Pipelines 2010: Climbing New Peaks to Infrastructure Reliability Renew, Rehab, and Reinvest 2010 ASCE

Transcription

1 1492 Integrated Decision-Support Framework for Municipal Infrastructure Asset Khaled Shahata 1 and Tarek Zayed 2 ABSTRACT Integration planning of Infrastructure systems reveals a changeling decisions facing Canadian municipalities for planning repair/renewal of road network, water distribution network, wastewater distribution network. Decision-making for these networks requires the incorporation of a massive amount of data collection, building business processes, identifying decision variables and optimization. The objective of this research is to establish a methodology to facilitate decision making process that ensures reliable and optimum decision regarding corridor rehabilitation for road, water and wastewater network. This proposed framework employs the following tasks: (1) analyze risk; (2) conduct performance evaluation; (3) assess the current physical condition of the pipe and road segment; (4) collecting data and performing data gap analysis; (5) document a conceptual business process diagrams; (6) develop decision analysis trees; and (7) implementing optimization of repair/renewal cost and defining the best replacement interval via genetic algorism (GA). In order to demonstrate the model features, a case study has been utilized from the City of Guelph, ON, Canada. The model is developed via genetic algorism (GA) using GIS platform. The results assist in setting priorities for integrated corridor rehabilitation and anticipated to generate a capital planning program for the city s infrastructure. In conclusion, this framework helps Canadian municipalities evaluate and select feasible optimal assets for integrated corridor rehabilitation. Keywords: * Decision-Support Framework, Decision-making, Asset management, GIS, standard municipal data, integrated infrastructure management, roads, water, wastewater network. 1 Asset Management Engineer, AECOM, Mississauga, Ontario, Canada, PH ext. 256; FAX ; Khaled.Shahata@aecom.com. 2 Associate Professor, Department of Building, Civil, and Environ. Eng., Concordia University, Montreal, Quebec, Canada, PH (514) ext. 8779; FAX (514) ; tzayed@alcor.concordia.ca. 1

2 1493 INTRODUCTION Efficient management of municipal infrastructure systems is currently more focused on the overall integrated multi-disciplinary aspects of Infrastructure Management process. The development and deployment of integrated infrastructure management systems is currently more of a necessity to maintain our infrastructure assets. Integrated infrastructure management would facilitate information flow across various disciplines and activities, which in return would improve the availability, reliability, and consistency of infrastructure information, resulting in timely and more efficient decisions. The need to adopt an integrated approach to infrastructure management is widely recognized in industry and academia (InfraGuide, 2003; Halfawy et al., 2002; Grigg, 1999 and Lemer,1998). Throughout the last two decades, municipalities have made significant investments in implementing software tools that focus on infrastructure management processes [Vanier, 2001& Halfawy et al., 2006]. The majority of the software tools were developed to function as stand-alone systems, and many have limited or no capability for sharing or exchanging information with other tools. Halfawy et al. (2006) reviewed commercial asset management systems in Canada. This paper aims to providing asset managers with an objective review of existing systems and technologies, and to identify a number of considerations that need to be addressed in the process of selecting an asset management system. It also highlights areas where further research and development are needed in order to extend the scope and capabilities of existing systems. The Infraguide report suggested a systematic integration approach for renewal of municipal road, sewer and water system in an integrated manner (InfraGuide, 2003). This approach consists of 5 tasks: The approach is based on collecting data inventories, condition assessment and performance evaluation. This evaluation can be completed independently for road, sewer and water systems. It should be noted that critical infrastructure assets should be dealt with separately from the non-critical assets thought out all phases of the process. Finally Task 5- developing a sound renewal plan which includes economic analysis, coordination with growth needs regulations, and risk management. The report set a high level framework for municipalities to adopt which will standardize procedures and integration their asset management initiatives. It is recommended for Municipalities and governmental official to join forces to enforce and standardizing these practices among all municipalities. There are several applications in development of municipal infrastructure management systems [ Lee & Deighton, 1995; Quintero et al., 2003; Ferreira & Duarte, 2005; and Halfawy, 2008)]. Municipalities have recognized the need to combine and to explore relationships among various data resources, and, when possible, to make them available to other department. Ferreira and Duarte (2005) proposed base liner referencing system (LRS) for an Integrated Infrastructure Management System is an attempt to provide a common referencing platform in which different data types can be represented spatially in a network structure in a standardized manner. The base LRS is intended to be the nucleus of the relational 2

