice-shaped. They were made from slurry cement because this was judged to be more effective and reliable to prevent liquefaction. 2. In order to exclude expansion joints without using crack causing joint, moderate-heat Portland cement was mixed with fly ash and substituted for 20% of the cement for the concrete wall, and concrete was placed after the middle of October when the temperature drops below 22 degrees centigrade. This was according to result of the concrete mix proportion test and the thermal cracking analysis. 3. The structure of the tank which is 52m and has a capacity of 6,000m 3 was checked by dynamic analysis because the behavior of the tank is greatly affected by the ground characteristics in the event of an earthquake. A four-step dynamic analysis was performed that took into consideration the ground characteristics, the structure and the influence of retained water. The safety of the elevated tank was checked from the result of the four analyses confirming the adequacy of each model. The authors believe that those three seismic-proof design schemes may be useful for other water facilities. ACKNOWLEDGEMENTS The authors would like to thank Mr. Masaya Goto and Mr. Kimiyasu Otake of Nippon Jogesuido Sekkei Co.,Ltd for the geo-technical studies and dynamic analysis of the elevated tank and Mr. Hiroya Takahashi of Kagata Corporation for the thermal cracking analysis. 283

302 REFERENCES (1) Seismic Design and Construction Guidelines for Water Supply Facilities Japan Water Works Association, 1997 (2) Design and Construction Manual for Deep Mixing Method Public Works Research Center, 1999 (3) General Specification of Concrete; Construction Part, Design Part, 1996, Proof Audit Design Part, 1999 Japan Society of Civil Engineers (4) Design Criteria of Concrete with Fly Ash (Draft) Japan Society of Civil Engineers, 1999 (5) Recently Mass Concrete Technique Japan Society of Civil Engineers, 1996 (6) Reducing the Clacks in Mass Concrete Japan Society of Civil Engineers, 1981 (7) Design Criteria and Explain Pre-Stressed Concrete Tank for Water Facilities Japan Water Works Association, 1998 (8) Dynamic Analysis and Seismic Design, Part 1 and 2 Japan Society of Civil Engineers,

303

304

305

306

307

308

309

310

311

312

313

314

315 3 rd US-Japan Workshop on Water System Seismic Practices TECNICAL SESSION VIII Panel Discussion Important Aspects to Include in a Water System Improvement Program Introduction Craig Davis, Los Angeles Department of Water and Power Seismic Resistant Design David Lee, East Bay Municipal Utility District Risk Assessment and Analysis Masakatsu Miyajima, Kanazawa University Seismic Risk Management Masanobu Shinozuka, University of California, Irvine Seismic Preparedness and Readiness Makoto Matsushita, Kobe Water Works Bureau Seismic Performances and Post-Earthquake Recoveries Jiin-Song Tsai, National Cheng Kung University Session Closing Craig Davis, Los Angeles Department of Water and Power 297

316 3 rd US-Japan Workshop on Water System Seismic Practices 298

317 Panel Discussion - Important Aspects to Include in a Water System Seismic Improvement Program Introduction Craig Davis: Are you wide awake after such a good lunch in the sun? I know we have had a long 3 days. I know I am feeling a little tired, but I am feeling very excited about this afternoon s panel discussion. Hopefully my introduction will not be boring; it will be very lively for you to wake up and discuss. The idea that you will see is to have everybody join into discussions, under at least controlled sessions, so we can acquire more information. This is session number eight, our last session outside of the closing at the end, it is the panel discussion. The purpose of the panel discussion is to stimulate and encourage discussion amongst all members, all attendees here, not just the people up in the front. The topic, as you know, is Important Aspects to Include in a Water System Seismic Improvement Program. The topic is second to the idea of discussing. Discussion is the main theme and we have a topic to help spur that along, keep it controlled, and also to learn. It is a very general topic if you really think about it. Everything that we have presented so far, and discussed, and on the tour yesterday has something to do with what somebody thinks is an important aspect to include in a seismic improvement program for some reason. This panel discussion is just to further learn from that. The goal is to formulate the discussion among participants. Discussion is key as you see. Provide a better understanding of the important seismic improvement aspects and develop useful information for improving our respective seismic improvement programs. We will document the recorded information that we have from here and a few of us will process that information and put it in the proceedings. Now a nice goal, and hopefully we can do this, is that we will come up with something with our respective knowledge and experiences that will be useful outside of the proceedings. Something that our organizations and other organizations around the world will be able to say: Look what the United States AwwaRF participants and Japan JWWA participants and the Taiwanese TWWA participants that have gotten together several times and look what they have come up with from their knowledge. One idea that has been brought up several times throughout the past few years is that in this workshop, now that we have gotten together a few times, maybe we should start having some kind of a product out of this. So that is a secondary goal that we will have here; secondary to DISCUSSION! The panel members, actually I see that you are not up front because I did not invite you yet, so we will do it this way: I was going to have you stand up so as I mention your name would you please stand up and enter the front with what ever materials you would like because we have a table here for you. David Lee from the East Bay Municipal Utility District will have a topic of Seismic Resistant Design. Professor Shinozuka from the University of Southern California (now at the University of California at Irvine) who [will have a topic of Seismic Risk management], Shino you do not necessarily have to stay up here but at least they will see you. Shino had a homeland security project that came up this week and was unfortunately unable to attend the workshop but he has spared time to come this afternoon and he is making his PowerPoint presentation for his five-minute summary. So Shino, if 299

318 you need to go work on your stuff, please feel free to. Makoto Matsushita from the Kobe Waterworks Bureau will have Seismic Preparedness and Readiness. Professor Miyajima from Kanazawa University will have Risk Assessment and Analysis. Professor Tsai from the National Cheng Kung University will have Seismic Performances and Post Earthquake Recoveries. As you see, what we have here is one practitioner experienced from the United States and one researcher experienced in water system seismic improvement programs from the United States, one Japanese practitioner, Japanese researcher, and one Taiwanese researcher. Unfortunately, we do not have room for six so we would not have time. We added the Taiwanese after we had organized this as well, so we were unable to add a sixth person. We have two recorders, Mr. Le Val Lund and Mr. Frank Collins. Frank will be doing the hand written and Le Val will be summarizing on the board for people to see what the best portions of the discussion are. We have two chairpersons. One will be myself and the other is Professor Jean-Pierre Bardet from the University of Southern California, who also got called away just prior to being able to attend this workshop but he has found time to help us out during the panel discussion after his quick return from Chicago. I have created some discussion aids to help carry on and initiate some of the discussions. I say I, that is not correct, we have created these. One was a survey of important aspects to include in a water system seismic improvement program that I sent out prior to the workshop and had, I believe, some very good response. The other was the session recordings that have been going on and those will be used as appropriate and felt necessary by the panel members for their summary. This is an example survey. This is the example you all saw, I sent it out by making the Department of Water and Power example. We wrote down an aspect, categorized the topic, and wrote down the reason for why we feel it is important. All of those are important to include because as Mr. Matsushita pointed out, and is well taken, I think many of us well understand that you might have an aspect and you might categorize it on a topic for what ever you re thinking at that time, or for what ever your purpose is, your intent is at the time that you come up with this idea, but if you think about that very same aspect in a slightly different manner it might be under a different topic. You might want to improve or you might include putting in new redundant pipes as a risk assessment because you just analyzed it, or you might call it something else because you re planning for it, or you might call it something else because you just constructed it, but it is all the same thing. You are putting in new pipes, so it depends on your perspective and that s why it s important to have this reason. The second thing that we have come up with over the past few workshops that is important for the reason, so if we can have that included as part of the discussion and the recording where ever possible, is because some aspects some organizations have never thought of, and so it is important to understand why you think it is important. For example, if it has never occurred to the Department of Water and Power and we don t want to think about it, that doesn t mean is it not important. Maybe we should understand why it might be important, but then again maybe we don t have that type of an issue here but other people will, so it is important to the exchange of information. We summarized the results and I don t expect you to actually read all of this. We had 13 survey responses. These are the six categories and these are the different topics. Basically, it is roughly 13 times five different topics that we have and these are how they 300

319 broke down. It is pretty good, except for this other category, which I think is great because we got two new ideas. It has a limited response but that might be the most important ideas that we come up with. The organization of this panel discussion will be that we have these 5 topics. Each topic will have 20 minutes, or each panel member will have 20 minutes up in the front. They will provide a 5-minute summary of what they think is important from their point of view and their experiences, and maybe they will be able to include experiences that were just obtained from discussions in this workshop or some of the survey responses and then those will be recorded as we proceed. In conclusion, we will have a brief discussion on, this is after this panel session is all done, within this 2-hour period if we have time in fact we will need to make some kind of time, we will have a brief discussion on considerations of future workshops. Should we continue with this idea? We have had three now, should we have a fourth and what would be the goals of that? We ll try this. What would be the goals of that? Ok we re not going to discuss where, when, or anything like that in this session, but we do want to discuss what you guys think we should do with your workshop, this is our workshop, and what should be some goals. What is everyone thinking; it would be good to get some feedback on that. Now that will be at the very end after the 20-minute sessions are done. Interpretations just as the discussions went on, in fact I wanted some interpretations already so we will step back a little bit. We will, the questions you know basically after anyone says something we will have some discussions. And if any of the, we will not interpret the 5-minute summaries unless the panel members believe that it is important to make sure certain things are said in multiple languages so that the information is out there because we can not screen out, it is not appropriate to screen out the people who do not speak English not as well as we do because I certainly couldn t being doing this in Japanese. I skipped you [the interpreters]. I was going to have him summarize a few things. Seishi Nonaka. Japanese interpretation. 301

320 302

321 Seismic Resistant Design David Lee. Ok, I guess we can get started. I am really honored to be here. I really feel excited for this whole section. The reason why is, this is called a workshop. This is not just another conference with everybody going there to make their presentations and leave. I think we probably accumulated a lot of questions during the past two and one half days and this is where we are asking each other questions and making comments and maybe start a debate and argument. This is the spirit of this workshop. I am glad that John didn t wear a tie, and he was gone for a couple of minutes and I was in a panic because if John s not there, no one s going to start it. For the workshop, I think that if you would like to follow me or you do it later when discussions are getting off, you can take off your jacket. Then please release your tie because there is no reason to have a fight with the tie on. It might get jerked. All right, so first I would like to introduce my partner Le Val. He is not just here to record, he s going to be here to tell you that you did it wrong. The discussion should range from very simple questions like what happened this morning. Paul Somerville asked Dr. Tsai: What do you mean we cannot predict the seismic or the earthquake, you are going to take my job away? That was one question. All of the other range could be, ok lets talk about ductile iron pipe is better or steel is better. My session is 20-minutes and Mr. Matsushita is recording me. It is seismic resistant design. Le Val and I thought about it. With all these presentations we had maybe we can separate them into sources and transmission. Any questions you can think about from the presentations related to the sources and transmissions; this means the water supply, the dams, the transmissions, the aqueducts, the tunnels, we saw the base isolation, we saw the collapsed dams, and we saw the big pipelines, tunnels. The second one is treatment, such as we add emergency generators. I heard the LADWP tell me, I hope Marty is here, we ve got these two emergency generators and we don t need any power after the earthquake. These two will operate the whole plant. I would like to know how big the capacity is and when you say you operate the whole plant what does that mean? Does that include a fluoridation system or just the essential systems? Storage, the tanks, we saw the DYK tanks. I haven t seen a tank in Japan use the pre-stress concrete DYK tanks but I saw a lot of concrete tanks, duel tanks. LADWP is adding flexible joints, Flex-tend, EBBA, and all that. In pumping plants we know that they are usually very rugged and they survived the Northridge by our friends design here. I would like to know the use of the Japanese standard or the US standard because he can do both. Distribution is probably a big area. The pipes, we have steel pipes, ductile iron pipes. Today Dr. Suzuki introduced a new methodology to estimate the damage of the pipelines. You know I was wondering if there was a way we can get that program, free! We have to make that decision. What ever the future is we have to have that. Its value will change all the current practices right now. It is not just good for steel pipes, or he will share that with Kubota for a few million dollars. Within this, if you can think about all of the questions we will just open it up for discussion. Anybody? Seishi Nonaka. Japanese translation. 303

322 Craig Davis. One format that I was going to introduce right now is that we plan to do is that we have the room nicely split down the middle. I counted and there was one more on this side, so you need to ask one more question. We will start by asking, we will split it on each side, so being that this is primarily the US side and that is where David Lee is from we will start with some discussion open on this side [US], but the next one will be split onto this side [Japan]. So that we are not leaning. Marty. Marty Adams. Can I ask the first question? David Lee. Yes, please Marty Marty Adams: We were discussing at our lunch table, we weren t quite arguing, but we were discussing the fact that we know that the large physical structures can be made to withstand an earthquake, but is it really practical to presume that the distribution system and the piping could be constructed that it would actually withstand a quake and not break, or was that really impractical to assume that you could actually achieve that. Seishi Nonaka. Japanese interpretation. David Lee. Anybody have any comments? Craig Davis. So Marty, if I were to reassess what you just said. Marty Adams. What about those I had lunch with? Craig Davis. Who did you have lunch with? Who had lunch with you? Ah, Frank. David Lee. And Don. Ok Frank will answer it first then Don will be next. Frank Collins. Well we were discussing that we can make some of the facilities bullet proof and we can design the heck out of them. But really when it comes down to it, when the pipes go through some particularly bad seismic areas, how much can we do, and how much money should we spend to prevent something. It s almost impossible to prevent, if the earth is going to move then the pipes are going to break, so it is a matter of trying to determine that, and that is what we were discussing. I think there are some people who probably disagree with that, there s people that think we can design it so it won t break. Seishi Nonaka. Japanese interpretation. Don Ballantyne. Taking that one step further, that s the view in the United States. If you look at the various mitigation programs there is very little investment into pipeline systems, particularly in distribution systems; but the opposite of that is in Japan many systems have aggressive pipeline replacement programs. In the United States we re very happy if we can replace 1% of pipe in the system a year. Marty indicated that you re replacing all of the pipe in 200 or 250 year period, whereas in Japan you were talking 304

323 about somebody had a program where they are replacing it all in 15 years. That s a lot different and we re trying to understand that. Seishi Nonaka. Japanese interpretation. David Lee. Any other comments? I propose that we don t even have to stand up, just sit there and talk. You don t have to stand up; If you prefer to stand up that s good, that s better. Craig Davis. I heard one correction from Dr. Suzuki. It was a 50, five zero, year period not 15, one five, that Don had said. Don Ballantyne. It was 15. Craig Davis. Somebody had a 15 according to Don s understanding. Norio Iijima. Translated by Seishi Nonaka. In Japan we know that there is a whole bunch of bad soil, which is very corrosive and cast iron piping system that are underground are so weak and damaged. However, in the future we plan to use and we are using ductile iron pipe and we anticipate them to last one hundred years or more. David Lee. Ok, Dr. Suzuki you are not needed, steel pipe is out. Ok, Marilyn. Marilyn Miller. One thing I ve noted in visiting Tokyo, and we talked about this at lunch, is that Tokyo the size of geographical are served is similar to East Bay MUD s but they have 10 times more people. So for each kilometer of pipe you replace you benefit 10 times more customers in Japan than you do in the US. That may be a factor to consider in these decisions as well. Craig Davis. Very interesting point. Seishi Nonaka. Japanese interpretation. David Lee. Ok, John is next. John Eidinger. In Japan now you have a little yellow book for seismic design of water systems that says new pipe should have 1% strain capability. In the United States we have absolutely no standards for seismic design for pipelines. Every pipe that s purchased, almost every pipe has no seismic design at all. When we buy ductile iron pipe it has slipon joints which we see sometimes fails and with corrosion over 50 more years maybe will fail much more. We need in the United States a standard for seismic design for distribution pipe and we don t have one now. We need a manufacturer to sell some pipes that meet these criteria and we need to educate the utilities to demand to purchase this. It will cost more, but at least for new construction we need something because here in the United States, at least California, 70% of all new pipe is PVC, and only 30% maybe 30% is ductile iron, maybe 1 or 2% is steel welded steel for very big pipes or important pipes 305

324 but mostly PVC still. There aren t, if you open our standards from AWWA, there is no information about earthquake design in them and what the recommendations there are on it they are bad recommendations. This is what we are putting in the ground, it only makes things worse. Seishi Nonaka. Japanese interpretation. David Lee. Any other comments in this topic? Ok, please. Koichi Murata. Translated by Mitsuo Takasue. Ok, his question is so far we have been discussing the materials of the pipes and how to make the pipes more stronger to resist the earthquakes. He was saying there must be another way to prevent some kind of water leakage, like a rupture, there is way to correct the water in another way, so we may look the other way on how to mitigate the disasters. David Lee. Alright. Ok. Seishi Nonaka. Excuse me. There was a recommendation from my expert translator saying that the idea of having redundancy, that area should be expanded maybe that is going to benefit more. That is what he also mentioned. Hiroshi Yamada. Translated by Mitsuo Takasue. What he was saying is, based on the experience from the N earthquake. One of the idea is to separate all the utilities into many different blocks. So even a reservoir has many, many reservoirs and so you don t have to make everything into one unit. There is lots of backup system in case of an earthquake disaster. David Lee. Ok, Dr. Tsai you had a comment. Professor Jiin-Song Tsai. Thank you. I just heard John say that s there s no seismic criteria for the pipeline in AWWA and I realize I understand that in Taiwan we are probably in a similar situation. I am an academic researcher and we have real practice people here, Mr. Wang and I believe you experienced his presentation about all kinds of pipeline materials damaged in the Chi-Chi earthquake and I would like to communicate with him before a little bit of time and ask him to give his opinion on this. Ping-Hsin Wang. Mandarin translated by Professor Jiin-Song Tsai. Until Chi-Chi earthquake we realize the pipeline materials are very vulnerable to seismic in Taiwan. The A-type joint is quite frequently used, had been used for 20 years until Chi-Chi earthquake. This kind of A-type joint is not available for those countries, United States and Japan, for our understanding. We are looking for a new type of joint to replace older pipeline joint currently. David Lee. Ok, I think that s so fast for 20 minutes already. So I ll turn it to Professor Shino. 306

325 Craig Davis. No, we will have Professor Miyajima actually will be second and while he is getting ready I would like to maybe follow up to cover the space with responding to Frank s point. I would like to through out as a proposal that we should always be able to design to have the system function within a reasonable period of time, no matter what the damage. I think that should be a very achievable concept, within a cost effective manner. Now how that s done is the goal I think. Marty Adams. Can I propose something as a goal from the first topic? Being as the AWWA has no current seismic design standard for pipelines but the Japanese water works does have a standard, I would think it would be worth taking a look at existing data to find out if the standard does in fact make a difference in the earthquake performance or not so, that we can find out if it is a valid to look into adopting one or if there is truly no difference in what really happens in an earthquake. Craig Davis. Thank you for the suggestion. Did you get that Val? Or actually Frank you re the one that needs to get that. Frank Collins. I m trying. John Eidinger. Perhaps AwwaRF should fund an effort to develop a suitable seismic standard for design of new PVC, ductile iron, steel, what ever kind of pipes we have in the United States. Change all our C9 units and whatever with a section for seismic design. Elizabeth Kawczynski. We need a proposal. Craig Davis. That would be similar to what they did for tanks after the 1971 earthquake. Seishi Nonaka. Japanese interpretation. Don Ballantyne. In response to Marty s comment. There was 240 km of seismic joint pipe in Kobe exposed to the worst soils that were there and there were no failures. Mitsuo Takasue. Japanese interpretation. Makoto Matsushita. Excuse me, at that time in 1995 Kobe earthquake we had 250 km of seismic pipes. After that we have installed more than 200 km in eight years so we have now almost 500 km. Don Ballantyne. There was 250 km exposed in the Kobe earthquake Makoto Matsushita. Yes, and no damage. 307

326 308

327 Risk Assessment and Analysis Craig Davis. Ok, Ok. As the moderator, I am going to speak up and speak fast that it is time to move on to Professor Miyajima. I love the conversation. I have to apologize that we need to move on. Please, if there is any dead air space in the near future in the next hour and a half, please hold those thoughts. I would appreciate it. Professor Miyajima. Professor Masakatsu Miyajima. Thank you, I would like to talk about my topic of risk assessment and analysis. Firstly, I would like to introduce the result of survey for each waterworks bureau on the topic of risk assessment and analysis. First of all there is inspection assessment of existing system components, and estimation of damage to facilities, and provide redundancy, and GIS-based analysis of pipe networks. I would like to introduce simple computational method of estimation of pipe damage. The first day Mr. Koichi Murata introduced the simple method for the Osaka Waterworks Bureau. This is the number of damage to pipe is estimated by this simple equation. C p is the correction factor by pipe type, C d is the correction factor by pipe diameter, and C f is the correction factor by occurrence of liquefaction, and L is piping length in a cell so that the number of damage to pipe in a cell is estimated by this simple equation. Theory in Japan the strong ground motion record is easily obtained after the earthquake. This is a map of Kobe and this peak ground acceleration was obtained at each site and we can draw like this here so we can get easily the distribution of estimated damage ratio like this. I consider the relation between each topic in this discussion session and the horizontal axis is time before the event and after the event. My topic is risk assessment and analysis and the seismic performance must pay attention during the earthquake and post-earthquake recovery and event is of course after the earthquake. Risk management is conducted by using the risk assessment and the seismic resistant design is also used by the results of the risk assessment analysis. The seismic preparedness and readiness also used the results of the risk assessment and analysis I think. The risk assessment and analysis is based on this topic I think. Finally I introduce the results of the survey for each waterworks bureau. Risk assessment and analysis, the number of risk assessment and analysis was largest in these six topics. This is the basis of this topic I think. These four topics is focused before the earthquake, so many waterworks bureau pay attention before the earthquake. Thank you very much. Lets discuss. I would like to give the microphone to Craig. Craig Davis. Ok, so with that let s discuss. Are there any thoughts that Professor Miyajima has brought to mind? This time we would like to start on the opposite side of the room than we did over here. Maybe we could cancel the lighting here or something. Thank you. Are there any ideas on this topic of Risk Assessment and Analysis from the... Yoshihisa Iwasaki. Translated by Seishi Nonaka. The risk assessment should include the, or the strategic planning should have some influence on the risk assessment, what do you think? Professor Masakatsu Miyajima. Yes I agree with you. I showed the figure the result of risk assessment is used on the interpretations and some stuff like that. The risk 309

328 assessment is the basis of another 5 topics I think. The time axes is not so important. I said the risk assessment is the basis of another 5 topics. So I agree with you. Before the earthquake assessment is conducted but the result is used before, and during, and after the earthquake. Seishi Nonaka. Japanese interpretation. Craig Davis. Ok, thank you. Any additional comments? This side of the room. Paul. Paul Somerville. In a few parts of the United States, including Los Angeles and San Francisco areas and also I think Seattle, we now have shake map, which is a system for vary rapidly making ground motion maps very soon after the earthquake so that the emergency responders can know how strongly the ground shook and react appropriately. In Japan I would like to find out a little bit about the status there. I know that the keynet and kiknet in Japan at the moment are very fine and very dense networks, but at the moment I think they do not provide a very rapid information, but I know there are other networks, for example JMA and the Fire Bureau and so on. I d like to know: In Japan what kind of data are available on ground shaking in almost real time following the earthquake? Seishi Nonaka. Japanese interpretation. Kiyoshi Naito. Translated by Seishi Nonaka. I m from Yokohama and I do deal with map he says. The map shows one data for every 500 m square area. It is like a grid and each point is separated by 500 m. Based on such an evaluation magnitude 7.0 and larger they have the policy of using earthquake resistant pipe, which is supposed to be the strongest among all the piping systems. Corrections, it is not magnitude 7.0, the intensity is 400 gal., 0.4g. Professor Masakatsu Miyajima. In Japan the real time information delivery to the public, the Japan Metrological Agency seismic intensity is very popular. After the one or two minutes after the event the TV informs the seismic intensity to the public. Another system Tokyo GIS has a strong ground motion network so they can know the distribution of the ground shaking just after the earthquake. In the Kansai area, Kancentral, we call Kancentral, has an instrument network, so as a member of that association they can obtain the pgv or pga by paging the machine just after the earthquake. John Eidinger. How long, 10-minutes, 5-minutes? Professor Masakatsu Miyajima. Within 2 or 3 minutes. I am a member so I got the information for a not so large earthquake, small earthquake. Mitsuo Takasue. I think I have a suggestion. Since he can speak fluent Japanese, so instead of us translating in Japanese, that would be more accurate. Professor Masakatsu Miyajima. Japanese interpretation. 310

329 Koichi Murata. Translated by Seishi Nonaka. I am from Osaka municipal utility. We have 1900 locations where maximum velocity is recorded and 11 stations where the magnitude and accelerations and velocity are recorded. Craig Davis. Anybody, any other issues on seismic risk? Marty Adams. Marty Adams. Just a question on the understanding of risk assessment. Are most people doing it as an analysis or assessment on the component level to evaluate the susceptibility of a particular pipeline or facility or on a service level to the customers in terms of actual loss of service or loss of water pressure or something of that sort? That s my question. Craig Davis. Anyone want to respond to that? Mitsuo Takasue. Japanese interpretation. Professor Masakatsu Miyajima. Not only the component, but system risk assessment is very important I think. Craig Davis. We are going to close out this session, but I am going to do it with a raise of hands based on what Marty Adams suggested because I want to get a perspective of what we do at the Department. We have historically really looked at a component. I know that other people have, we have intuitively worried more about the system as its really redundant and created things more on an intuitive level and did analytical studies on a component. Other people have done the opposite and have selected some components to analyze and then went a little more intuitive that some components will resist and then looked at the system. So, by a raise of hands I m going to ask: who worries more about a component than the system as a whole? How many do, without the perspective and the interpretation, how many do components? Seishi Nonaka. Japanese interpretation. Craig Davis. Ok, so who worries more about the component? Two. Who worries about the system? So system analysis is much bigger in Japan, but there s lot of people over here that raised their hand and there was a lot of abstainment. I was curious. We are being informed again to stop. 311

330 312

331 Seismic Risk Management Craig Davis. Professor Shinozuka, Professor Shinozuka will present a slightly different perspective, but on a topic that was a recurring theme several times throughout the last two days in many different ways. What s that? This is Professor Shinozuka from the University of Southern - Not any longer, UCI. Professor Masanobu Shinozuka. Good afternoon. Thank you very much for the introduction. I made some preparation for this workshop. In the struggle of getting through some urgency at my office I thought it was 15 minutes for summary and 5 minutes for discussion. Then I just found out this morning it is the other way around. So I may have too much material so I will skip some of them. I only made very, shall we say, challenging proposition. I am very much interested in for many years in this concept of performance design or performance based engineering. Now in order to pursue that, to make a very long story short, you have to have performance criteria. I m not quite sure if we really tried that very hard. People talk about performance, performance, performance. Now what is performance? Particularly, when it comes to a very complex system as a lifeline, water delivery system, power distribution, etc., what I came up here is some global concept of performance. In fact I think the discussion we just had indicated that component vs. system, risk vs. management. These are very complex issues. I am an academic researcher and I think I am allowed to be very academic. However, I don t say that practice is not very important. It is the other way around, but let me talk about this academic concept. We have a number of dimensions in which we have to consider performance. One is technical where we usually concentrate, second organizational, as indicated here as a dimension, social, and economic. Now, within each category we have to consider two very important components, one is robustness that means how seismic resistant your system is, that one number. Two, how rapidly you can restore its intended performance, original performance, that is on the right hand side rapidity issue. Then instead of going abstract fashion as I have started out, let s talk about water. It s right in the middle. Water system we have to worry about robustness and we have to worry about rapidity and restoration. When we talk about, this is now only in technical dimension because that would be easiest for us to understand. In water, technical robustness requires that we have to maximize the availability of operational water supply after earthquake. This is of pre-earthquake level, so let s say 95% of performance is still maintained that would be robustness criteria. On the other hand, if you look at rapidity criterion, then we might say that all ready?- maximize provisional target water supply level should be 95% within one day, or something like that. These are two very important criteria for the performance. Now I can only go through two pages of PowerPoint. Craig Davis. Professor Shinozuka, we will allow you a few more minutes. Professor Masanobu Shinozuka. Ok, thank you. This is the more carefully put down the criteria. This is an example, I am sorry I can not talk about the Los Angeles Water System, we are working on it now. Some years ago we worked on Memphis Water System and this is the damage pattern of the Memphis City area water system and the 313

332 earthquake magnitude 7 epicentered at Marked Tree. This PowerPoint indicates that what kind of loss we would experience in terms of flow rate and the redder the picture looks the severer the damage ratio is. This is a kind of result you can get when we talk about performance criteria for robustness. Water system, this is technical second line, we can find that 5% of pipes sustain breaks and we can make the system restored to the level of 99% of the performance within one week. This is the kind of things we can simulate and test out our performance criteria. I ll stop here and I ll leave the audience with a discussion. Craig Davis. Thank you Professor Shinozuka. So we are on this side of the room. Are there any comments, suggestions, questions? Frank. Frank Collins. I thought the presentation was very good and I think that like the Department of Water and Power and Cities like the City of San Diego, we are taking into account those exact issues. Rather than sometimes looking at each component, we re trying to look at redundancy and the water sources. The same as the LADWP, the City of San Diego has many sources of water spread over a wide area. If any particular event occurs we re looking at how robust they system is to handle that and we re trying to keep the percentage and the number of customers out of service to a minimum. That s one of the things were focused on. That s what we re doing as well. Craig Davis. Thank you Frank. Any other comments? Maybe from this side? From this side? We have a whole bunch on this side. Marilyn I think was first. Marilyn Miller. I ve been thinking a lot about our seismic improvement program. We ve been working on it for eight years, have built a lot of improvements and so what do we do when it is finished? My thoughts have been that we need to start instituting some inspection programs to make sure that the systems continue to function that we ve put in, that they are maintained properly. I think that this becomes an ongoing program of risk management. It isn t just do an upgrade now and forget it. So that s something I think we all need to keep in mind. Seishi Nonaka. Japanese interpretation of Frank s and Marilyn s comments. Craig Davis. You did Frank s also? Seishi Nonaka. Yes Professor Masanobu Shinozuka. Could I make a comment on the second point? My thought about monitoring systems is that utilities usually have SCADA systems, that s supervisory control and data acquisition system. That was developed for normal operational purposes. I d like to see the extension of SCADA system concept as we see the emergence of the very advanced sensors indispensary. We might have many sensors throughout the system and we could use that for the detection of any damage due to earthquake however small that might be. When it is large of coarse it could be very useful to identify critically damaged locations. 314

333 Craig Davis. Any question on this side? Oh, translate, translate. Seishi Nonaka and Mitsuo Takasue. Japanese interpretation. Professor Masanobu Shinozuka. Ok, let me speak in Japanese. Don Ballantyne. I wanted to comment about performance objectives. Many utilities have adopted performance objectives, but in all cases have been done locally and there are not no national standards in that regard, comparable to let s say building codes. I concur with you, it s something that should be considered. The flip side to that there s a group within, let s say FEMA, believe that those decisions should be made locally and should not be made on a national basis as far as performance levels within systems. Craig Davis. Any feedback on that? Oh, translation. Professor Masanobu Shinozuka. Japanese interpretation. Professor Masanobu Shinozuka. Don, can I make a comment on that? There is a publication entitled Making the Nation Safer. This came out from the National Academy Press, supported by the National Research Council. There are a number of issues they listed for water supply and wastewater systems. One of them they point out that the lack of standardization is a weakness of a system fro terrorist attacks point of view. There are so many other things but that s one of the things they point out. Professor Masanobu Shinozuka. Japanese interpretation. Craig Davis. Any comments? Takashi Furuya. Translated by Seishi Nonaka. I m from Yokosuka and we have been working on this risk management for the past 25 years. However, we have a concern about the original water source, which is 30 to 50 km away. We simply assume that when there s an earthquake we are going to loose that water supply, period. However, because of that we have established a number of groundwater wells. We believe that since an earthquake comes and we loose the main supply line, we can activate the groundwater supply, which can provide 20 m 3 per person for 60 days. In addition, our motto is to utilize the piping system that is the strongest for piping over 150 mm. The main supply line can supply - we don t have just one supply line to supply water, 100% of the supply, but we have five. One is able to supply half a day of supply. As you can tell we are talking about redundancy here. Paul Somerville. In his presentation, Professor Shinozuka described the performance objectives in probabilistic terms. I m wondering if in Japan, when the performance criteria are prepared are they expressed in probabilistic terms? 315

334 Professor Masanobu Shinozuka. Can I answer? I don t think they do use probabilistic terms. It is very important that we recognize when you engage in performance engineering criteria you have to have some leeways. When you do design for example, in old days it says that you do a bridge shall not deflect expressed in terms of its span, for example. See then you can prove that you designed that way, yet still it can fail. On the other hand if you say bridge shall not fail then where is the protection for us engineers? The only way out is that chance is linked to very small probability. If you take it away you ll be troubled. That s how I feel, and I ve been saying that a long time. Professor Masanobu Shinozuka. Japanese interpretation. David Lee. I understand what the professor said, but my comment is do we want to be just like a weatherman? Professor Masanobu Shinozuka. I think that s a serious question. A weatherman, if they make a mistake whether it rains or shines, the consequence is minimal. Plus now you may get wet all over, but what is that, that s not so bad. On the other hand, the probability we are talking about is very, very serious. Yet you have to have something. I think the engineering seismologist can attest how difficult it is to provide a probabilistic number, but this is the best we can do. What else can we do? That would be my answer. Craig Davis. We need to close down the session. Actually I had request earlier from several people earlier so I will close down with request for two very, very brief comments and it s on this. Hold on Val. We had a comment that the United States has no performance objective criteria established uniformly. The question was posed by several people. Are there any performance objective established in Japan and are there any performance objective criteria established in Taiwan? So I have one answer that needs no further elaboration other than interpretation, there is no performance objective criteria in Taiwan. We d like to know if someone would like to comment briefly from Japan. Seishi Nonaka. Japanese interpretation. Professor Masanobu Shinozuka. I do not really think. Seishi Nonaka. Please speak in Japanese Professor Masanobu Shinozuka. Japanese interpretation to Craig Davis. Craig Davis. I speak English Professor Masanobu Shinozuka. Oh, I m sorry. What was your question or comment? Craig Davis. The question is briefly, is there any performance objective criteria established for the country of Japan? There is none for the country of the United States. Locally there might be individually determined criteria. Is there an across the board? 316

335 Le Val Lund. National Professor Masanobu Shinozuka. Japanese interpretation. Yoshihisa Iwasaki. Translated by Professor Masanobu Shinozuka. As far as I can understand there is one - is that a utility company? [Yes] - that prescribed to performance design criteria, but as far as the government is concerned - Is that the Ministry of Construction? [Yes] even if they prescribe performance criteria, they leave it up to the local utility companies to do their own design and they can claim that they meet the performance criteria the government is imposing. So it s sort of in between. Paul Somerville. Are they bringing any probabilistic aspects to the performance criteria? Professor Masanobu Shinozuka. I don t think so. 317

336 318

337 Seismic Preparedness and Readiness Craig Davis. We need to move on now. Ok, we will hand this over to Makoto Matsushita from the Kobe Waterworks Bureau. Makoto Matsushita. The topic assigned to me is Seismic Preparedness and Readiness, but I decided to combine this topic and the other topic at my own risk. Anyway, I picked four key words here. One is redundancy, the second one is recovery process, the third one is communication, the fourth one is facility aging and system reconstruction. This one is the other topic. The first one, redundancy, it means transmission mains and water supply pipelines and distribution. This is Japanese English I think. These are to make redundancy to avoid critical damage of water supply system. The second one, recovery process, is very important to shorten the recovery period in an emergency situation. Also, response plan also has an important role for recovery, quick recovery. Third one, communication, it means, communication means industrial customer, and residential customer, and other stake holders, government, or other agencies. Communication is for the investment and to make decision. This I explained the other day and Le Val Lund told me this is informed disclosure. The fourth one is also important I think, the facility aging and system reconstruction, which includes pipe replacement, that is mostly in Japan and the replace to electric valves, this is bold faced I think, and water recycling is the demand, and adaptation. These are adaptation to paradigm shift, which means social changes including population decrease in Japan, and higher requirement for water quality, and functional improvement, and water demand increase or decrease, I don t know; anyway, the adaptation to the new world, new society. Finally, I conclude this presentation; we should utilize the opportunity of the seismic improvement program for the other purposes as much as possible. Not only earthquake, for other purposes that s the best way I think. Finally, according to the Japanese custom, I put the eyes on this Daruma. This is called Daruma, Japanese folk art, here and here. This is not a real Daruma, but I have a real Daruma and I would like to ask Craig to put the eyes here in closing the workshop. Thank you. [Presented Daruma to Craig Davis] Craig Davis. Ok. Is this traditional Japanese? Makoto Matsushita. Folk art. Folk art is a toy, just a toy. This is a persistent person so, oh no. Hopefully he sometimes returns to stand up. Craig Davis. Mr. Ishii, Ok, moving on with the session, are there any comments starting with this side of the room? Kenetsu Kojima. Translated by Seishi Nonaka. I come from the northern end of Japan, Honshu, it s the mainland and it s kind of in the countryside and we have experienced a number of earthquake damages. Nobody in the nation recognized it as an important event and nothing was taken and we weren t seriously taken. However, when Kobe earthquake came and 6,000 and some people died, the entire nation participated in the restoration and all that. Can we talk about some kind of standardization strategy? 319