3 1494 database and the GIS. Ferreria only examined the road network using his proposed technique; also he didn t introduce optimization to his proposed model. On the other hand, Halfawy (2008) has proposed a detailed framework to address a number of issues, including: asset life-cycle data modeling, sharing and management, systematization of municipal processes, and integration of disparate software tools in a flexible and modular architecture. Halfway emphasizes on data modeling and integration of software tools, rather than the details of individual integration processes (e.g. risk assessment, renewal planning, and optimization). Halfway did identify decision trees and/or business barrier between various municipal departments. Therefore, formalizing a more-detailed integrated business process model is considered necessary. RESEARCH OBJECTIVES The objective of this research is to establish a methodology to facilitate decision making process that ensure reliable and optimum decision regarding pursing corridor rehabilitation for road, water and wastewater network. In order to fulfill this objective the following sub objectives are identified: 1. Conduct a comparable study for current asset management practices. 2. Investigate/ develop condition assessment, performance analysis, and risk models for roads, water, and wastewater infrastructure. 3. Establish a conceptual business process diagrams to identify business drivers as well as barriers and obstacles for integrated corridor rehabilitation. 4. Develop an optimal decision support system for corridor rehabilitation which satisfies two main objectives: 1) minimizing the total replacement cost required to maintain the infrastructure at a specified level of service; and 2) minimizing the economic loss of replacing any asset earlier than its anticipated useful life. 5. Create decision trees to serve as best-practice guidelines for integrated corridor rehabilitation. 6. Develop a prototype tool that implements the developed methodology and recommends an optimal replacement strategy. METHODOLOGY The plan set to achieve the objectives of this research consists of three main phases: I. review of current asset management practices; II. development of integrated asset management framework and model; and III. prototype development and implementation. The decision-support framework for integrated asset management model is illustrated in Figure 1. This proposed framework employs the following steps: (1) perform overall Risk analysis at project and network level; (2) conduct performance evaluation for each asset class independently; (3) assess the current physical condition of the pipe and road segment; (4) collect data and perform data gap analysis; (5) document a conceptual business process diagrams to identify business drivers as well as barriers and obstacles for integrated corridor rehabilitation; (6) develop decision analysis trees to set priorities for integrated 3

4 1495 corridor rehabilitation; and (7) implement optimization of repair/renewal cost and defining the best replacement interval via genetic algorism (GA). Figure 1 Research Methodology Flow Diagram CASE STUDY AND DATA ANALYSIS In order to demonstrate the model features, a case study has been utilized from the City of Guelph, ON, Canada. The City of Guelph has a population of approximately 115,000. The total operating budget is approximately $255 million with a capital budget of roughly $81 million. Task 1: Risk Assessment Framework Risk assessment, based on the objective assessment of the probability and consequences of asset failure, represents a practical and effective means of identifying and prioritizing capital and maintenance requirements. Although failure modes will differ, the risk dimensions should remain constant (economic, environmental, health & safety, etc.). The parameters affecting cost of rehabilitation and replacement of Road/Sewer/Water (R/S/W) infrastructure assets were selected based upon four overall Criticality Indices as follows: a) economic: influence of the asset s failure on monetary resources; b) operational: influence of the asset s failure 4

5 1496 on operational ability; c) social: influence of the asset s failure on society; and d) environmental: influence of the asset s failure on the environment. Next, the parameters that influence each of the previous Criticality Indices were established and are outlined as shown in Figure 2. Some of these parameter where missing (e.g. contaminated soil, utility locations, soil types in some areas). These missing parameters were assigned an overall weight of zero until the required data are collected and populated. Road segment Sewer mains Water mains Economic parameters Operational parameters Pipe size - Business disruption critical customer (major water users, hospitals, health clinics) Pipe shape - Damage to surrounding infrastrture (e.g. Near gas, electricity, etc) Pipe bury depth - Pipe size - Contaminated soil - Road width - - Low accessibility - Pavement /pipe material Environmental parameters Land use Water body proximity Road right of the way - - Sensitive area (watershed, environmental land reserves, etc) Road type (expressway, arterial, collector, local, etc) - - Soil type Traffic count Social parameters No diversion road Land use Transit route Figure 2 Risk model parameters Road segment Sewer mains Water mains A methodology to transfer these subjective factors into a point system was adopted. Figure 3 shows the economic parameters matrices which outlines this importance ranking system. Each Criticality Index was sub-categorized and weighted by a percentage multiplier that represents its relative importance among the other subcategories. These index category weightings must add up to 100%. The bottom level of each matrix consists of a point system that assigns a relative consequences importance score of 1 to 5 to each parameter value or group of values. Parameter Values can be a characteristic value or a Boolean Yes/No value that indicates whether the asset has the characteristic or not. A Parameter Value with a score of 5 would represent a major impact on large portion of the infrastructure system while a score 1 represent no significant impact on current asset. The Consequence of failure model can be formulated as in Equation 1. Where (OCI) represents the Overall Criticality Index; Equation 1 5