338 Makoto Matsushita. It is a comment and he agreed with me about the communication is very difficult, very important it seems. Is that right? [yes]. Any other questions, and other comments please. This side or Japanese side. Ok, this side Murakami. Keiichi Murakami. Translated by Seishi Nonaka. I m from Hanshin area municipal utility and we have experienced the Hanshin great earthquake. What we did immediately after the earthquake is to come to unite with other government agencies and other connections and came to an agreement and made a contract to jointly work together incase of a disaster like that. Translated by Mitsuo Takasue. He also mentioned they had in the Kansai region and that they had the same kind of contract with the east portion of the Kanto region and they have a similar kind of mutual assistance programs. They have a program right now but they haven t had any disasters yet, so they haven t tried how this mutual assistance would work in the future. Kiochi Murata. Translated by Seishi Nonaka. I m from Osaka area municipal utility and we have prepared manuals to follow to plan against a disaster and also how to restore the system and we provide training on a regular basis. Makoto Matsushita. Can I make some comments on his comment now? In Japan, the larger cities have mutual aid agreements with each other. For example, Kyoto, Osaka, Kobe, Hiroshima, north to south, located north to south. If we have, if we Kobe have a disaster and north side Osaka and south side Hiroshima come to Kobe and make support. If Osaka have disaster Kyoto and Kobe will go there and support his recovery. How about US side mutual scheme, someone, anybody speak? Craig Davis. Anybody from the United States? Marilyn. Marilyn Miller. In California there are a number of mutual aid arrangements. I know when the Northridge Earthquake happened East Bay MUD crews were helping, and so goes the length of the whole state and also in the immediate Bay Area. I m not sure how it is in other states though. Seishi Nonaka and Mitsuo Takasue. Japanese interpretation. Makoto Matsushita. Ok, Mr. Fukuda. Katsutoshi Fukuda. Translated by Seishi Nonaka. Japan Water Works Association has divided Japan into seven different districts and they all have the connections and they identified a strategy to help each other. Makoto Matsushita. Ok, any other comment? Marilyn. Marilyn Miller. I d like to speak to the issue of communications a little bit. Maybe it goes to the point made earlier that the rural communities don t get the attention when they have the damage. I think we as a water industry, it sounds crass, but we could take 320

339 advantage of the opportunities to publicize the need for seismic improvements when there is a disaster that happens. I know at East Bay MUD it was just fortuitous when we started our program that we had Loma Prieta, Northridge, and Kobe within a short span and there was a lot of public support, but perhaps we can utilize that as a strategy to gain support for the kinds of programs we know are important. Seishi Nonaka. Japanese interpretation. Makoto Matsushita. Ok, I would like to have a question or a comment from the US side. I presented the last idea as system reconstruction is very important when we think about earthquake improvement. Presentation was only dedicated only to the earthquake area. We are now thinking about earthquake and water quality improvement. Another policy is to get together and put in one project. How do you think about this idea US side? John Eidinger. I believe you are correct as most US utilities do it the way you do it. Makoto Matsushita. Japanese interpretation. Bill Heubach. I think it an excellent idea and it s something we re trying to do at Seattle Public Utilities. The problem that we re struggling now is in our pipeline replacement program is trying to how to prioritize pipelines for replacement and how these different considerations factor into the prioritization. I was wondering if anybody had any ideas or any comments on what they re doing to prioritize based on these different considerations. Seishi Nonaka. Japanese interpretation. Kiyoshi Naito. Translated by Seishi Nonaka. The first priority is obviously to get a hold of the water supply and issue it to the people, and then think about the next priority. Makoto Matsushita. Marilyn please. Marilyn. Oh, oh, Marilyn first. Craig Davis. That s his boss, so. Marilyn Miller. Well with regard to pipeline we re kind of doing it the other way around. We have pipelines we know we need to replace for maintenance reasons or what ever and we use seismic issue as a way to prioritize those. Another thing we found if you re looking at particularly storage tanks, if you need to seismically upgrade a tank, a number of ours all had water quality problems because they were too big and the water quality degraded. So we were able to actually replace them with smaller tanks instead of upgrade the old one. So we achieved two purposes there cost effectively. Seishi Nonaka. Japanese interpretation. Makoto Matsushita. Thank you very much. The panel is over, so but we have a last comment, briefly please. 321

340 Craig Davis. Mr. Singley. Glenn Singley. Ok, here s just another twist and another idea, something that we are working on again with MCEER. There is a concern about after a large event would occur, what areas do you try to target for recovery, and how long would that recovery be? We re trying to develop some simple models to show the economic impact of the effects of different areas being impacted by an earthquake and how would the economics of the region dictate where you would start in some of your recovery process. So, which part of the system do you need to recover first to get that economic engine back going if it is a large scale problem? So that s being modeled now and hopefully sometime in the future we ll have some answers to that. Seishi Nonaka. Japanese interpretation. 322

341 Seismic Performances and Post-Earthquake Recoveries Craig Davis. Ok, thank you Makoto. Very, very good job! We will move on to Professor Tsai for ah. What was your topic? Professor Jiin-Song Tsai. Seismic performance and post earthquake recoveries. Craig Davis. Thank you. Professor Jiin-Song Tsai. Ok. We re probably a little bit behind schedule. In my presentation I think I probably very little time and just briefly survey a portion of the interesting topic to our all friends here and I just want to highlight some key words here, which I and give an introduction. Still warming up, warming up? All right, there are eight topics, eight topics completed here. What I would like to say is that all these questions I organized in a three-layer system. Here, we look at here the first person I notice is the service goal. What is a service goal we d like to give to our concerns, concerning seismic performance and recoveries? In one business, I d like to say we people on this topic we probably have to consider what we are, what we are. In one business we re always talking about a vision, what kind of vision we want to provide to our customers. What kind of leading environment or kind of needs they can have. Then we can set up our mission. The first layer may be the honor or the people want the business or want to think about it. We work on this and it s just the same mission or vision we want to put our effort on. The second layer then we base on both vision and mission, we give our goals, several kind of goals and service goals is one of them. Of the seismic condition, what kind of service goal we wish to give after those damaged things. This tells us the goal is a second level thing we have to consider. The third layer we re talking about the performance, in an earlier section we talked about performance, here I looked here and found system. So we re talking about system performance. This is the three layers I found in our topics here. I believe in this section we probably will focus on system performance based on, probably here we re talking about inspection and some previous experience. Here they re talking about learn, study post-earthquake response in other areas, which is concerned about previous experience. We want to invite those previous experiences into our system performance, and also we re talking about inspection I just mentioned and another here I just saw previous experience. People from Kobe area, they re talking about mutual aid scheme. Their experience learned from that disaster, they wish to give those ideas in this section. I think this is all I want to say in my portion and I wish to give you some idea, system of thinking, and we discussed about seismic performance and post earthquake recovery. Can someone help me? Craig Davis. I will help you. Oh, need to kill some lights. Thank you Jianping. Professor Jiin-Song Tsai. Before that, one comment first. An old saying in Chinese, people of compassion love mountains. Look at the picture there are some mountains there. Craig Davis. Ah, on the side wall? 323

342 Professor Jiin-Song Tsai. Yeah, and people of wisdom love water. So that s what we are. Love here the moral and enjoy. So friends go on. Craig Davis. Ok, let s use water wisdom. Do we have any points of discussion or comments from Professor Tsai s opening summary? I forget which side we re on, we ll go right own the middle. Marty. Marty Adams. I ll try to wrap up without 10 comments from the last three groups. Craig Davis. Ok, go right ahead. Marty Adams. In agreement with much of what has been said, I believe that a seismic performance standard has to be set at a local level because it s really dictated by what the political environment and social environment will accept as a level of response and reliability, and cost. I think that s probably why there s not a national standard. First of all there s not a national risk in the United States, as we know here in California there is a risk. I think that certainly, as mentioned with Marilyn from East Bay MUD, when there is a disaster just as when there is any kind of problem, certainly security risk is a great example, there is an awareness and opportunity to have people and make investment and commit to a higher level of service when that s what s on their mind. It won t take but a year or two afterwards when people move on with their life and have other concerns. I think the reality is we have performance standards for water and power that we want to adhere to in our daily service; but yet we understand the reality that an earthquake is considered by the population to be an act of god and to be uncontrollable and that some level of degradation of service or outage is not only acceptable but expected. So I think that certainly the idea Professor Shinozuka in his performance evaluation, he acknowledged that maybe a certain percentage of pipes break or a certain percentage of recovery is established in a certain amount of time, it acknowledges the fact that you can not be bullet proof. So the question is: How long is too long?; How much is too much?; And how can you reasonably invest in shrinking those numbers down with any kind of reliability? That s my comment. Craig Davis. Any response to that? Seishi Nonaka. Japanese interpretation. Professor Jiin-Song Tsai. Any comment or anything want to say a question on opposite ocean? Yeah, please. David Lee. Marty, in response to your question how much is too much, how much is enough, you know in the East Bay MUD seismic improvement program, we presented to our board of directors four packages with associated costs and risk for them to consider. So I would say in our program is so called informed decision. It is not a consent, or you know a consent to be I think is to ask someone to agree with you; or informed disclosure is just tell them, but I guess our program we asked the board of directors to make a 324

343 decision with all of the available information. So that s why they paid, we had the so called packages. I am pretty sure a lot of the districts, utilities are doing that way too. We gave them four packages and they picked the third one. Professor Jiin-Song Tsai. Thank you Dave. Seishi Nonaka. Japanese interpretation. Professor Jiin-Song Tsai. Thank you. Yeah please. Norio Iijima. Translated by Seishi Nonaka. The cost associated with the strategic plan and also repair, we know that the customer end up paying some of it. However, the municipal electric companies do pay. In our country, in Tokyo, as far as Tokyo is concerned we have some identified costs that is to be paid by the customers and paid by the municipal electric distribution company. Is it a good idea to draw a line between customers and the municipal electric utility? Professor Jiin-Song Tsai. Thank you. It is quite clear that s a balance between your customer and your honorable government. I believe in Taiwan we have just a newest experience compared to here. In Chi-Chi earthquake we have had a lot of damage in central area and we have experienced people here, Mr. Wang, in talking about the recovery and how much it costs. It is beyond our expectations of the cost that we prepare for seismic matter. I will translate by myself. Ping-Hsin Wang. Translated by Professor Jiin-Song Tsai. A big number. It costs, how many zero? Nine zero. Six plus nine zero. In Taiwan dollar it s divided by thirty, so it is two plus eight zero US dollars. It s far beyond our expectation for budget, any kind of budget. So the government gave a special to the water something company to handle this kind of situation. Next one given of Taiwan area, which is the most serious damaged area. In that area it serves 0.8 million people for how many cubic meter? 1.1 million cubic meter per day of water. This is concerned about customer portion. After the earthquake, within one week the whole system in that area recovery and very little, or nearly none, complaints from customers. First reason is military involved and controlled all the orders and any kind of supplement route which can quite all those operation very rapid. One independent communication system was prepared by the water company which was very useful in that moment. Independent communication system. All subcontractor all employee are called out for this special mission within one day, I think using that communication system. To back up water source system use at that moment, the first one was 49 deep well system. The second was Lee Dam, how far from Taijon the other dam? Ok fine. Ok Japanese translation. Yeah Ok. Let me organize too many information. 49 well provide 0.1 million cubic meter per day and the back up dam provide 0.5 million cubic meter per day. So back up system start up for the disaster area within one day. Mitsuo Takasue and Seishi Nonaka. Japanese interpretation. 325

344 Professor Jiin-Song Tsai. Thank you, and thank you all. I think the time is allotted and thank you Mr. Wang and thank you Marty and I give the time to. Craig Davis. We need to be done by 4, no 3:20. We are 20 minutes behind schedule due to a 20 minute delay and so we ll account for that as being on schedule. I am a project manager and so that s the way things work. We are compressing one last item into the break. The last item which is I mentioned would be extremely brief, is extremely briefer now, on considerations of future workshops and we will at the moment eliminate the goal topic. I would like to take a short break just because it is really warm and this took a while. No more than 5 minutes. There are refreshments out there. Lets stretch and please talk with your colleagues or consider to yourself the idea of the importance to you for having To keep your table up, were keeping disasters from happening having future workshops along this subject as we have in the past and in this very successful one here. So let s take a break and we will be back here by 25 after. 326

345 Session Closing Craig Davis following the break. You all had a 3 minute homework assignment, or I guess a hall work assignment, and I would like to know, we will summarize by interest by a show of hands, because we have to move on. I wanted to have a group discussion but we will not have the opportunity to do that. There s a dead fly in here on my head, excuse me, I don t know how that happened. By a show of hands who is interested in carrying on this topic with future workshops in some location. Please let us know of an interest. Water System Seismic Improvements. Mitsuo Takasue. Japanese interpretation. Craig Davis. A show of hands please, please. From my perspective that is almost, there is a consultant that is not interested, that s almost unanimous. So we will take that as an extremely positive suggestion. The idea of when or where will be discussed amongst TWWA, JWWA, and AwwaRF in the very near future. We will not discuss that topic at the moment that will be informed to everybody later. Now I would like to move into the closing ceremony. I would like to start this and now I know I will be doing the final closeout too. 327

346 328

347 3 rd US-Japan Workshop on Water System Seismic Practices CLOSING REMARKS Craig A. Davis Workshop Coordinator Waterworks Engineering Los Angeles Department of Water and Power Jiin-Song Tsai Professor National Cheng Kung University Taiwan Water Works Association Kenei Ishii Director Water Works Engineering General Institute Japan Water Works Association Elizabeth Kawczynski Senior Account Manager American Water Works Association Research Foundation Glenn Singley Director Water Engineering and Technical Services Business Unit Los Angeles Department of Water and Power 329

348 3 rd US-Japan Workshop on Water System Seismic Practices 330

349 Closing Remarks Craig A. Davis Workshop Coordinator Waterworks Engineer, Los Angeles Department of Water and Power I would like to thank you all for attending and making this what I perceive as a very successful workshop. My expectations were far exceeded. You all performed great discussions. The information presented and all of the presentations were outstanding, even the ones that were not seismically oriented; all of you did a great job of relating the information to seismic issues and the importance of how all of this interacts. I thank you for that. I would like to personally than the two Japanese interpreters for doing the hardest work at this workshop. I have seen some interpretations, but I would say this was outstanding. They far exceeded my expectations. But I knew they would do well because they are engineers and they work for the LADWP. Jiin-Song Tsai Professor, National Cheng Kung University On behalf of the Taiwan Water Works Association We appreciate this opportunity and feel very honored to be invited the US - Japan Workshop on Seismic Performance. Before I came here we had a lot of discussions with TWWA concerning how to devote ourselves more to this big family. We also show our hand to carry on the next workshop. I just talked to Elizabeth Kawczynski during lunch, we have an idea to propose a major project in TWWA and also in our society to find more sponsors. In this project we, our group members, wish to study all the topics, and subjects, and interesting problems and questions in previous workshops and organize and assist in one of interesting topics. We wish to propose to TWWA and to invite people here to be involved in this project. We will do this probably in one year or two years I believe; we will have more discussion and do more organizing. Please excuse me; this is an idea that just popped up during lunch. In the workshop we are going to propose will become a final report type of workshop, we can discuss the things and come up with a conclusion as a documentation which is useful in the water business. In fact, seismic is one kind of risk in this business and whether we want to make it more profitable and more valuable, we want to solidify this idea in our workshop. I wish this dream may come true in the future. This is what we want to devote the TWWA to hold this workshop. This is my closing comments and thank you Dr. Davis. 331

350 Kenei Ishii Director Water works Engineering General Institute Japan Water Works Association On behalf of all the Japanese participants and Japan Water Works Association, I would like to express my gratitude to you. I deeply and sincerely appreciate you impeccable preparation and hospitality of this workshop. Now we realize there are some different points and aspects of the seismic countermeasures among the United States, Taiwan, and Japan. For instance the replacement of old pipelines, support of the trunk lines, and seismic assessment. I think it is important to know the deference. The knowledge of the difference will bring us new concepts or new ideas to develop the seismic problem in the future. Therefore, this is very precious; this is so precious that we have no chance other than this workshop to exchange the seismic information relating to waterworks with foreigners. Ah, American people speak English fluently, while it is hard for me to speak English. It is true that English is the standard language in the world, but what is spoken in English is not always standard. This is a new Japanese proverb. By the way, I hear the next workshop would be held in Taiwan at the opening ceremony; I am hoping. Taiwan is a very near country to Japan. I will dispatch our own members to Taiwan. I am looking forward to seeing you again in Taiwan. Thank you very much. Elizabeth Kawczynski Senior Account Manager American Water Works Association Research Foundation It is so good to see you all again and it is so good for us to get together to exchange ideas and perspectives on the hazards that we face. I know and I can feel that we all have learned something more from this workshop. I certainly want to take the opportunity to thank the project manager [Craig Davis] and all of his helpers; he couldn t have done it without them. We certainly appreciate the cooperation of the JWWA and TWWA for participating. Thank you very much to our subscribers for taking the time to make presentations; thank you to the consultants; thank you to everyone for a beautiful, wonderful experience. We certainly look forward to doing this all again somewhere. 332

351 Glenn Singley Director Water Engineering and Technical Services Business Unit Los Angeles Department of Water and Power Thank you Craig, I too want to add my appreciation to all of you who have participated. We are honored to have been able to host this program. I think we have all learned a lot from all the countries that have been involved and all the different agencies involved. We have had comic relief when our Japanese interpreters have interpreted our English and repeated English back to us, and Japanese and repeated Japanese back to us. I think they did a wonderful job, but of coarse I don t know; they could have been telling baseball scores or something as far as all I knew. We have been very privileged having all of you here. The weather cooperated; the smog cooperated and stayed away. We hope that all of you travel to your various destinations very safely and think very hard on the things that we learned, and we will be looking forward to the next time that we get together. Thank you. 333

352 334

353 Closing Ceremony For the Closing Ceremony Dr. Davis invited Mr. Ishii, Mrs. K., Mr. Iijima, and Mr. Singley to the front of the room. Mr. Ishii and Mrs. Kawczynski were presented with the Daruma, provided by compliments of Mr. Matsushita, to hold while Mr. Iijima and Mr. Singley each painted one eye. Following the Daruma eye painting ceremony, Dr. Davis officially closed the workshop. 335

354 336

355 3 rd US-Japan Workshop on Water System Seismic Practices APPENDICES Appendix I: Appendix II: Appendix III: Appendix IV: Appendix V: Appendix VI: Workshop agenda Discussion results Future Directions for Improvement in Water System Seismic Practices: Topics for Future Discussion and Development Important Aspect to Include in a Water System Seismic Improvement Program Survey of Important Aspects to Include in a Water System Seismic Improvement Program Los Angeles Department of Water and Power Facilities Tour Workshop Participants A-1

356 3 rd US-Japan Workshop on Water System Seismic Practices A-2

357 3 rd US-Japan Workshop on Water System Seismic Practices Appendix I: Workshop agenda A-3

358 3 rd US-Japan Workshop on Water System Seismic Practices A-4

359 3rd US-Japan Workshop on Water System Seismic Practices Appendix I: Workshop Agenda August 5, :00-8:00 PM Welcome Reception Hosted by AwwaRF New Otani Hotel, Rooftop Garden August 6, :00-9:00 AM Registration John Ferraro Building, LADWP Conference Center, A-level Continental breakfast provided at conference center 9:00-10:00 AM Opening Remarks (Conference Center) 9:00-9:10 Introduction Craig Davis, LADWP and Elizabeth Kawczynski, AwwaRF 9:10-9:25 Greetings Gerald Gewe, Los Angeles Department of Water and Power 9:25-9:40 Greetings Norio Iijima, Tokyo Waterworks Bureau 9:40-9:55 Greetings Ping-Hsin Wang, Taiwan Water Supply Corporation 10:00 10:15 Break (drinks and snacks provided) 10:15 12:05 Technical Session I * (Conference Center) Theme: Risk Assessment and Analysis I Chairpersons: Marilyn Miller (EBMUD) and Kenei Ishii (JWWA) 10:15 Seismic Upgrades for 20 Suburban Water Utilities in the San Jose Bay Area (G&E Engineering Systems Inc., Oakland, California) 10:35 Government Policies on Earthquake Disaster Prevention Measures Related to Water Supply in Japan (Health Service Bureau, Ministry of Health, Labor and Welfare, Tokyo, Japan) 10:55 Anti-Seismic Measures of Existing Water Supply Facilities - A Case Study of an Anti-Seismic Plan of Inagawa Water Treatment Plant (Hanshin Water Supply Authority, Kobe, Japan) 11:15 San Francisco Public Utilities Commission Capital Improvement Program (San Francisco Public Utilities Commission) 11:35 Seismic Measures for Waterworks in Yokosuka City (Yokosuka City Waterworks Bureau, Japan) A-5 John Eidinger Yoshihisa Iwasaki Keiichi Murakami Jeet Bajwa Takashi Furuya 11:55 Session Discussion (10 minutes) Chairpersons 12:05 1:00 Lunch, provided at Conference Center/Cafeteria * Session Recorders and Chairpersons are listed in the table in the back of the Agenda

360 3rd US-Japan Workshop on Water System Seismic Practices August 6, 2003 continued 1:00 2:50 Technical Session II * (Conference Center) Theme: Seismic Performances, Preparedness, and Readiness Chairpersons: Charles Pickel (Memphis) and Nobuhisa Suzuki (Japan Water Steel Pipe Assoc.) 1:00 Damages and Motion of Pipelines in Liquefied Ground (Japan Ductile Iron Pipe Association) 1:20 Seismic Evaluation of Water Supply System in Health Facilities (Kanazawa University, Japan) 1:40 The Research of Damages of Public Water Supply Pipelines during the Ji- Ji Taiwan Earthquake on September 21, 1999 (Taiwan Water Supply Corporation, Yuang-Shang, Taiwan) 2:00 Emergency Operation Planning - How Contra Costa Water is Building Earthquake Response Capabilities In Calm to Excel in Emergency (Contra Costa Water District, California) 2:20 Emergency Water Supply Facilities of Hachinohe Regional Water Supply Authority (Hachinohe Regional Water Supply Authority, Aomori, Japan) Toshio Toshima Masakatatsu Miyajima Ping-Hsin Wang Stephan Welch Kenetsu Kojima 2:40 Session Discussion (10 minutes) Chairpersons 2:50 3:10 Break (drinks and snacks provided) 3:10 5:00 Technical Session III * (Conference Center) Theme: Seismic Risk Management and Post Earthquake Recovery Chairpersons: Frank Collins (Parsons) and Hiroshi Yamada (Waterworks Bureau Tokyo Met. Gov.) 3:10 Seismic Damage Simulation of Distribution Pipeline Based on the Monitoring Data Collected by a Seismometer Network (Osaka Municipal Bureau, Japan) 3:30 Prioritization of the City San Diego Water Department s Capital Improvement Program (City of San Diego Water Department) 3:50 Seismic Practices Evaluation of Kobe Water System using Risk Management Approach (Kobe City Waterworks Bureau, Japan) 4:10 Knowledge Management in Engineering - a Business Improvement Methodology Applied to Seismic Risk Management (Thames Water Utilities, UK) 4:30 Emergency Restoration for Water Supply Following Earthquake Disasters in the City of Yokohama (Yokohama Waterworks Bureau, Japan) Koichi Murata Michael E. Conner Makoto Matsushita Jim Woodhams Kiyoshi Naito 4:50 Session Discussion (10 minutes) Chairpersons 5:00 Group Photograph (Location: Los Angeles County Music Center fountain) * Session Recorders and Chairpersons are listed in the table in the back of the Agenda A-6

361 3rd US-Japan Workshop on Water System Seismic Practices August 7, :30-8:00 Continental Breakfast (provided in front of auditorium, A-level) 8:00-9:00 AM Technical Session IV * (Auditorium) Theme: Los Angeles Water System and Tour Chairperson: Le Val Lund Tour Introduction John Ferraro Building, Auditorium, A-level 8:00-8:20 Introduction and Overview of LA Water System Martin Adams, Los Angeles Department of Water and Power 8:20-9:00 History of Los Angeles Water System Seismic Improvements Craig A. Davis, Los Angeles Department of Water and Power 9:00-9:15 Break (on your own in cafeteria) 9:15-9:30 Board Busses Hope Street 9:30-12:30 Travel to Magnolia Trunk Line, Discuss LA Water System in route - Marty Adams 10:00-10:30 Magnolia Trunk Line Construction Site, Evelyn Cortez-Davis 10:30-11:00 Travel to Van Norman Complex, Discuss LA Water System in route - Marty Adams 11:00-1:00 Tour Van Norman Complex facilities, Craig Davis and Marty Adams 1:00-1:30 Lunch, provided ( Candy Land picnic area, Van Norman Complex) 1:30-3:00 Tour Van Norman Complex facilities, Craig Davis, Marty Adams, Le Val Lund 3:00-3:30 Travel to Hollywood Water Quality Improvement Project, Discuss LA Water System in route - Marty Adams 3:30-4:30 Tour Hollywood Water Quality Improvement Project - Steven Cole 4:30-5:00 Travel to Hotel/LADWP, Discuss LA Water System in route - Marty Adams 5:00 Return to Hotel/LADWP office 6:00 8:30 Banquet Hosted by AwwaRF and LADWP New Otani Hotel, Ballroom 1 6:00-7:00 Reception - Hosted by MCEER 7:00-8:00 Banquet (Ballroom 1) 8:00-8:30 Technical Session V * (Ballroom 1) -- Keynote Speaker Theme: System Performance and Management Lessons Learned from the World Trade Center Disaster for Water Supply Management Thomas D. O Rourke, Cornell University * Session Recorders and Chairpersons are listed in the table in the back of the Agenda A-7

362 3rd US-Japan Workshop on Water System Seismic Practices August 8, :30-8:00 Continental Breakfast (provided at conference center) 8:00-9:50 AM Technical Session VI * (Conference Center) Theme: Risk Assessment and Analysis II Chairpersons: David Lee (EBMUD) and Masakatsu Miyajima (Kanazawa Univ.) 8:00 Maintenance Management System for Lowering Possible Seismic Damages onto Water Works Facilities (National Cheng Kung University, Tainan, Taiwan) 8:20 An Overview of the Metropolitan Water District of Southern California s Seismic Program (Metropolitan Water District of Southern California, Los Angeles) 8:40 A Fast Simulation Method for Predicting Seismic Responses of an Extensive Water Distribution Network (Japan Water Steel Pipe Association, Kawasaki, Japan) 9:00 Multi-Hazard Risk Assessments, Elements in Common with Seismic, Security, and Other Risk Studies (ABS, Seattle, Washington) 9:20 A Study on the Development of a Backup System in a Big Urban Area (Japan Water Research Center, Tokyo, Japan) Jiin-Song Tsai Clark Sandberg Nobuhisa Suzuki Don Ballantyne Yoshiharu Sorakuma 9:40 Session Discussion (10 minutes) Chairpersons 9:50 10:10 Break (drinks and snacks provided) 10:10 11:40 Technical Session VII * (Conference Center) Theme: Seismic Resistant Design Chairpersons: Endi Zhai (Group Delta Consultants) and Cheryl Chi (National Cheng Kung Univ.) 10:10 Reinforcement Work of the Embankment of the Yamaguchi Reservoir (Waterworks Bureau Tokyo Metropolitan Government, Japan) 10:30 Seismic Upgrade of East Bay Municipal Utility District's Mokelumne No. 3 Aqueduct (East Bay Municipal Utility District, Oakland, California) 10:50 Seismic-Proof Design for the Structures of the Sinanogawa Water Treatment Plant in Niigata (Niigata Waterworks Bureau, Japan) 11:10 A Practical Approach to Mitigation of Earthquake Pipeline Damage (Seattle Public Utilities, Washington) Hiroshi Yamada Bruce Maison Hitoshi Hasegawa William Heubach 11:30 Session Discussion (10 minutes) Chairpersons 11:40 12:40 Lunch, provided at Conference Center/Cafeteria * Session Recorders and Chairpersons are listed in the table in the back of the Agenda A-8

363 3rd US-Japan Workshop on Water System Seismic Practices August 8, 2003 continued 12:40 2:40 Technical Session VIII * (Conference Center) Theme: Panel Discussion Important Aspects to Include in a Water System Seismic Improvement Program Introduction: Craig Davis, Los Angeles Department of Water and Power Panel: David Lee, East Bay Municipal Utility District Masakatsu Miyajima, Kanazawa University Masanobu Shinozuka, University of California, Irvine Makoto Matsushita, Kobe Water Works Bureau Jiin-Song Tsai, National Cheng Kung University Seismic Resistant Design Risk Assessment and Analysis Seismic Risk Management Seismic Preparedness and Readiness Seismic Performances and Post-Earthquake Recoveries Chairpersons: Craig Davis, Los Angeles Department of Water and Power Jean-Pierre Bardet, University of Southern California 2:40 3:00 Break (drinks and snacks provided) 3:00 3:25 Closing (Conference Center) 3:00-3:05 Craig A. Davis 3:05-3:10 TWWA Jiin-Song Tsai 3:10-3:15 JWWA Kenei Ishii 3:15-3:20 AwwaRF Elizabeth Kawczynski 3:20-3:25 LADWP Glenn Singley Workshop Closing Craig Davis * Session Recorders and Chairpersons are listed in the table in the back of the Agenda A-9

364 3rd US-Japan Workshop on Water System Seismic Practices Table of Session Chairpersons and Recorders. Technical Session Chairpersons Recorders Session I: Risk Assessment and Analysis I Marilyn Miller, East Bay Municipal Utility District Kenei Ishii, Japan Water Works Association Don Ballantyne, ABS Session II: Seismic Performances, Preparedness, and Readiness Charles Pickel, Memphis Light, Gas, and Water Nobuhisa Suzuki, Japan Water Steel Pipe Assoc. Endi Zhai, Group Delta Consultants Session III: Seismic Risk Management and Post Earthquake Recovery Frank Collins, Parsons Hiroshi Yamada, Waterworks Bureau Tokyo Metropolitan Government Marilyn Miller, East Bay Municipal Utility District Session IV: Los Angeles Water System Le Val Lund, Consultant Frank Collins, Parsons Session V: System Performances and Management Session VI: Risk Assessment and Analysis II David Lee, East Bay Municipal Utility District Masakatsu Miyajima, Kanazawa University Session VII: Seismic Resistant Design Endi Zhai, Group Delta Consultants Cheryl Chi, National Cheng Kung University Session VIII: Panel Discussion Important Aspects to Include in a Water System Seismic Improvement Program Glenn Singley None Craig Davis, Los Angeles Department of Water and Power Jean-Pierre Bardet, University of Southern California John Eidinger, G&E Engineering Systems Inc. Tetsuo Tobita, Kyoto University Le Val Lund, Consultant Frank Collins, Parsons A-10

365 3 rd US-Japan Workshop on Water System Seismic Practices Appendix II: Discussion results Future Directions for Improvement in Water System Seismic Practices: Topics for Future Discussion and Development A-11

366 3 rd US-Japan Workshop on Water System Seismic Practices A-12

367 3rd US-Japan Workshop on Water System Seismic Practices Appendix II: Discussion Results Future Directions for Improvement in Water System Seismic Practices: Topics for Future Discussion and Development The workshop had a primary focus of encouraging discussions between all organizations and countries; Ten-minute formal discussion periods were reserved at the end of each session along with several scheduled break times to encourage informal discussion, and a panel discussion was held in Session VIII with the intent of including the entire workshop group in an open focused discussion. A listing of chairpersons and recorders is provided with the agenda in Appendix I. Chairpersons lead and encouraged the organized discussion periods. Written records were prepared for each session by pre-assigned recorders to document important aspects brought up in the presentations and during the discussions. The written recordings were made with the intent of aiding in identifying: 1. Topics in need of further discussion and development, which may also potentially serve as a focus of future workshops (summarized in this appendix). 2. Information useful for creating a list of important aspects to include in a water system seismic improvement program (summarized in Appendix III). The following two sections (identified as Panel Discussion Session VIII and Sessions I to VII Recordings ) summarize results of recordings made concerning the topics that are in need of further discussion and development. These topics were principally identified based on the level of interest demonstrated during the discussions, and may also be considered as future directions for improvements in water system seismic practices. The panel discussion provided the greatest opportunity for information exchange between all workshop participants; consequently these recordings serve as the primary outcome for major topics of future interest resulting from this workshop. However, as the panel discussion transpired it was clear that there are many additional topics of interest to be discussed, but there was not sufficient time. Therefore, results from the panel discussion should not be considered an exhaustive list of topics. To complement the panel discussion results, the recordings made from the first seven sessions were also summarized and placed under a separate section in this appendix. These results present a consolidation of all the recordings made from each of the recorders, summarized by major topics. The topic results are presented in an outline form, but are not intended to be listed in any priority order. Summary: The results show that the US, Japan, and Taiwan have similar seismic improvement concerns. In some cases organizations from the different countries have similar approaches to evaluating and resolving problems, such as design and evaluation of dams, reservoirs, treatment facilities, etc. In other cases there are differing approaches to problems, such as available post-earthquake water supplies, pipe replacement programs, etc. However, even within the differing ideas the general concept remains the same; For example, all agree that post-earthquake emergency supplies must be available; most US organizations approach the problem by trying to ensure supplies through the existing normal operating system, where as several Japanese organizations approach the problem by adding separate tanks, cisterns, and other means of holding water in the event an earthquake results. There was not adequate A-13

368 3rd US-Japan Workshop on Water System Seismic Practices information available to understand how Taiwan approaches this issue. As another example, all countries agree that certain pipes, such as cast iron, behave poorly in earthquakes and the bad performance must be considered as a part of the improvement program; whether it is from pipe replacement, post earthquake response and repair, improving system performance by adding redundancy and isolation capabilities, etc. It is important to understand the similarities in approaches to resolving seismic problems. However, even greater value may be attained from these workshop gatherings by inquiring further into the differing approaches to solve the problems, providing opportunity to identify the common aspects, as shown in the above examples, then further discussing the differing methods used to solve problems to create greater learning possibilities. A-14

369 3rd US-Japan Workshop on Water System Seismic Practices Panel Discussion, Session VIII August 8, 2003 Panel Member David Lee, East Bay Municipal Utility District Masakatsu Miyajima, Kanazawa University Masanobu Shinozuka, University of California, Irvine Makoto Matsushita, Kobe Water Works Bureau Jiin-Song Tsai, National Cheng Kung University Topic Seismic Resistant Design Risk Assessment and Analysis Seismic Risk Management Seismic Preparedness and Readiness Seismic Performances and Post-Earthquake Recoveries Seismic Resistant Design: There was discussion concerning practical pipe design to withstand earthquake effects, how much money should be spend to prevent pipeline damage, and what factors should be considered. Cost effectiveness is a function of population, density, and multiple benefits. Japanese agencies are currently replacing old pipelines on a seemingly higher rate than the US agencies; it was stated that Japan is replacing pipe on a 50 to 100 year cycle and the US on a 200± year cycle. Japanese agencies are installing earthquake resistant pipelines (S-type joints) in known seismic hazard areas; US and Taiwan agencies have no comparable product. Japan has pipe seismic design guidelines and the US and Taiwan does not. The needs to provide redundancy, make key supply points earthquake resistant, break the distribution system into blocks and supply storage into many parts were discussed. Regional redundancy may be more important than localized earthquake resistance. Topics for discussion: 1) Regional redundancy and back up in water supply systems. 2) US/Taiwan design guidelines for pipe joint seismic design, including licensing Japan ductile iron pipe S-joint in other countries. 3) Pipeline replacement cycles for seismic improvement. Risk Assessment and Analysis Results of risk assessment and analysis serve as the basis for determining risk management, emergency response, and post-earthquake recovery strategies, as well as seismic resistant design. There was discussion concerning how different agencies approach the seismic vulnerability question: Do agencies evaluate specific components or the overall system/network? The question was posed to the group with a show of hands. Most agencies evaluate the overall system performance. There was also discussion on the needs and use for near real time ground motion assessments in post-earthquake response. Topics for discussion: 1) Overall water system seismic performance assessments (vs. specific components). 2) Needs and use of near real-time ground motion assessments. A-15