6 1497 (W Index i ): weight for each Index i (e.g. Economic, Operational, Social, and Environmental factor); (W Var j ): weight for each Variable j (e.g. pipe size, depth, material, etc.); and (S Var j ): Score for each Variable j (scores ranges from 1 to 5). The risk model and the overall Criticality model can be formulated as in Equation 2. Risk = Probability of Failure (PoF) Overall criticality index (OCI) Equation 2 Index Weight Variable Weight Pipe Size Pipe Depth Economic 0.25 Material 0.15 Land Use Low Accessibility Road type 0.15 Processing None None None Spatial intersect with 20m buffer zone around land use polygons (ave R.o.W.width) spatially intersect water bodies, railroads, and escarpment None Notes sewers with 225 mm diameter pipes are a concern because fittings for these pipes are either very difficult to find or they must be fashioned by City staff The City of Guelph has experienced problems associated with tuberculation and reduced flow in copper and galvanized watermains. Updated to reflect "High Density" in the downtown area specified by design. ROW are estimated based on road class type Valid Entries water water water R/S/W R/S/W R/S/W Value Score Value Score Value Score Value Score Value Score Value Score OTHER 1 OTHER 1 No 1 Local CI, DI, 3 Park 2 Yes 5 Collector CONC 4 Residential 3 Arterial Cu or GALV 5 Comm/Indst 4 Expressway HighDensity 5 HWY 5 Sewer Sewer Sewer Value Score Value Score Value Score OTHER Brick Figure 3 Economic Parameters Risk Matrices The establishment of risk indices, parameters, parameter scoring and weightings outlined within the previous matrices were a result of expert opinion from City of Hamilton, City of Guelph, Region of Durham, AECOM and several consultations of literature review (e.g. Fares, 2008; UMA-AECOM, 2008) using a mixed Delphi- Analytical Hierarchy Process (Delphi-AHP) (Tavana et al., 1993). Task 2: Performance Management Framework Infrastructure exists to service its users; performance measures are used to link asset and system operation to these business objectives. Performance management aims in defining a baseline for service levels for various asset classes investigated in this dissertation. Performance management is divided into three stages, first identify key performance indicators for each asset class, then establish and measure performance, finally linking the key performance indicators to a desired level of service. 6

7 1498 Asset management experts were engaged to examine business drivers and identify how 'service' defined for each asset class, rationalize performance measures that link business goals with infrastructure operation, and establish 'scales' and threshold values representing the acceptable operating range for each metric. If sufficient information is available to estimate deterioration profile for each R/S/W system component then it would be used to develop advanced deterioration rates to recommend best replacement time. Otherwise generic deterioration models will be adapted from literature until required data are available. Example for performance management evaluation includes: Road network: Statistical analysis of defects, pavement management system (PMS) results, results of visual inspections Sewer network: Structural integrity of pipelines (Statistical analysis of defects), results of CCTV inspections, Health and economics considerations from flooding damage, Infiltration/inflow considerations, and River water quality objectives. Water network: Statistical analysis of repairs and leaks, breakage history, water quality issues, remote field testing, field sampling, and testing. Task 3: Condition Rating Framework At a basic level, Condition Assessment can be a simple and effective means of identifying and prioritizing infrastructure needs. Building upon Task 2 outcomes, best practice and extensive experience, Task 3 will establish a framework for condition assessment of R/S/W system infrastructure, relating ratings and performance definitions across asset classes, and establishing a mechanism for rolling-up component condition to an overall system level. The analysis of condition includes the evaluation of structural indicators (the physical condition of an asset) and performance indicators (the capacity or utilization of an asset). 3.1Road Network The pavement condition rating evaluation consists of a pavement distress survey, and the calculation and reduction of data into a surface condition scale through pavement management application software. Stantec s data collection vehicle was used to complete field data collection component on the paved roads. More specifically, this unit surveyed the paved network and collected pavement distress, road roughness, and digital video data. The Pavement Management System (PMS) database contains pavement rehabilitation historical data which saves information on all types of previous treatments, their application date, and associated costs. 3.2 Sewer network The Water Research Centre (WRc) rating system and operator/inspector certification provides a consistent assessment of sewer structural condition. The intent of the condition assessment process is to attach a condition grade to each reach of sewer based on the worst defect in that reach. This is a two step process involving the assignment of a preliminary internal condition grade (ICG) based on the raw defect scores obtained from CCTV inspections using WRc rating system. The condition grades are broken down into five categories, or discrete condition states, based on the 7