370 3rd US-Japan Workshop on Water System Seismic Practices Seismic Risk Management There was discussion on the use of seismic performance objectives and criteria for the water supply industry and use of probability in defining the performance objectives. There are no national or state established water system seismic performance criteria in the US, Japan, or Taiwan. Some local utilities have established some performance criteria. The importance of standardization was identified. There were comments concerning national vs. local standards and performance objectives. Seismic improvement programs are an on-going program of risk management. SCADA systems are useful for determining earthquake damage. Topics for discussion: 1) Seismic performance criteria and objectives. 2) US/Japan/Taiwan water system seismic design standards/ guidelines. 3) Considerations of national vs. local standards. Seismic Preparedness and Readiness It was noted seismic preparedness is a combination of available system redundancy, in place recovery strategy, and a good pre- and post-earthquake communication plan. Also, as noted by almost all agencies represented, it is important to emphasize the multi-benefit purposes (e.g., water quality, increased capacity, etc.) of doing seismic upgrades. In almost all cases the seismic upgrades improve normal operation capacity, redundancy, and reliability, making water systems more robust during the day to day operations and for earthquake. Other points of discussion included mutual assistance agreements, response and recovery plans, and prioritization of seismic improvements and post-earthquake recoveries. Topics for discussion: 1) Comprehensive seismic readiness/recovery strategy. 2) Multi-benefit aspects of seismic improvements. Seismic Performance and Post-Earthquake Recovery Seismic performance and recovery can be viewed as consisting of three layers: (1) seismic mission, (2) service goals, and (3) system performance. There was a continuation of discussion concerning the development of seismic performance goals/standards. Each agency typically develops local performance goals. The goals are discussed with all stake holders (customers, operators, managers, engineers, etc.) taking into account social, political, financial, seismic hazards, and system requirement considerations and a consensus of the appropriate performance goal is reached. The term informed decision was identified as being appropriate for this process (as opposed to consent or informed disclosure ). There was also discussion concerning the large damage repair costs associated with urban earthquakes like the Chi-Chi earthquake and how much the public should be charged directly for mitigating earthquake damage in municipal utilities. Topics for discussion: 1) How to establish localized seismic performance and post-earthquake recovery goals. 2) Financial considerations for seismic preparations and post-earthquake recovery. A-16

371 3rd US-Japan Workshop on Water System Seismic Practices Sessions I to VII Recordings August 6, 7, and 8, The importance of post-earthquake water supplies. a. Emergency response and recovery water supplies must be available. b. Most US organizations seem to approach the problem by trying to ensure supplies through the existing normal operating system. c. Several Japanese organizations approach the problem by adding tanks, cisterns, and other means of holding water in the event an earthquake results. d. The value of adding water storage for post-earthquake response was questioned by some participants. How should this be evaluated and determined? 2. The amount of money to be spent on mitigation to reach a post-earthquake water outage time goal. What should the post-earthquake water outage time be? a. The United States seem to be concerned with water outage times on the order of hours to days. b. The Japanese seem to be concerned with water outage times on the order of weeks to months. 3. Seismic improvements need to be performed in collaboration with other necessary system changes and improvements (e.g., water quality, aged and degrading facilities, system growth, increased capacity, etc.). a. This is a common issue in the US, Japan, and Taiwan. b. Basic planning question: Should the seismic improvements be the primary goal and the other infrastructure improvements be secondary and incorporated with the seismic improvement program; or will the focus be primarily on infrastructure upgrades and the seismic improvements be performed as a secondary opportunity? 4. What is a practical rate of pipeline replacement and what should drive the need? a. The Japanese seem to place a high priority on pipeline replacement for seismic and other issues. b. The United States do not seem to place a high importance on pipe replacement for seismically related issues. c. It would be appropriate to evaluate to pipe rehabilitation and replacement rate in the United States for reasons other than seismic (e.g., several organizations have a high rate of rehabilitation for improving water quality and old degraded pipes without incorporating seismic concerns) and make direct comparisons with the Japanese organizations. d. Value may be attained from better understanding the deviating philosophies between the countries regarding pipe replacement. 5. Practical solutions are needed for mitigating effects of pipeline damage. a. This is a common issue in the US, Japan, and Taiwan. b. What are potential solutions that can be considered? 6. Financial concerns bring the greatest problems to implementing and completing a water system seismic improvement program. A-17

372 3rd US-Japan Workshop on Water System Seismic Practices a. This is a common issue in the US, Japan, and Taiwan. b. How should a program be financed? c. What should be financed and how should it be determined? d. What are the factors competing with seismic improvements? 7. Japan is establishing a financial subsidy program specific to water system earthquake improvements. a. The United States and Taiwan do not have a similar subsidy program. 8. Customer/rate payer input is needed for identifying the amount of money to be spent for seismic improvements. a. This is a common issue in the US, Japan, and Taiwan. b. What are the primary issues and how should this be accomplished? 9. Water systems need to work closely with organizations and facilities where water needs are critical following an earthquake (e.g., fire departments, hospitals). a. This is a common issue in the US, Japan, and Taiwan. b. What factors need to be considered? 10. Risk and vulnerability assessments and associated models are needed to prioritize and implement appropriate seismic improvements. a. This is a common issue in the US, Japan, and Taiwan. b. What are appropriate assessments to perform and how should they be modeled and implemented? c. Below are several recorded notes related to this topic: Security vulnerability assessments can build on seismic vulnerability analyses. The same for Y2K analyses. There is commonality among the elements of vulnerability and mitigations of vulnerability. The information becomes more valuable the more you use it. Multi-hazard risk assessments, accounting for hazard probability, facility vulnerability, loss consequence, and correlation factors, are useful for quantifying, screening, and prioritizing system risks and improvements based on potential impacts from different hazards (e.g., earthquake, flood, security, wind, etc.). Other analysis methods such as fault-tree and other cost-benefit aid in quantifying the system reliability and justifying the mitigation. Having a model that simulates seismic impacts on water systems could be used as a tool for managing immediate post-earthquake response. The model could be run with information on the actual earthquake magnitude and location to show rough locations of pipe breaks and other damage. This information could be used to direct damage assessment teams and to begin immediate planning for resources to restore the system. 11. Implementing a broad based seismic improvement program requires the management of risks, hazards, knowledge, emergency response, system operations, priorities, budgets, resources, etc. of seismic issues and their relation to other water system needs. A-18

373 3rd US-Japan Workshop on Water System Seismic Practices a. This is a common issue in the US, Japan, and Taiwan. b. What are appropriate methods and tools for managing water system seismic improvements? c. Below are several recorded notes related to this topic: Risk management is a useful tool for prioritizing water system improvements in relation to the seismic and other risks and improvement needs. There is a holistic nature to risk management. Risk management looks at multiple risks all at once. This also applies to prioritization of mitigations. This may provide the best approach for customers and it increases flexibility for unanticipated risks. It is also an on-going process, not an end-point Knowledge management provides an approach to organizing and sharing the information developed in the US-Japan workshops. Potential components: o A who s who database with photos, links to papers and presentations. o A web site to access the information Knowledge management could be in the form of joint training and planning among water agencies, particularly those in close geographic proximity to maximize mutual assistance efficiency. Maintenance management approaches are being investigated to determine applicability for improving water system seismic performance. It is important to be proactive in seismic improvements and not just be reactive and plan to repair everything after an earthquake try to prevent damage and also understand that damage will occur so also plan to repair damage following an earthquake. o Plan ahead and provide system redundancy and isolation capabilities. 12. Training in emergency restoration is needed. A-19

374 3rd US-Japan Workshop on Water System Seismic Practices A-20

375 3 rd US-Japan Workshop on Water System Seismic Practices Appendix III: Important Aspect to Include in a Water System Seismic Improvement Program A-21

376 3 rd US-Japan Workshop on Water System Seismic Practices A-22

377 3rd US-Japan Workshop on Water System Seismic Practices Appendix III: Important Aspects to include in a Water System Seismic Improvement Program Information brought forth during the workshop was assembled, processed, and summarized to provide a listing of goals and tasks, included in Tables 1 and 2, that are important for a broad based water system seismic improvement program. The workshop brought together numerous water system practitioners to discuss earthquake related issues, and provided an excellent opportunity to identify aspects that each of them and their respective organizations consider important to include in a water system seismic improvement program. Table 1 lists 6 seismic improvement program goals. Table 2 lists 32 primary tasks and several subordinate tasks that are important for accomplishing the goals; the primary tasks are general in nature and considered important for a seismic improvement program. Whereas the subtasks in Table 2 are more specific and important for accomplishing the primary task, but may not be applicable for all water organizations. The goals and tasks are not in any priority order nor considered independent; they must be accomplished in an integrated approach to develop a water system that can perform adequately following earthquakes. Results from many of the tasks will need to be combined to perform additional tasks in support of all the goals. The tasks are grouped under the listed goals in Table 2 only due to their affinity and natural relationship with a particular goal, but in many cases serve to support all the goals. All of the aspects show a clear focus of: The purpose of a water system seismic improvement program is to ensure the safe provision of water following an earthquake. Tables 1 and 2 are intended to provide goal and task summaries useful for accomplishing a broad based water system seismic improvement program. The tables are not intended to be a comprehensive and exhaustive list of goals and tasks; they should be considered a working list of items useful for consideration in a seismic improvement program that can be modified as additional goals and tasks are identified and further knowledge on the subject is gained. Even though Tables 1 and 2 provide extensive lists and most likely cover the majority of important aspects, it would be inappropriate to assume that all aspects important for inclusion in a water system seismic improvement program could be identified from this single workshop. Additional documents, including past workshop proceedings, are recommended to be reviewed to provide additional items for inclusion in Tables 1 and 2. Seismic Improvement Programs for water systems require very broad approaches and the inclusion and integration of many different components including, management, risk analysis, numerous engineering fields, geology, seismology, technical research, emergency preparedness and response, socio-economic issues, financial, community involvement and development, working knowledge of specific water operations and field crews, and so on. The information in Tables 1 and 2 was processed by the editor by first performing a thorough review of the survey responses presented in Appendix IV, summarizing the information into goals and tasks using an affinity process, then adding supplemental information from the workshop opening statements, papers, presentations, and discussions. A-23

378 3rd US-Japan Workshop on Water System Seismic Practices Table 1. Goals for Water System Seismic Improvement Program Goal Description 1 Provide adequate post-earthquake water supply throughout service area. 2 Reduce earthquake damage to facilities. 3 Ensure minimum level system functionality and rapid system recovery. 4 Achieve a rapid emergency response. 5 Accomplish a well planned, cost-effective, and publicly responsible seismic improvement program to ensure public safety. 6 Continually develop and improve earthquake disaster prevention capabilities *. * Items included in addition to survey responses. Information obtained from the survey responses is not marked. A-24

379 3rd US-Japan Workshop on Water System Seismic Practices Table 2. Tasks for Achieving Water System Seismic Improvement Program Goals Task Description Goal 1: Provide adequate post-earthquake water supply. 1 Forecast water outage time 2 Estimate post-earthquake water supply needed * Account for: a. Supply source outage time * b. Water loss through broken pipes and damaged facilities * 3 Assure water availability throughout service area a. Install infrastructure to isolate and/or store water for post-earthquake usage in order to assure water availability throughout system * i. Isolate pipes and use as underground emergency storage tanks ii. Install separate emergency water storage tanks and cisterns throughout system accessible within limited distance from every point within service area * b. Isolate tanks and reservoirs to keep from draining * c. Evaluate reliability of water supply from sources (i.e., wholesale agencies, aqueducts, storage reservoirs, etc.), improve as necessary d. Prepare for mobile water deliveries and temporary water supply systems * e. Develop and install emergency backup connections to other systems * 4 Prepare for post-earthquake water deliveries; reduce impacts from lack of stored emergency water (e.g., bottled water, portable tanks, water trucks, distribution facilities, etc.) a. Ensure adequate supplies, facilities, personnel, etc. to accomplish this task 5 Establish post-earthquake system operation requirements a. Set system performance as a priority, build upon individual components b. Limit water loss from damaged facilities c. Prevent secondary disaster impacts (e.g., fire, disease, etc. following earthquake) * Goal 2: Reduce earthquake damage to facilities. 6 Evaluate, strengthen, and upgrade existing facilities a. Dams and reservoirs b. Purification plants, pumping stations, etc. c. Pipelines and pipe network i. Replace cast iron with special seismic ductile iron pipe ii. Connect to multiple supply sources iii. Replace old pipes susceptible to corrosion and having poor earthquake resistance with new stronger and corrosion resistant pipes (e.g., Tokyo replacing lead service connection pipes with stainless steel) * iv. Use flexible pipe joints * 7 Design all new facilities for seismic resistance 8 Provide adequate and continued facility maintenance to help safeguard against seismic damage (in addition to facility seismic upgrades) * A-25

380 3rd US-Japan Workshop on Water System Seismic Practices Task Description Goal 3: Ensure minimum level system functionality and rapid recovery. 9 Identify seismic and geo-hazards (landslides, liquefaction, fault movement, etc.) 10 Model system performance a. Revisit periodically and compare with new knowledge gained since last system model 11 Identify system seismic vulnerabilities a. Inspect system components (repeat periodically) * b. Provide accurate supply source, transmission, and distribution assessments 12 Prepare damage estimate(s) a. Determine probability of damage b. Use GIS and fragility relationships as appropriate 13 Estimate the probability for continued operation 14 Identify needs for system improvements and reconstruction 15 Mitigate component damage effects on system functionality (e.g., damage to pipelines, reservoirs, filtration plants, pumping stations, etc.) a. Provide seismic resistant power supply (normal and backup) b. Implement block distribution system c. Provide system redundancy to expected damage areas i. Water storage (as much as possible) ii. Supply and distribution pipelines iii. Utilize multiple water supply sources/points d. Provide isolation capabilities within the system i. Install remote valve operation capabilities e. Ensure continued and uninterrupted system operation in lightly damaged and undamaged regions * Goal 4: Achieve a rapid emergency response. 16 Forecast recovery crews needed a. Ensure adequate response and recovery support staff * 17 Develop system repair and operation plans [for expected damage] a. Stockpile material supplies and prepare storage sites i. Prepare a dispersed stock of materials, personnel, equipment 18 Develop emergency preparedness and response plans a. Coordinate emergency support with other cities and water utilities b. Incorporate community emergency planning c. Establish mutual aid scheme (formal and informal relations with other organizations) d. Coordinate post-earthquake response with municipal department an emergency service agencies (e.g., fire, police, city, county, state agencies) e. Prepare to provide food and water rations to repair crews * f. Develop damage assessment teams (with pre-assigned reporting location) * 19 Provide disaster prevention training (include other cities and agencies) a. Perform periodic emergency drills (annually) A-26

381 3rd US-Japan Workshop on Water System Seismic Practices Task Description Goal 5: Accomplish a well planned, cost-effective, and publicly responsible seismic improvement program to ensure public safety. 20 Establish seismic improvement/mitigation program objectives 21 Establish risk management structure 22 Perform cost-benefit analysis a. Estimate cost for infrastructure improvements b. Provide an economic justification for improvements (to management, politicians, and customers), be fiscally responsible 23 Achieve a multi-year fiscal commitment (e.g., 10 year financial plan) 24 Estimate cost for post-earthquake response and recovery * 25 Attain adequate funding for long term seismic improvement program * a. Understand that a proper seismic improvement program requires major investment over long period of time * b. Include infrastructure and post-earthquake recovery costs c. Consider bond sales, water revenue, government grants and subsidies * 26 Communicate with customers and public a. Understand customer s needs and interests b. Keep appraised of social changes and demands c. Gain customer support and acceptance of projects and programs 27 Prioritize improvements to be made a. Coordinate seismic improvements with other system improvements and upgrades * (e.g., deteriorating facilities, water quality improvements, etc.) 28 Establish plan for performing work to improve facilities and system performance (design, construction, emergency response, etc.) Goal 6: Continually develop and improve earthquake disaster prevention capabilities *. 29 Provide education, training, and knowledge exchange * a. Learn from past earthquake experiences * b. Learn from other water organization experiences * c. Network with others who are working on water system seismic improvement aspects * d. Train managers, engineers, operators, and field personnel on seismic issues * e. Provide staff development * 30 Perform engineering development and research activities related to water system seismic performance * 31 Develop water system guidelines and goals for seismic improvement aspects 32 Apply knowledge learned from disasters, management systems, and risk assessments other than earthquakes and water systems for the improvement of water system seismic practices * * Items included in addition to survey responses. A-27

382 3rd US-Japan Workshop on Water System Seismic Practices A-28

383 3 rd US-Japan Workshop on Water System Seismic Practices Appendix IV: Survey of Important Aspects to Include in a Water System Seismic Improvement Program A-29

384 3 rd US-Japan Workshop on Water System Seismic Practices A-30

385 3rd US-Japan Workshop on Water System Seismic Practices Appendix IV: Survey of Important Aspects to Include in a Water System Seismic Improvement Program A pre-workshop survey, presented at the end of this appendix, was sent to all workshop participants. The survey requested descriptions of up to 5 aspects considered important to include in a water system seismic improvement program and to categorize each aspect into one of the following main topics: 1. Risk Assessment and Analysis 2. Risk Management 3. Seismic Preparedness and Readiness 4. Seismic Resistant Design 5. Seismic Performances 6. Post-Earthquake Recoveries and Events 7. Other, specify a topic The survey was sent to all 26 workshop participants and 13 survey responses were received (6 from Japan and 7 from the United States), giving a total of 64 different aspects considered important to include in a water system seismic improvement program from 13 different organizations. Table I presents the basic survey response statistics. The survey responses are summarized by Organization (O) in Tables O-1 to O-13 and by main Topic (T) in Tables T-1 to T-6. As shown in Table I, the survey responses were fairly well distributed among the 6 main topics, with 2 additional topics added by the East Bay Municipal Utility District (EBMUD) and the Kobe City Waterworks Bureau (KCWB) as discussed later in this appendix. The Japanese and United States had a generally similar number of aspects identified for each main topic, except for the two topics of Seismic Performances and Post- Earthquake Recoveries and Events; there were four United States and no Japanese responses for Seismic Performances and four Japanese and no United States responses for Post- Earthquake Recoveries and Events. The difference in response statistics between the two countries my simply be a difference in interpretation of the main topic meaning, but the aspects identified by each country were similar and therefore were grouped together in Table T-5. Hereafter, these two main topics will be discussed as a single topic. Table I shows that the majority of aspects, 28% of all aspects identified, were categorized under the main topic of Risk Assessment and Analysis. The combined topics for Seismic Performances and Post-Earthquake Recoveries received the minimum number of responses, 12.5% of all aspects identified. The other three main topics received between 16% and 23% of all aspects identified. Table O-1 shows the EBMUD identified an additional aspect as Communication with Customers and Public. Table O-9 shows that the KCWB identified an additional aspect as Adaptation to Paradigm Shift and Social Changes. The presentation in by Mr. Matsushita from the KCWB in Sessions III and VIII described this idea, among other things, as a necessity for the Water Supplier to understand society and the customer s needs and interests, how they may change over time, and how this may be achieved through proper communication. Thus, these two organizations are identifying similar aspects and indicate A-31

386 3rd US-Japan Workshop on Water System Seismic Practices that an additional main topic of Customer Service and Satisfaction should be included within seismic improvement programs. Additional comments were provided by the KCWB, at the bottom of Table O-9, indicating how seismic improvements can be categorized under several, if not all, of the main topics listed. A similar aspect was identified by the EBMUD in Table O-1 (Aspect 1), stating that Risk Assessment and Management are part of the same process. These and other similar comments provided during the workshop identify how important it is to take a comprehensive integrated approach to a seismic improvement program and that the categorization of a seismic improvement into one of the major topics is primarily a function of the authors perspective on a particular seismic improvement and/or the phase of work that is being considered. The categorization of the different aspects into the main topics was only intended to be a tool for acquiring information on seismic improvements. The aspects listed in the following tables were determined from actual practicing engineers performing seismic improvements to water organizations in the United States and Japan. They were determined from the practitioners: Direct experience from earthquakes that have affected their systems, Understanding seismic hazard, system operational needs, and earthquake engineering aspects of system and components. Based on the survey results, the following conclusions are made: 1. The survey provided very important information useful for identifying aspects needed for inclusion in a broad based seismic improvement program. 2. The survey responses indicate that the U.S. and Japan have similar interests and concerns regarding seismic improvement aspects 3. The survey responses provided the base information used for identifying the goals and primary tasks needed for a comprehensive water system seismic improvement program described in Appendix III. 4. All of the aspects show a clear focus toward the primary objective of a water system seismic improvement program, which is to ensure the safe provision of water following an earthquake. 5. At present time, many organizations are concerned with Seismic Risk Management, Assessment and Analysis aspects, which are apparent by the large number of responses to these two topics (33 of 64 responses, 52% of total responses). 6. Customers need to be included in seismic improvement program process. 7. There are several aspects indicating that the U.S. and Japan organizations have similar approaches to resolving seismic related problems, and others where there are differing approaches. 8. It is important to take a comprehensive approach in developing a seismic improvement program. A-32

387 3rd US-Japan Workshop on Water System Seismic Practices Table I. Survey Statistics for Different Seismic Improvement Program Topics Seismic Risk Post- Seismic Seismic Preparedness Assessment Earthquake Organization Resistant Risk and and Recoveries Design Management Readiness Analysis and events East Bay Municipal Utility District Los Angeles Department of Water and Power Memphis Light Gas and Water City of San Diego Water Department San Francisco Public Utilities Commission 20 Suburban Water Utilities, San Jose Seattle Public Utilities Hachinohe Regional Water Supply Authority Kobe City Waterworks Bureau Niigata Waterworks Bureau Osaka Municipal Waterworks Bureau Bureau of Waterworks, Tokyo Metropolitan Government Yokohama Waterworks Bureau Seismic Performances Other topics Total (13) A-33

388 3rd US-Japan Workshop on Water System Seismic Practices Survey Results from Different Organizations Table O-1. Survey result from East Bay Municipal Utility District Aspect Seismic Improvement Topic Reason for Importance 1 Risk Assessment & Management These two aspects are really part of the same process. They need to be iterative as part of the decision-making process in planning an improvement program. Priority-setting based on post-earthquake performance goals enables an agency to define the system improvements in an efficient manner. 2 Planning for coordination and outreach to cities, counties, fire agencies, emergency response agencies. As EBMUD s Seismic Improvement Program (SIP) nears completion, we will be working with these entities more closely. Seismic Preparedness & Readiness Water system improvements need to a part of overall community emergency planning. 3 Revisit original SIP modeling of system performance. Seismic Performance System performance needs to be a priority, building on performance of individual components. It should also be revisited regularly to incorporate lessons learned in studies of post-earthquake responses in other areas. 4 Claremont Corridor Seismic Improvements upgrades to a critical water distribution tunnel. Other Communication with customers and public Table O-2. Survey result from Los Angeles Department of Water and Power Involvement of customers all through the process is important to gaining initial support and to maintaining the overall goal in the face of site-specific opposition to individual seismic improvements Aspect Seismic Improvement Topic Reason for Importance 1 Provide redundancy in water supply pipelines in critical supply areas 2 Provide as much in-city water storage and sources of supply as possible Risk Assessment and Analysis Risk Assessment and Analysis 3 Evaluate and strengthen water storage dams Seismic Resistant Design 4 Perform inspection and assessment of existing system components and identify weaknesses Risk Assessment and Analysis Allows multiple sources for water transmission to damaged area of distribution network. Proved useful following 1971 and 1994 Mw 6.7 earthquakes Allows multiple sources for water supply following earthquake damage to distribution network. Proved useful following 1971 and 1994 Mw 6.7 earthquakes. Public safety and ensure post-earthquake water supply. Severe damage in 1971 prior to performing. No significant damage in 1994 after strengthening dams in LA. Identify water system components vulnerable to earthquake damage. 5 Prioritize facilities to be strengthened Risk Management Complete the most critical facilities first A-34

389 3rd US-Japan Workshop on Water System Seismic Practices Table O-3. Survey result from Memphis Light Gas and Water Aspect Seismic Improvement Topic Reason for Importance 1 Provide new butterfly header valves with remote operators that can be operated from a SCADA system. The butterfly valves will either replace or be added existing header. Seismic Preparedness and Readiness Allow for electric operators to be placed on valves. Old style gate valves will not accommodate electric operators. 2 Generators at all major water treatment plants. Risk Management Allows for continuous operation of both water production wells and high service pumps in emergency situation 3 Seismic retrofit and mitigation of existing plants and well fields. Risk Management Increases the probability that the plants and well fields will survive a major seismic event. 4 Seismically design of all new modifications to existing plants and new plants and well fields. Risk Management Increases the probability that the plants and well fields will survive a major seismic event year seismic plan with financial commitment to above. Risk Management Complete seismic retrofit/mitigation program. Table O-4. Survey result from City of San Diego Water Department Aspect Seismic Improvement Topic Reason for Importance Identify water system components vulnerable to earthquake damage. 1 Perform inspection and assessment of existing system components and identify weaknesses 2 Set operational service goals to establish required performance Risk Assessment and Analysis Seismic Performances If there are no service goals there is nothing to measure performance against 3 Perform a cost-benefit analysis to determine the optimal level of seismic improvements Risk management To determine what improvements need to be done to meet the established performance goals 4 Prioritize facilities to be strengthened Risk Management Complete the most critical facilities first, maximize the benefit to cost 5 Provide redundancy in water supply pipelines in critical supply areas Seismic Preparedness and Readiness Allows multiple sources for water transmission to damaged area of distribution network. A-35

390 3rd US-Japan Workshop on Water System Seismic Practices Table O-5. Survey result from San Francisco Public Utilities Commission Aspect Seismic Improvement Topic Reason for Importance 1 Provide redundancy for the Irvington Tunnel Risk Assessment and Tunnels 85% of the water from H. Hetchy flows through this tunnel. Last tunnel inspection was done 30 years ago. The tunnel crosses a fault line. 2 Upgrade structural and general rehabilitation work on Water Storage Reservoirs Seismic Resistant Design City reservoirs were inspected for the seismic event. Recommendation was to strengthen the roofs, columns and beams. 3 Inspect and assess the existing facilities and identify weaknesses Seismic Performances All facilities and pipelines were inspected for their reliability during operation and seismic event. Inspection revealed that several facilities are vulnerable to age and are not safe during a seismic event. 4 Prioritize facilities to be strengthened Risk Management Complete the CIP program priorities established on the seismic and age criteria. 5 Pipeline repair and readiness Risk Management Develop an emergency response plan for repair and operation of the Transmission Pipelines following a probable seismic event. Table O-6. Survey result from 20 Suburban Water Utilities, San Jose Peninsula South Bay Area (G&E Engineering Systems Inc.) Aspect Seismic Improvement Topic Reason for Importance 1 Clear and accurate assessment of wholesale water system vulnerabilities Risk Assessment Retail customers cannot developed their own cost effective seismic upgraded program until a comprehensive program on the part of the wholesaler is available 2 Forecast damage to distribution system pipes (4" to 24" diameter) due to geohazards (liquefaction, faulting, slides, shaking), in consideration of pipe materials and local soils / corrosion aspects 3 Improve backbone pipeline network to all areas with high value or high risk of fire after earthquake, so that water outages are limited to those areas Risk Assessment About 90% of water outages are attributed to damage to pipelines, especially small diameter pipes. A reasonable forecast of the quantity of pipeline damage, coupled with availability of pipe crews, will yield the forecasted water outage times. This provides information for fire following earthquake and economic analysis. Seismic Resistant Design Rapid restoration of water supply to high economic value customers, or for fighting fires, are the best ways to minimize water impacts to earthquakes. Most of the capital budget should be allocated for this purpose 4 GIS-based analysis of pipe networks due to damage / non-damage from past earthquakes Risk Assessment Study of causes of pipe damage (joinery, material, corrosion, diameter, etc.) is still in need of improvement. We will not be able to make the best "fixes" until we have a better understanding of what is "wrong" 5 Capital Improvement Program Seismic Resistant Design Educate all decision makers inside the water utility, and perhaps also the general public, as to the costs and cost-effectiveness of seismic mitigation measures A-36

391 3rd US-Japan Workshop on Water System Seismic Practices Table O-7. Survey result from Seattle Public Utilities Aspect Seismic Improvement Topic Reason for Importance Pipeline damage can rapidly reduce water availability throughout the system 1 Mitigate pipeline damage effects on system functionality Seismic Performance 2 Make sure newly constructed facilities are seismic-resistant Seismic Resistant Design Much less expensive to make new facilities seismic resistant than to upgrade existing facilities 3 Provide economic justification for seismic improvement Risk Management Utilities are becoming more focused purely on economic considerations 4 Emergency preparedness and response plans so engineers can quickly inspect facilities and crews can repair damaged facilities Seismic Preparedness Allows quicker system restoration and reduces customer impacts 5 Evaluate and upgrade (if necessary) critical facilities such as dams, chemical storage for water treatment, etc that could result in loss of life Risk Assessment and Analysis Saves lives Table O-8. Survey result from Hachinohe Regional Water Supply Authority Aspect Seismic Improvement Topic Reason for Importance 1 Estimate earthquake damage to pipelines and give priority to them in improving Risk Management Cost-benefit performance. Started after Sanriku-Harukaoki Earthquake (M7.5) in Perform drill following manual on countermeasure to disasters (earthquake) once a year Seismic Preparedness and Readiness Cope with disasters (earthquake) rapidly. 3 Install emergency water cut-off valves in distribution pipes to use pipes as reservoir in emergency Post-Earthquake Recoveries and Events Secure water supply to nearby residents and water wagons for its efficient operation in emergency. Started after Tokachi-oki Earthquake (M 7.9) in Evaluate and strengthen Hakusan purification plant, our main purification facility Seismic Resistant Design Strengthen water supply system. 5 Connect pipelines to use multiple water resources Seismic Resistant Design Use other water resources when groundwater is turbid and so on. Ensure post-earthquake water supply. A-37

392 3rd US-Japan Workshop on Water System Seismic Practices Table O-9. Survey result from Kobe City Waterworks Bureau * Aspect Seismic Improvement Topic Reason for Importance Kobe has not enough water resources and depends on the supply from HWSA. 1 Provide redundancy in water transmission main (New transmission line) Seismic preparedness and readiness 2 Reduction of pipe failure risks (Replacement) Seismic preparedness and readiness Distribution network spreads so widely and exposed under the high risk of strong quake. This is for the lessons learned from the Kobe Earthquake 3 Provide enough emergency water in downtown area (Emergency water supply system) Seismic preparedness and readiness In order to avoid confusion of water bureau and customers due to lack of water. 4 Extensive mutual aid scheme Post-Earthquake Recoveries and events Dispersed stock of personnel and equipments (Risk Transfer) 5 System Reconstruction Other (Adaptation to Paradigm shift and Social changes) Increase of aged facilities, Functional improvement Population decrease, Higher claims from customers (water quality etc.) We should utilize the occasion of SIP for other purposes as possible. *Generally, one seismic Improvement includes many aspects or topics. For example New transmission main includes Risk Assessment and Analysis, Seismic Resistant Design and Seismic preparedness and readiness. Table O-10. Survey result from Niigata Waterworks Bureau Aspect Seismic Improvement Topic Reason for Importance 1 Proceed with a project for the block distribution system Seismic Preparedness and Readiness Limit damaged area of distribution network and start quick restoration of the damaged system. 2 Replacement of cast iron pipe with seismic joint ductile iron pipe Seismic Resistant Design Ensure post-earthquake water supply. 3 Perform inspection and assessment of existing system components and identify weaknesses Risk Assessment and analysis Identify water system components vulnerable to earthquake damage. 4 Prepare materials for emergency delivery of fresh water Post-earthquake recoveries and events Ensure post-earthquake emergency water supply. 5 Install a stable power supply system for each plant, normal and backup power and a generator Seismic Resistant Design Ensure post-earthquake water supply. A-38

393 3rd US-Japan Workshop on Water System Seismic Practices Table O-11. Survey result from Osaka Municipal Waterworks Bureau Aspect Seismic Improvement Topic Reason for Importance 1 Provide redundancy in water supply pipelines in critical supply area Risk Assessment and Analysis Allows multiple sources for water transmission to damaged area of distribution network. 2 Evaluate and strengthen water supply pipelines network Seismic Resistant Design Carries out the minimum of the leakage influence. 3 Convincing managers of the importance for seismic improvements Risk Management Require investment of Benefit / Cost for seismic improvements. 4 Prioritize facilities to be strengthened Risk Management Complete the most critical facilities first Table O-12. Survey result from Bureau of Waterworks, Tokyo Metropolitan Government Aspect Seismic Improvement Topic Reason for Importance 1 Estimation of damages of facilities Risk Assessment and The basis for planning the reinforcement work of the facilities Analysis 2 Prioritization of facilities to be reinforced Risk Management To carry out reinforcement works efficiency under financial trouble 3 Replacement of cast iron pipe with special seismic joint ductile iron pipe Seismic Resistant Design To secure post-earthquake water supply. Even now, cast iron pipes are left over not so much but at important places. 4 Establishment of network of transmission pipes and block distribution system Risk Assessment and Analysis To minimize the damaged area of water supply 5 Preparation of emergency water supply facilities Post-earthquake recoveries and events To secure post-earthquake water supply and human lives. A-39

394 3rd US-Japan Workshop on Water System Seismic Practices Table O-13. Survey result from Yokohama Waterworks Bureau Aspect Seismic Improvement Topic Reason for Importance 1 Strengthening the aseismatic properties of waterworks facilities 2 Making water transmission and distribution facilities earthquake resistant 3 Making water transmission and distribution facilities earthquake resistant 4 Establishment of earthquake disaster countermeasures sites Risk Assessment and Analysis Risk Assessment and Analysis Risk Assessment and Analysis Seismic Preparedness and Readiness Earthquake damage is pressed down to a/the minimum.. Aseismatic countermeasure is hypothesizing a/the south Kanto earthquake (M7.9; scale of 6 7 seismic intensity). Reinforcing existing reservoirs and aqueduct spans to improve aseismatic properties Reinforcing pipelines to improve aseismatic properties (areas where liquefaction and/or violent shock is expected) Establishment of sites where materials and equipment are stockpiled for emergency restoration work Development of facilities to accommodate support staff 5 Establishment of risk management structure Risk Assessment Emergency activities manual 6 Strengthening disaster prevention training Risk Assessment Training in reciprocal support with other cities (among cities providing reciprocal support, etc.) A-40

395 3rd US-Japan Workshop on Water System Seismic Practices Survey Results for Different Seismic Improvement Program Topics Table T-1 Seismic Resistant Design Aspect Seismic Improvement Organization Reason for Importance 1 Evaluate and strengthen water storage dams 2 Upgrade structural and general rehabilitation work on Water Storage Reservoirs 3 Capital Improvement Program 4 5 Improve backbone pipeline network to all areas with high value or high risk of fire after earthquake, so that water outages are limited to those areas Make sure newly constructed facilities are seismic-resistant Los Angeles Department of Water and Power San Francisco Public Utilities Commission 20 Suburban Water Utilities, San Jose 20 Suburban Water Utilities, San Jose Seattle Public Utilities Public safety and ensure post-earthquake water supply. Severe damage in 1971 prior to performing. No significant damage in 1994 after strengthening dams in LA. City reservoirs were inspected for the seismic event. Recommendation was to strengthen the roofs, columns and beams. Educate all decision makers inside the water utility, and perhaps also the general public, as to the costs and cost-effectiveness of seismic mitigation measures Rapid restoration of water supply to high economic value customers, or for fighting fires, are the best ways to minimize water impacts to earthquakes. Most of the capital budget should be allocated for this purpose Much less expensive to make new facilities seismic resistant than to upgrade existing facilities 6 Evaluate and strengthen Hakusan purification plant, our main purification facility Hachinohe Regional Water Supply Authority Strengthen water supply system Connect pipelines to use multiple water resources Replacement of cast iron pipe with seismic joint ductile iron pipe Install a stable power supply system for each plant, normal and backup power and a generator Evaluate and strengthen water supply pipelines network Replacement of cast iron pipe with special seismic joint ductile iron pipe Hachinohe Regional Water Supply Authority Use other water resources when groundwater is turbid and so on. Ensure post-earthquake water supply. Niigata Waterworks Bureau Ensure post-earthquake water supply. Niigata Waterworks Bureau Ensure post-earthquake water supply. Osaka Municipal Waterworks Bureau Bureau of Waterworks, Tokyo Metropolitan Government Carries out the minimum of the leakage influence. To secure post-earthquake water supply. Even now, cast iron pipes are left over not so much but at important places. A-41