8 1499 potential for collapse and the likelihood of further deterioration. A final performance grade, or Structural Performance Grade (SPG), will be assigned based on the consideration of supplementary data such as soil conditions, frequency of surcharging and cursory data as shown in Figure 4. SPG is assigned by qualified individual based on the impact/significance of observed defects on performance. The grade reflects the probability of collapse where, on a scale of 1 to 5, 1 is pristine and 5 the sewer collapsed or collapse is imminent. Figure 4 Sewer Structural Performance Assessment Procedure 3.3 Water network Condition assessment of water mains is challenging compared to other infrastructure assets because they are typically underground, operated under pressure, and mostly they are inaccessible. A condition assessment model, using artificial neural network (ANN) and analytical hierarchy process (AHP), is developed in order to set up rehabilitation priority for water mains. Various factors are incorporated in the developed model, such as physical characteristics (i.e.: pipe type, size, age, breakage rate), in situ environmental (i.e.: Cathodic protection, ground water level, soil type, surface type, and road type), and operational (i.e.: Hazen-Williams factor, operational pressure) (Al Barqawi, 2006). This model was altered and updated based on the availability of data for the existing case study. Also a new factor was introduced to this model which is the Hydraulic Capacity of water mains assets. Head loss (HL/L) was selected as the main criterion to assess hydraulic capacity. Head loss is required on an individual asset basis, and is obtained from the City s water hydraulic model. Task 4: Data Collection and Analysis A questionnaire and workshop interviews were conducted with municipal experts and consultants in (City of Hamilton, City of Guelph, Region of Durham, City of North Bay, AECOM and several consultations of literature review (e.g. Al Barqawi, 2006; Chughtai, 2007; and Fares, 2008). These questionnaires were designed to collect the opinions of practitioners regarding the main factors affecting the risk, performance 8

9 1500 indicator, and condition rating. In addition to having access to GIS geodatabase, maintenance records, and various consultant reports for City of Guelph, Ontario, Canada. It is anticipated to conduct another workshop interview with York region and an additional workshop with City of Guelph to establish level of service parameters. Task 5: Business Process Diagrams This Task aims to document a conceptual business process diagrams to identify business drivers as well as barriers and obstacles for integrated corridor rehabilitation. Identify Issues and needs pertaining to integration of decision making for infrastructure assets. These processes involve the generation and manipulation of large data sets with complex interrelationships.this will require that all R/S/W Asset be tracked in some sort of data repository. Although WAM (Synergen) is the core corporate work order system that supports tracking assets, many external data repositories do exist. Examples of these data repositories include: GIS, Pavement Management System (PMS), and Other Miscellaneous Spreadsheets. Most of these processes still remain largely heuristic and subjective in nature, documented in the form of guidelines or manuals of best practices. Task 6: Decision Matrix & Decision Trees Analysis The purpose of this module is to equip the asset manager with a consistent methodology for decision-making during the integrated Corridor rehabilitation planning cycle. Within a planning cycle, the asset manger must make one of three decisions for the each asset in the road right of the way: a) Schedule Intervention; b) Schedule Inspection; and c) Revisit at next planning cycle. Task 7: Optimization Implementation Integrated infrastructure management has multiple objective functions, which imply optimizing a number of different objectives at the same time. In practice, problems with multiple objectives are reformulated as single-objective problems by either forming a weighted combination of the different objectives or else replacing some of the objectives by constraints. Optimizing integrated infrastructure rehabilitation requires the simultaneous optimization of more than one objective function. Such as minimize replacement cost, minimize the economic loss of early replacement of an asset, maximizing network condition rating, minimize consequence of failure and maximize performance. Optimization using genetic algorism (GA) allows searching for decision variable that minimizes the objective function while satisfying certain constrains. ANTICIPATED RESULTS The selection of the most suitable rehabilitation decision depends on the segmentation approach used for road, water, sewer assets. Figure 5 (a), (b), &(c) shows a sample condition rating results of road, sewer mains, water mains segments respectively for City of Guelph, ON. Each R/S/W segment is color coded based on their physical condition. Figure 5 (d) shows a systematic approach for defining the available corridor rehabilitation segmentation option. Each of this segmentation 9