396 3rd US-Japan Workshop on Water System Seismic Practices Table T-2 Seismic Risk Management Aspect Seismic Improvement Organization Reason for Importance 1 Risk Assessment & Management East Bay Municipal Utility District 2 Prioritize facilities to be strengthened Los Angeles Department of Water and Power 3 Generators at all major water treatment plants. Memphis Light Gas and Water Seismic retrofit and mitigation of existing plants and well fields. Seismically design of all new modifications to existing plants and new plants and well fields. 10 year seismic plan with financial commitment to above. Perform a cost-benefit analysis to determine the optimal level of seismic improvements Memphis Light Gas and Water Memphis Light Gas and Water These two aspects are really part of the same process. They need to be iterative as part of the decision-making process in planning an improvement program. Priority-setting based on post-earthquake performance goals enables an agency to define the system improvements in an efficient manner. Complete the most critical facilities first Allows for continuous operation of both water production wells and high service pumps in emergency situation Increases the probability that the plants and well fields will survive a major seismic event. Increases the probability that the plants and well fields will survive a major seismic event. Memphis Light Gas and Water Complete seismic retrofit/mitigation program. City of San Diego Water Department To determine what improvements need to be done to meet the established performance goals 8 Prioritize facilities to be strengthened City of San Diego Water Department Complete the most critical facilities first, maximize the benefit to cost 9 Prioritize facilities to be strengthened 10 Pipeline repair and readiness 11 Provide economic justification for seismic improvement San Francisco Public Utilities Commission San Francisco Public Utilities Commission Seattle Public Utilities Complete the CIP program priorities established on the seismic and age criteria. Develop an emergency response plan for repair and operation of the Transmission Pipelines following a probable seismic event. Utilities are becoming more focused purely on economic considerations 12 Estimate earthquake damage to pipelines and give priority to them in improving Hachinohe Regional Water Supply Authority Cost-benefit performance. Started after Sanriku-Harukaoki Earthquake (M7.5) in Convincing managers of the importance for seismic improvements Osaka Municipal Waterworks Bureau Require investment of Benefit / Cost for seismic improvements. 14 Prioritize facilities to be strengthened Osaka Municipal Waterworks Bureau Complete the most critical facilities first 15 Prioritization of facilities to be reinforced Bureau of Waterworks, Tokyo Metropolitan Government To carry out reinforcement works efficiency under financial trouble A-42

397 3rd US-Japan Workshop on Water System Seismic Practices Table T-3 Seismic Preparedness and Readiness Aspect Seismic Improvement Organization Reason for Importance Planning for coordination and outreach to cities, counties, fire agencies, emergency response agencies. As EBMUD's Seismic Improvement Program (SIP) nears completion, we will be working with these entities more closely. Provide new butterfly header valves with remote operators that can be operated from a SCADA system. The butterfly valves will either replace or be added existing header. Provide redundancy in water supply pipelines in critical supply areas East Bay Municipal Utility District Memphis Light Gas and Water Water system improvements need to a part of overall community emergency planning. Allow for electric operators to be placed on valves. Old style gate valves will not accommodate electric operators. City of San Diego Water Department Allows multiple sources for water transmission to damaged area of distribution network. 4 Emergency preparedness and response plans so engineers can quickly inspect facilities and crews can repair damaged facilities Seattle Public Utilities Allows quicker system restoration and reduces customer impacts 5 Perform drill following manual on countermeasure to disasters (earthquake) once a year Hachinohe Regional Water Supply Authority Cope with disasters (earthquake) rapidly. 6 Provide redundancy in water transmission main (New transmission line) Kobe City Waterworks Bureau Kobe has not enough water resources and depends on the supply from HWSA. 7 Reduction of pipe failure risks (Replacement) Kobe City Waterworks Bureau Distribution network spreads so widely and exposed under the high risk of strong quake. This is for the lessons learned from the Kobe Earthquake Provide enough emergency water in downtown area (Emergency water supply system) Proceed with a project for the block distribution system Establishment of earthquake disaster countermeasures sites Kobe City Waterworks Bureau Niigata Waterworks Bureau Yokohama Waterworks Bureau In order to avoid confusion of water bureau and customers due to lack of water. Limit damaged area of distribution network and start quick restoration of the damaged system. Establishment of sites where materials and equipment are stockpiled for emergency restoration work Development of facilities to accommodate support staff A-43

398 3rd US-Japan Workshop on Water System Seismic Practices Table T-4 Risk Assessment and Analysis Aspect Seismic Improvement Organization Reason for Importance Provide redundancy in water supply pipelines in critical supply areas Provide as much in-city water storage and sources of supply as possible Perform inspection and assessment of existing system components and identify weaknesses Perform inspection and assessment of existing system components and identify weaknesses 5 Provide redundancy for the Irvington Tunnel Clear and accurate assessment of wholesale water system vulnerabilities Forecast damage to distribution system pipes (4" to 24" diameter) due to geohazards (liquefaction, faulting, slides, shaking), in consideration of pipe materials and local soils / corrosion aspects GIS-based analysis of pipe networks due to damage / non-damage from past earthquakes Evaluate and upgrade (if necessary) critical facilities such as dams, chemical storage for water treatment, etc that could result in loss of life Perform inspection and assessment of existing system components and identify weaknesses Provide redundancy in water supply pipelines in critical supply area Los Angeles Department of Water and Power Los Angeles Department of Water and Power Los Angeles Department of Water and Power City of San Diego Water Department San Francisco Public Utilities Commission 20 Suburban Water Utilities, San Jose 20 Suburban Water Utilities, San Jose 20 Suburban Water Utilities, San Jose Seattle Public Utilities Saves lives Niigata Waterworks Bureau Osaka Municipal Waterworks Bureau Allows multiple sources for water transmission to damaged area of distribution network. Proved useful following 1971 and 1994 Mw 6.7 earthquakes Allows multiple sources for water supply following earthquake damage to distribution network. Proved useful following 1971 and 1994 Mw 6.7 earthquakes. Identify water system components vulnerable to earthquake damage. Identify water system components vulnerable to earthquake damage. 85% of the water from H. Hetchy flows through this tunnel. Last tunnel inspection was done 30 years ago. The tunnel crosses a fault line. Retail customers cannot developed their own cost effective seismic upgraded program until a comprehensive program on the part of the wholesaler is available About 90% of water outages are attributed to damage to pipelines, especially small diameter pipes. A reasonable forecast of the quantity of pipeline damage, coupled with availability of pipe crews, will yield the forecasted water outage times. This provides information for fire following earthquake and economic analysis. Study of causes of pipe damage (joinery, material, corrosion, diameter, etc.) is still in need of improvement. We will not be able to make the best "fixes" until we have a better understanding of what is "wrong" Identify water system components vulnerable to earthquake damage. Allows multiple sources for water transmission to damaged area of distribution network. A-44

399 3rd US-Japan Workshop on Water System Seismic Practices 12 Estimation of damages of facilities 13 Establishment of network of transmission pipes and block distribution system Strengthening the aseismatic properties of waterworks facilities Bureau of Waterworks, Tokyo Metropolitan Government Bureau of Waterworks, Tokyo Metropolitan Government The basis for planning the reinforcement work of the facilities To minimize the damaged area of water supply Earthquake damage is pressed down to a/the minimum.. Aseismatic countermeasure is hypothesizing a/the south 14 Yokohama Waterworks Bureau Kanto earthquake (M7.9; scale of 6 7 seismic intensity). Making water transmission and distribution Reinforcing existing reservoirs and aqueduct spans to 15 Yokohama Waterworks Bureau facilities earthquake resistant improve aseismatic properties Making water transmission and distribution Reinforcing pipelines to improve aseismatic properties 16 Yokohama Waterworks Bureau facilities earthquake resistant (areas where liquefaction and/or violent shock is expected) 17 Establishment of risk management structure Yokohama Waterworks Bureau Emergency activities manual 18 Strengthening disaster prevention training Yokohama Waterworks Bureau Training in reciprocal support with other cities (among cities providing reciprocal support, etc.) A-45

400 3rd US-Japan Workshop on Water System Seismic Practices Table T-5 Seismic Performances and Post-Earthquake Recoveries Aspect Seismic Improvement Organization Reason for Importance Revisit original SIP modeling of system performance. Set operational service goals to establish required performance Inspect and assess the existing facilities and identify weaknesses Mitigate pipeline damage effects on system functionality East Bay Municipal Utility District City of San Diego Water Department San Francisco Public Utilities Commission Seattle Public Utilities System performance needs to be a priority, building on performance of individual components. It should also be revisited regularly to incorporate lessons learned in studies of post-earthquake responses in other areas. If there are no service goals there is nothing to measure performance against All facilities and pipelines were inspected for their reliability during operation and seismic event. Inspection revealed that several facilities are vulnerable to age and are not safe during a seismic event. Pipeline damage can rapidly reduce water availability throughout the system 5 Install emergency water cut-off valves in distribution pipes to use pipes as reservoir in emergency Hachinohe Regional Water Supply Authority Secure water supply to nearby residents and water wagons for its efficient operation in emergency. Started after Tokachi-oki Earthquake (M 7.9) in Extensive mutual aid scheme Kobe City Waterworks Bureau 7 8 Prepare materials for emergency delivery of fresh water Preparation of emergency water supply facilities Dispersed stock of personnel and equipments (Risk Transfer) Niigata Waterworks Bureau Ensure post-earthquake emergency water supply. Bureau of Waterworks, Tokyo Metropolitan Government To secure post-earthquake water supply and human lives. A-46

401 3rd US-Japan Workshop on Water System Seismic Practices Table T-6 Other topics Aspect Seismic Improvement Topic Organization Reason for Importance 1 Claremont Corridor Seismic Improvements - upgrades to a critical water distribution tunnel. 2 System Reconstruction Communication with customers and public Adaptation to Paradigm shift and Social changes East Bay Municipal Utility District Kobe City Waterworks Bureau Involvement of customers all through the process is important to gaining initial support and to maintaining the overall goal in the face of site-specific opposition to individual seismic improvements Increase of aged facilities, Functional improvement Population decrease, Higher claims from customers (water quality etc.) We should utilize the occasion of SIP for other purposes as possible. A-47

402 3rd US-Japan Workshop on Water System Seismic Practices A-48

403 3rd US-Japan Workshop on Water System Seismic Practices Survey Request A-49

404 3rd US-Japan Workshop on Water System Seismic Practices Survey Request on Important Aspects to Include in a Water System Seismic Improvement Program To enhance the workshop panel discussion I am requesting all persons and organizations involved in the workshop to spend a few minutes to describe up to 5 aspects that they consider important to include in a water system seismic improvement program and a very brief description as to why it is considered important. Please also identify the aspect to be within one of the following main topics: 8. Risk Assessment and Analysis 9. Risk Management 10. Seismic Preparedness and Readiness 11. Seismic Resistant Design 12. Seismic Performances 13. Post-Earthquake Recoveries and Events 14. Other, specify a topic A list of these same topics and the types of aspects that are included within them is provided in an additional file sent with this document. For your reference, an example list of 5 important aspects for the Los Angeles Department of Water and Power is provided at the end to this document. In addition, some additional examples are provided based on an assessment of reports from the 2 nd workshop in Tokyo, These aspects do not have to be in any priority order, nor a comprehensive list of the aspects you consider important to include in a seismic improvement program. The information provided will be presented in the workshop proceedings and used as reference during the panel discussion on August 8, It is not necessary to identify a seismic improvement aspect for each of the above topics. It is more important to first identify the seismic improvement aspects that are important, then place it under one of the topics listed above. It is acceptable for all aspects to be listed under one topic, or only a few topics, as in the case for the Los Angeles Department of Water and Power. In addition, it is not necessary for the aspects you list to be described in the paper to be presented at the workshop. We would prefer that you only spend a few minutes to fill out the attached table and try to identify seismic improvement aspects from your personal experiences and those that your organization has considered, had to deal with, implemented, etc. If responses are focused on aspects commonly considered to your respective organizations, we may be able to identify interesting trends that will aid in evolving the current state of knowledge in Water System Seismic Practices. This exercise is not intended to place a burden on your time, but will prove very valuable in identifying and itemizing aspects that different organizations consider important in their seismic improvement programs. Therefore, we request that you please just take a few minutes to fill out as many (up to 5) important seismic improvement aspects; if you can not complete 5 please complete as many as possible. A-50

405 3rd US-Japan Workshop on Water System Seismic Practices Please return by August 4, 2003 to: Craig A. Davis by Organization: Los Angeles Department of Water and Power Aspect Seismic Improvement Topic Reason for Importance 1 Provide redundancy in water supply pipelines in critical supply areas Risk Assessment and Analysis Allows multiple sources for water transmission to damaged area of distribution network. Proved useful following 1971 and 1994 Mw 6.7 earthquakes 2 Provide as much in-city water storage and sources of supply as possible 3 Evaluate and strengthen water storage dams 4 Perform inspection and assessment of existing system components and identify weaknesses Risk Assessment and Analysis Allows multiple sources for water supply following earthquake damage to distribution network. Proved useful following 1971 and 1994 Mw 6.7 earthquakes. Seismic Resistant Design Public safety and ensure post-earthquake water supply. Severe damage in 1971 prior to performing. No significant damage in 1994 after strengthening dams in LA. Risk Assessment and Analysis Identify water system components vulnerable to earthquake damage. 5 Prioritize facilities to be strengthened Risk Management Complete the most critical facilities first Additional examples: 1. Topic: Seismic Preparedness and Readiness Seismic Improvement Aspect: Purchase and store materials to aid in earthquake recovery 2. Topic: Seismic Performances a. Research and evaluation of large diameter steel pipe performances during past earthquakes 3. Topic: Seismic Resistant Design a. Replacement of cast iron pipe with special seismic joint ductile iron pipe 4. Topic: Risk Management a. Identifying funding sources for seismic improvements b. Convincing managers of the importance for seismic improvements 5. Risk Assessment a. Perform seismic evaluation of embankment dams. b. Develop mitigation strategies for ground failure impacts on distribution pipelines A-51

406 3rd US-Japan Workshop on Water System Seismic Practices Please complete and return this table. Organization: Aspect Seismic Improvement Topic Reason for Importance A-52

407 3 rd US-Japan Workshop on Water System Seismic Practices Appendix V: Los Angeles Department of Water and Power Facilities Tour A-53

408 3 rd US-Japan Workshop on Water System Seismic Practices A-54

409 3rd US-Japan Workshop on Water System Seismic Practices Appendix V: Los Angeles Department of Water and Power Facilities Tour Tour Itinerary 9:30 am Depart from John Ferraro Building (JFB) - LADWP 10:00 am Arrive Magnolia Trunkline 10:30 am Depart from Magnolia Trunkline for Van Norman Complex (VNC) 11:00 am Arrive VNC 1:00 pm Lunch VNC Candy Land 1:30 pm Continue Tour of VNC 3:00 pm Depart VNC for Hollywood Water Quality Improvement Project (HWQIP) 3:30 pm Arrive HWQIP 4:30 pm Depart form HWQIP for Hotel/JFB 5:00 pm End of Tour A-55

410 3rd US-Japan Workshop on Water System Seismic Practices Figure A-1. Los Angeles Department of Water & Power facilities tour map. A-56

411 3rd US-Japan Workshop on Water System Seismic Practices Magnolia Trunkline Project Purpose of Project The Magnolia Trunk Line Project is part of a systemwide improvement program that LADWP is undertaking in and around the San Fernando Valley. The impact of the Northridge Earthquake in 1994 highlighted the desirability of having greater flexibility within the water distribution system. Flexibility would allow for the repair and/or maintenance of water lines with minimum disruption of water service to the community. According to the planning efforts of LADWP, the Magnolia Trunk Line is needed to convey water between two major reservoirs, the Los Angeles Reservoir and Stone Canyon Reservoir. This project will allow for increased reliability and greater flexibility to meet the needs of the customers in the San Fernando Valley and other portions of Los Angeles. Project Description The Project consists of 14,350 feet of 54-inch-diameter steel pipe and five buried vaults. The pipeline will be installed by conventional trenching methods. Pipe jacking, which is similar to tunneling, will be used to install the pipe across the following streets: Kester Avenue, Cedros Avenue, Van Nuys Boulevard, Hazeltine Avenue, and Fulton Avenue. Location The project is located along Magnolia Boulevard from Noble Avenue to Coldwater Canyon Avenue, in Sherman Oaks, California. A portion of the trunk line will also be built in Willis Avenue between Magnolia Boulevard and Otsego Street. Construction Schedule Construction Start January 2003 Estimated Finish June 2004 Estimated Project Cost $17,915,000 A-57

412 3rd US-Japan Workshop on Water System Seismic Practices Figure A-2. Magnolia Trunk Line area map. Figure A-3. Typical trench section of Magnolia Trunk Line. A-58

413 3rd US-Japan Workshop on Water System Seismic Practices Van Norman Complex Van Norman Complex Abstract The Van Norman Complex (Complex) is owned and operated by the City of Los Angeles Department of Water and Power. The Complex serves as a terminal for the First and Second Los Angeles Aqueducts and controls approximately 75% of the annual water supply for the City of Los Angeles. The Complex receives, stores, and distributes water from the Los Angeles Aqueducts and the California Aqueduct. Important water facilities on the Complex include the High Speed Channel, Bypass Channel, Tailrace, Los Angeles Aqueduct Filtration Plant, Los Angeles Dam and Reservoir, Upper and Lower San Fernando Dams, Lower Van Norman Bypass Dam, and Reservoir, Van Norman Pumping Stations Nos. 1 and 2, several debris basins and retaining embankments, and several miles of large diameter pipelines. The Complex also contains important Power System facilities such as the Rinaldi Receiving Station, Sylmar Converter Station, and two power plants. The Complex has been shaken with strong near-field ground motions from the 1971 San Fernando and 1994 Northridge Earthquakes; each earthquake having a moment magnitude of M w 6.7 and an epicientral distance of approximately 11 km from the Complex. In addition, the Complex overlies the ruptured faults that generating shaking for these two events and was subkected to tectonic ground deformations. In 1994, the largest ground motion velocities ever recorded from any previous earthquake were measured at the Rinaldi Receiving Station. Evaluation of geotechnical performances of natural and engineered soil structures and studies to understand the damage inflicted to water and power facilities on the Complex during the 1971 and 1994 earthquakes have led to significant improvements in geotechnical and lifeline earthquake engineering throughout the world. The tour will have stops at several water system facilities throughout the Complex. Facility descriptions will include: the function and importance in the Los Angeles Water System, past earthquake performances, geotechnical aspects affecting seismic performances, and improvements/seismic mitigations implemented. A-59

414 3rd US-Japan Workshop on Water System Seismic Practices Figure A-4. Van Norman Reservoir Complex tour map. A-60

415 3rd US-Japan Workshop on Water System Seismic Practices STOP 1: Tailrace, High Speed, and Bypass Channels The Tailrace, High Speed and Bypass Channels make up the three aqueduct channels on the Complex. The Tailrace Channel receives the discharge flow from the San Fernando Power Plant. The power plant takes water flowing under gravity from the First Los Angeles Aqueduct through a penstock, processes the water in turbine chambers, and discharges it through a tunnel and a riser into the Tailrace. The Tailrace is a large channel and effectively acts as a small reservoir, having little energy loss across its length. The Tailrace suffered significant damage from liquefaction-induced lateral spreading during the 1971 and 1994 earthquakes. The small compacted fill dike at the south end failed following the 1994 earthquake from piping through transverse cracks that developed as a direct result of earthquake shaking. The High Speed Channel (HSC) and Bypass Channel (Bypass) are two of the most important and most used facilities in the Water System. The HSC and Bypass deliver water from the First Los Angeles Aqueduct (FLAA), the Second Los Angeles Aqueduct (SLAA) and raw water connections from the Metropolitan Water District to the Los Angeles Filtration Plant for filtering and overall distribution to the City of Los Angeles. The HSC and Bypass Channels are critical channel segments along the FLAA and SLAA allowing various sources of supply to enter the city. STOP 2: Upper San Fernando Dam, Van Norman Pumping Station No inch Discharge Line, and Van Norman Pumping Station No. 2 The Upper San Fernando Dam (USFD) is a hydraulic fill dam, which was constructed between 1919 and The USFD originally impounded the Upper Van Norman Reservoir. The 1971 San Fernando Earthquake damaged the USFD causing up to 3 feet (0.9 m) of settlement, 5 ft (1.5 m) of downstream movement, and up to 2 feet (0.6 m) of longitudinal movement. Following the 1971 earthquake, emergency repairs were performed to keep the USFD in service until the completion of the Los Angeles Reservoir project. The USFD was removed from service in 1979, and since then has become a retention/diversion structure for storm water control in the Van Norman Complex. The Los Angeles Aqueduct Backwash Ponds are located just upstream of the USFD. The USFD sustained some longitudinal cracking, lateral movement and settlement as a result of the 1994 Northridge Earthquake. The 54-inch Discharge Line serves as the discharge for the Van Norman Pumping Station (VNPS) No. 1, which is used for emergencies in the event the VNPS No. 2 is out of service. The VNPS No. 1 is needed to provide service to the foothill Trunk Line System to the east of the Van Norman Complex. The 1971 San Fernando and 1994 Northridge Earthquakes damaged the 54-inch discharge line concrete piers and pile support systems. Damage was limited to the support system and the pipeline was able to remain in service following both earthquakes. Following the 1994 earthquake the 54-inch discharge line was placed on a base isolation system to minimize lateral movement in the pipeline. A-61

416 3rd US-Japan Workshop on Water System Seismic Practices The Van Norman Pumping Station No. 2 supplies filtered water to the higher water service elevations in the Northern San Fernando Valley. The pump station building contains eight 1500 horsepower vertical pumps, a 22 ft (6.7 m) by 47 ft (14.3 m) control room, electrical switchgear, and miscellaneous electrical and mechanical equipment. A clearwell with an overall capacity of 4.6 million gallons (17.4 million liters) is located adjacent to the pumping station to provide the required chlorine contact time and operating storage for the filtered water. The pumping station and Clearwell performed very well in the 1994 Northridge Earthquake. STOP 3: Lunch at Candy Land Picnic Area STOP 4: Sepulveda Trunk Line and Lower San Fernando Dam The Sepulveda Trunk Line is an 84 to 96 inch (213 to 244 cm) diameter pipeline that was put in service in July The 5-mile long (8 km) trunk line was built to replace the City Trunk Line; a large diameter riveted steel pipe that was built in The old 1913 pipeline had shown evidence of leaks and corrosion and was replaced to avoid a pipe failure. The Sepulveda Trunk Line is one of the most critical and important supply lines to the City of Los Angeles. The Lower San Fernando Dam (LSFD) is a hydraulic fill dam, which was constructed between 1912 and The LSFD originally impounded the Lower Van Norman Reservoir. During the 1971 San Fernando Earthquake the LFSD developed a major liquefaction induced slide in the upstream slope, loosing approximately 30 feet of the 35 feet of freeboard available prior to the earthquake. The slide nearly caused a major uncontrolled release of the reservoir. No water escaped downstream as a result of the earthquake, but the reservoir was drained as fast as possible and was taken out of service on February 25, The earthquake also caused collapse of Outlet Tower No. 1 and damage to Tower No. 2 in The LSFD was reconstructed following the 1971 earthquake to retain storm water. Currently the area upstream of the dam is a storm water detention basin. The outlet towers were replaced with ungated outlet lines. The LSFD performed relatively well during the 1994 earthquake with the upstream hydraulic fill slide debris liquefying and causing some upstream movement and settlement, resulting in longitudinal cracking in the crest and upstream berm. The 96-inch corrugated metal pipe installed as part of the Outlet Tower No. replacement suffered complete lateral collapse during the 1994 earthquake. The collapsed drainline was removed and replaced with a rigid pipe. STOP 5: Los Angeles Reservoir The Los Angeles Reservoir was constructed in 1977 to store treated water of the Los Angeles Aqueduct, and to replace the Lower and Upper San Fernando Dams, which were extensively damaged by the 1971 San Fernando Earthquake. The Los Angeles Reservoir has two compacted earthfill dams, the Los Angeles Dam (154 ft or 47 m high) and the North Dike (118 ft or 36 m). Both embankments are founded on bedrock, are zoned with A-62

417 3rd US-Japan Workshop on Water System Seismic Practices shell material on the upstream and downstream slopes, and contain a chimney drain made of coarse material. The Los Angeles Dam also has a clay zone upstream of the chimney drain. The interior slopes and perimeter roadways of the reservoir are lined with asphalt concrete pavement, but the reservoir bottom is not paved. The Los Angeles Reservoir performed pretty well during the 1994 Northridge Earthquake, with minor surficial cracking and small settlements of the embankment dams. The outlet tower had not problems with over 1g of motion in several consecutive cycles. The bridge connection to piers did not hold and one section displaced horizontally, but the bridge did collapse. Van Norman Complex References Bardet, J. P. and C. A. Davis, Lower San Fernando Corrugated Metal Pipe Failure, M. O Rourke ed., Proc. Fourth U.S. Conf. on Lifeline Earthquake Engineering, ASCE, San Francisco, Aug., pp Bardet, J. P. and C. A. Davis, 1996a, Engineering Observations on Ground Motion at the Van Norman Complex after the Northridge Earthquake, Bulletin of Seismological Society of America, Special Northridge Issue, Vol. 86, No. 1B, pp. S333-S349. Bardet, J.P., and C.A. Davis, 1996b, Performance of San Fernando Dams during the 1994 Northridge Earthquake, Journal of the Geotechnical Engineering Division, ASCE, Vol. 122, No. 7, pp Bardet, J.P., and C.A. Davis, 1996c, Study of Near-Source Ground Motion at the Van Norman Complex after the 1994 Northridge Earthquake, Proc. Workshop on Site Response Subjected to Strong Ground Motions, Port and Harbour Research Institute, Yokosuka, Japan, January 16-17, Vol. 2, pp Bardet, J. P. and C. A. Davis, 1998a, An Overview of the Investigations of the Performance of Geotechnical Structures in the Van Norman Complex after the 1994 Northridge Earthquake, Vol. II, Proceedings of the NEHRP Conference and Workshop on Research on the Northridge, California Earthquake of January 17, 1994, California Universities for Research in Earthquake Engineering, Aug , 1997, Los Angeles California, pp Brown, K., P. Rugar, C. Davis, and T. Rulla, 1995, Seismic Performance of Los Angeles Water Tanks, Proc. 4th U.S. Conf. on Lifeline Earthquake Engineering, ASCE, San Francisco, Aug., pp Brown, K.J., and C.A. Davis, 1996, The Lower Van Norman Bypass Reservoir Cover, In Gopu, V. K. A. (ed.), Proceedings of the International Wood Engineering Conference 96, New Orleans, LA, Oct , Vol. 3, pp A-63

418 3rd US-Japan Workshop on Water System Seismic Practices Bureau, G., S. Inel, C.A. Davis, W.H. Roth, 1996, Seismic Response of Los Angeles Dam, CA, during the 1994 Northridge Earthquake Proc. 16th Annual USCOLD Lecture Series, Los Angeles, CA, July 22-26, pp Cultrera, G., D. M. Boore, W. B. Joyner, and C. M. Dietel, 1999, Nonlinear Soil Response in the Vicinity of the Van Norman Complex Following the 1994 Northridge, California, Earthquake, Bull. Seism. Soc. Am., 89(5), pp Davis, C. A., 1999b, Performance of a Large Diameter Trunk Line During Two Near- Field Earthquakes, Proc. 5th U.S. Conf. on Lifeline Earthquake Engineering, ASCE, Seattle, Aug., pp Davis, C. A., 1999c, Case Study of the Granada Trunk Line During Two Near-Field Earthquakes, Proceedings of the Seventh Japan-US Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures Against Soil Liquefaction, Multidisciplinary Center for Earthquake Engineering Research, Technical Report MCEER , T. D. O Rourke, J. P. Bardet, and M. Hamada, eds., pp Davis, C.A., 2001, Retrofit of a Large Diameter Trunk Line Case Study of Seismic Performance, Proc. of 2 nd Japan-US Workshop on Seismic Measures for Water Supply, AWWARF/JWWA, Tokyo, Japan, Aug. 6-9, 2001, American Waterworks Association Research Foundation Project No. 2786, Session 2. Davis, C.A., and M.M. Sakado, 1994, Response of the Van Norman Complex to the Northridge Earthquake, Proc. 11th Annual Conf. of the Association of State Dam Safety Officials, Sept., Boston, Mass, pp Davis, C.A., and J.P. Bardet, 1994, Geotechnical Observations at the Van Norman Complex after the 1994 Northridge Earthquake, Proceedings of the Fifth Japan-US Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures Against Soil Liquefaction, National Center for Earthquake Engineering Research, Technical Report NCEER , T.D. O Rourke and M. Hamada, eds., pp Davis, C.A., and J.P. Bardet, 1995a, Northridge Earthquake -- Van Norman Complex Ground Movement, Proceedings of the Third International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, Vol.3, Paper 1409, St. Louis, Missouri, April. Davis, C. A. and J. P. Bardet, 1995b, Seismic Performance of Van Norman Water Lifelines, Proc. 4th U.S. Conf. on Lifeline Earthquake Engineering, ASCE, San Francisco, Aug., pp Davis, C.A., and J.P. Bardet, 1996a, Performance of Two Reservoirs During 1994 Northridge Earthquake, J. Geotech. Engrg. Div., ASCE, Vol. 122, No. 8, pp A-64

419 3rd US-Japan Workshop on Water System Seismic Practices Davis, C.A., and J.P. Bardet, 1996b, Performance of Four Corrugated Metal Pipes during the 1994 Northridge Earthquake Proceedings of the 6th Japan-US Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures against Soil Liquefaction, Tokyo, June, pp Davis, C.A. and J.P. Bardet, (1998), Seismic Analysis of Large Diameter Flexible Underground Pipes, J. Geotech. And GeoEnv. Engrg. Div., ASCE, Vol. 124, No. 10, pp Davis, C. A. and J. P. Bardet, (2000), Responses of Buried Corrugated Metal Pipes to Earthquakes, J. Geotech. And GeoEnv. Engrg. Div., ASCE, 126(1), pp Davis, C.A., J.P. Bardet, and J. Hu, 2002, Effects of Ground Movements on Concrete Channels Proc. 8th US-Japan Workshop on Earthquake Resistant Des. of Lifeline Fac. and Countermeasures for Soil Liquefaction, Tokyo, Japan, T. D. O Rourke, J. P. Bardet, and M. Hamada, Eds., Tech. Rep. MCEER. EERI, Earthquake Engineering Research Institute, 1995, Northridge Earthquake of 17 January, 1994 Reconnaissance Report, Vol. 1, Earthquake Spectra, Supplement C to Volume 11. Lee, K.L., H.B. Seed, I.M. Idriss, and F.I. Makdisi, 1975, Properties of Soil in the San Fernando Hydraulic Fill Dams," Journal of the Geotechnical Engineering Division, ASCE, GT8, pp Lindvall Richter Benuska Associates, 1994, Processed LADWP Power System Strong-Motion Records from the Northridge, California Earthquake of January 17, 1994, Prepared for the Los Angeles Department of Water and Power,LRB No , May 26, Lund, Le Val, and E. Matsuda, 1994, Lifelines, Northridge Earthquake, January 17, 1994, Preliminary Reconnaissance Report, Chapter 6, Earthquake Engineering Research Institute, pp O Rourke, T. D. and M. Hamada, Eds., 1992, Case Studies of Liquefaction and Lifeline Performance during Past Earthquakes, NCEER , Vol. 2, National Center for Earthquake Engineering Research, Buffalo, NY. Porcella, R.L., E.C. Etheredge, R.P. Maley, and A.V. Acosta, 1994, Accelerograms Recorded at USGS National Strong-Motion Network Stations During the Ms=6.6 Northridge, California Earthquake of January 17, 1994, Department of the Interior, U.S. Geological Survey, February, Open File Report Roth, W.H., S. Inel, C. Davis, and G. Brodt, 1993, Upper San Fernando Dam 1971 Revisited, Proc. 10th Annual Conf. of the Association of State Dam Safety Officials, Aug., Kansas City, MO, pp A-65

420 3rd US-Japan Workshop on Water System Seismic Practices Saul, R.B., 1975, Geology of the Southeast Slope of the Santa Susana Mountains and Geologic Effects of the San Fernando Earthquake, San Fernando, California, Earthquake of 9 February 1971, Chapter 57 California Division of Mines and Geology, Bulletin 196, pp Scott, R.F., 1973, The Calculation of Horizontal Accelerations from Seismoscope Records, Bulletin of the Seismological Society of America, Vol. 63, No. 5, pp Seed, H. B, K. L. Lee, I. M. Idriss, and F. I. Makdisi, 1973, Analysis of the Slides in the San Fernando Dams During the Earthquake of February 9, 1971, Report No. UCB/EERC 73-2 University of California, Berkeley, California. Subcommittee on Water and Sewage Systems, 1973, Earthquake Damage to Water and Sewerage Facilities, San Fernando, California, Earthquake of February 9, 1971, N. A. Benfer and J. L. Coffman eds., U. S. Dept. of Commerce, NOAA Spec. Rpt., Vol. II, pp Trifunac, M. D., M. I. Todorovska, and V. W. Lee, 1998, The Rinaldi Strong Ground Motion Accelerogram of the Northridge, California Earthquake of 17 January 1994, Earthquake Spectra, EERI, Vol. 14, No. 1, pp Weber, F. H., 1975, Surface Effects and Related Geology of the San Fernando Earthquake in the Sylmar Area, Chapter 6 in San Fernando, California, Earthquake of 9 February 1971, Oakeshott, G. B., Ed., California Division of Mines and Geology, Bulletin 196, Sacramento, CA. Youd, T. L., 1971, Landsliding in the Vicinity of the Van Norman Lakes, in The San Fernando, California, Earthquake of February 9, 1971, U. S. Geological Survey Professional Paper 733, pp A-66

421 3rd US-Japan Workshop on Water System Seismic Practices Hollywood Water Quality Improvement Project The Hollywood Water Quality Improvement Project is an excellent example of a successful heavy construction project in a potentially controversial urban setting. The Los Angeles Department of Water and Power was able to design and construct the project with the community s involvement throughout the entire process. The Project was undertaken by the Los Angeles Department of Water and Power (LADWP) to meet new water quality regulations from the California Department of Health Services. The project consists of two 30 million gallon buried water storage tanks, a large underground control vault, and seven tunnels of varying size, totaling over a mile in length. The main bypass tunnel has over one mile of 10.5 foot diameter tunnel with bolted, gasketed, concrete segments driven by an EPB tunnel Boring Machine. The storage tanks are some of the largest of their kind, at 385 ft diameter and 42 feet in height. The tanks are built of reinforced concrete and wrapped with 230 miles of steel strand cable to provide horizontal prestressing of the tank walls. The site location is in a very affluent community located near the Hollywood Sign, and the community involvement and cooperation was exceptional. Excavation on the site consisted of removing one million cubic yards of soil and rock and placing the material in fill sites on the property adjacent to the tank sites. The final grading of the fill sites uses a technique called Landform Grading to give the fills a more natural appearance. The project was awarded as a $74 million lump sum contract with an $11 million contingency to Kiewit Pacific. Kiewit subcontracted the tunnel work to Elmore Pipejacking. The construction management team consisted of Jacobs Associates, EPC, UCM, and Montgomery Watson Harza. Woodward Clyde/URS and Geosoils provided geotechnical design support. The design was done primarily by the Los Angeles Department of Water and Power. The primary deadline of this project was system operation and compliance by November The system was fully operational by July 2001, and was officially in compliance by November 2001, meeting or exceeding the schedule requirements. A-67

422 3rd US-Japan Workshop on Water System Seismic Practices Figure A-5. Hollywood water quality improvement project map. A-68

423 3 rd US-Japan Workshop on Water System Seismic Practices Appendix VI: Workshop Participants A-69

424 3 rd US-Japan Workshop on Water System Seismic Practices A-70