10 Pipelines 2010: Climbing New Peaks to Infrastructure Reliability Renew, Rehab, and Reinvest options will be considered for further analysis and optimization of the cost effective options. For example, if a sewer main is the driving asset for implementing corridor rehabilitation then the following options will be considered for replacement: I) replace sewer segment from manhole to manhole and the equivalent portion of water and road segment (in some cases, this option might not be practical. Each case will have to be dealt with separately) ; II) (if sewer segment < water segment) replace sewer mains equivalent to the water segment length, and road segment equivalent to water segment ( utilizing this option might result in replacement of non critical asset. Each case will have to be dealt with separately) ; III) replace sewer segment from manhole to manhole, water segment from node to node, and road segment equivalent to the longer of sewer / water segment. Figure 5 Sample Integration management framework results & analysis CONCLUSION In this research work, a decision-support framework for integrated asset management of road network, water distribution network, wastewater distribution network is established. The Integrated framework amalgamates performance indicators, condition assessment, and risk into Phase I of the integration. On the other hand phase II implements optimization of repair/renewal cost and defining the best replacement interval. A series of Workshop setting interviews were conducted with various municipalities departmental staff to gather all necessary information for framework guideline development. A case study has been utilized from the City of Guelph, ON, Canada. The proposed framework requires City staff to work together to develop a shared database of asset inventories, condition, performance and financial information to support decision making throughout the organization and community,

11 1502 resulting in efficient and effective management of infrastructure services. The developed risk and condition rating modules architects an integrated solution to provide enabling tools to support achievement of the integration of decision-making. The optimization phase of the framework, currently in progress, is being developed via genetic algorism (GA) using GIS platform. The result anticipated to generate a capital corridor rehabilitation program for the city s infrastructure. This framework helps Canadian municipalities to evaluate and select feasible optimal assets for integrated corridor rehabilitation. REFERENCES Al Barqawi, H. (2006). Condition Rating Models for Underground Infrastructure:Sustainable Water-Mains. Master Thesis, Concordia University, Montreal, QC. Chughtai, F. (2007). An Integrated Approach to Sewer Pipeline Condition Assessment. Master Thesis, Concordia University, Montreal, QC, Canada. Fares, H. (2008). Evaluating the Risk of Water Main Failure Using a Hierarchical Fuzzy Expert System. Master Thesis, Concordia University, Montreal, QC. Ferreira, A., & Duarte, A. (2005). A GIS-Based Integrated Infrastructure Management System. From Pharaohs to Geoinformatics, FIG Working Week 2005 and GSDI-8, Cairo, Egypt April Grigg, N. S. (1999). Viewpoint. INFRASTRUCTURE: Integrated Issue or Tower of Babel? Journal of Infrastructure Systems, 5, No. 4, pp ,. Halfawy, M. (2008). Integration of Municipal Infrastructure Asset Management Processes: Challenges and Solutions. Journal of Computing in Civil Engineering, 22,No. 3, pp Halfawy, M. R., Newton, L. A., & Vanier, D. J. (2006). Review of commercial municipal infrastructure asset management systems. ITcon, 11 (pg ). Halfawy, M. R., Pyzoha, D., & El-Hosseiny, T. (2002). An Integrated Framework for GIS- Based Civil Infrastructure Management Systems. Annual Conference of the Canadian Society for Civil Engineering. Montréal, Québec, Canada: Canadian Society of Civil Engineers. InfraGuide. (2003). An Integrated Approach to Assessment and Evaluation of Municipal Road, Sewer and Water Networks. Canada: InfraGuide best practice,national Guide to Sustainable Municipal Infrastructure. Lee, H., & Deighton, R. (1995). Developing Infrastructure Management Systems for Small Public Agency. Journal of Infrastructure Systems, Vol. 1, No. 4, Lemer, A. C. (1998). PROGRESS TOWARD INTEGRATED INFRASTRUCTURE- ASSETSMANAGEMENT. NRCC/CPWA Seminar Series Innovations in Urban Infrastructure (pp. 7-24). APWA International Public Works Congress. Quintero, A., Konaré, D., & Pierre, S. (2003). Prototyping an Intelligent Decision Support System for Improving Urban Infrastructures Management. European Journal of Operational Research,, 162, Tavana, M., Kennedy, D., Rappaprt, J., & Ugras, Y. (1993). An AHP-Delphi Group Decision Support System Applied to Conflict Resolution in Hiring Decisions. Journal of Managemant Systems, 5, No.1, pp UMA-AECOM. (2008). City of Hamilton Water Main Management Framework: Technical Memorandum No. 3 Water Main Rehab Optimization. Hamilton: UMA-AECOM technical report # D , September Vanier, D. (2001). Why industry needs asset management tools. Journal of Computing in Civil Engineering, Vol. 15, No., pp