10NCEE Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska NEW BUILDING DAMAGE AND LOSS FUNCTIONS FOR TSUNAMI C. A. Kircher 1 and J. Bouabid 2 ABSTRACT This paper describes new functions for determining the probability of damage to buildings and essential facilities due to tsunami inundation (flood) and tsunami lateral force (flow) hazards and for combining damage due to these hazard with that due to earthquake shaking (i.e., for evaluation of damage and los due to a local tsunami). The paper includes example values of damage and loss estimated using the new functions and compares these estimates with actual values of damage and loss for the 2012 Tohoku and other recent tsunami events. The new building damage and loss functions were developed as part of by a FEMA-funded project to develop a Tsunami Model for incorporation into the HAZUS Loss Estimation Technology. The objective that project is to develop a methodology to produce credible results of national and regional annualized tsunami losses and for evaluation of regional and local tsunami scenarios. 1 Principal, Kircher & Associates, Palo Alto, CA 94303 - cakircher@aol.com 2 Catastrophe Modeling Manager, Finance, Chubb Group of Insurance Companies, Warren NJ 07059 Kircher, CA, Bouabid, J. New Building Damage Functions for Tsunami. Proceedings of the 10th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.
10NCEE Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska New Building Damage and Loss Functions for Tsunami Charles A. Kircher 1 and Jawhar Bouabid 2 ABSTRACT This paper describes new functions for determining the probability of damage to buildings and essential facilities due to tsunami inundation (flood) and tsunami lateral force (flow) hazards and for combining damage due to these hazard with that due to earthquake shaking (i.e., for evaluation of damage and los due to a local tsunami). The paper includes example values of damage and loss estimated using the new functions and compares these estimates with actual values of damage and loss for the 2012 Tohoku and other recent tsunami events. The new building damage and loss functions were developed as part of by a FEMA-funded project to develop a Tsunami Model for incorporation into the HAZUS Loss Estimation Technology. The objective that project is to develop a methodology to produce credible results of national and regional annualized tsunami losses and for evaluation of regional and local tsunami scenarios. Introduction Building damage functions are developed for incorporation into a future Tsunami Model [1] of the HAZUS Loss Estimation Technology. These functions determine the probability of Moderate, Extensive and Complete damage to general building stock and buildings of essential facilities due to tsunami inundation (flood) and tsunami lateral force (flow) hazards. General building stock represents buildings of a given area (e.g., census block) grouped in terms of model building type, occupancy class and other building inventory characteristics. The same methods are used to estimate damage to essential facility buildings, although they are applied to specific buildings, rather than aggregated groups of buildings. For evaluation of local tsunami damage and loss, the probability of building damage due to tsunami is combined with the probability of building damage due to the earthquake that generated the tsunami. Figure 1 is a flowchart illustrating the components of the Tsunami Model and their relationship to building damage (i.e., direct physical damage to general building stock and essential facilities). Inputs to the calculation of tsunami damage to buildings include hazard data, namely tsunami flood (inundation height) and tsunami flow (momentum flux) data and earthquake damage data from the Earthquake Model (when evaluating local tsunami effects). Estimates of tsunami damage are used in the evaluation of damage to transportation and utility lifeline buildings [2], estimation of building-related debris, evaluation of sheltering needs (number of displaced households) and calculation of direct economic losses. The Tsunami Model uses the same model building types, occupancy classes, building damage states, etc., as those of the Earthquake Model, descriptions of which may be found in the HAZUS technical manuals for 1 Principal, Kircher & Associates, Palo Alto, CA 94303 - cakircher@aol.com 2 Catastrophe Modeling Manager, Finance, Chubb Group of Insurance Companies, Warren NJ 07059 Kircher, CA, Bouabid, J. New Building Damage Functions for Tsunami. Proceedings of the 10 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.
earthquake [3] and the Advanced Engineering Building Module [4]. Potential Earth Science Hazard Tsunami Flood (Water Depth) Tsunami Flow (Momentum Flux) Direct Physical Damage Earthquake Model Building Damage due to Earthquake General Building Stock and Essential Facilities Lifelines Transport Systems Lifelines Utility Systems Induced Physical Damage Direct Economic and Societal Loss Debris (building-related) Direct Economic Casualties Shelter Figure 1 Flowchart of the HAZUS Tsunami Model showing relationship of building (direct physical) damage to other model components Form of Damage Functions Building damage functions are in the form of lognormal fragility curves that relate the probability of being in, or exceeding, a discrete state of damage given the median estimate of the hazard parameter of interest (i.e., median peak inundation height or median peak momentum flux). Figure 2 illustrates fragility curves of Moderate, Extensive and Complete structure damage due to tsunami flow for an older mid-rise reinforced-concrete shear wall building. 1.0 0.9 0.8 Moderate Structural Damage Extensive Structural Damage Complete Structural Damage P[Moderate F = 4,500] = 0.28 = 0.73 0.45 = P[ Moderate F = 4.5k] - P[ Extensive F = 4.5k] Probability of Structural Damage 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 = 2,845 (β M,F = 0.74) = 4,987 (β E,F = 0.74) P[Extensive F = 4,500] = 0.18 = 0.45 0.27 = P[ Extensive F = 4.5k] - P[Complete F = 4.5k] P[Complete F = 4,500] = 0.27 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000 Median Peak Momentum Flux, F, (feet 3 /seconds 2 ) C = 7,129 (β C,F = 0.74) Figure 2 Example fragility curves of Moderate, Extensive and Complete structure damage due to tsunami flow (i.e., median peak momentum flux, F).
Conceptually, the form of the tsunami building damage functions is the same as the lognormal fragility curve format used by the HAZUS Earthquake Model. Each damage-state curve is defined by the median value and associated variability of the fragility parameter of interest. The variability of these fragility curves has two fundamental components, the variability of the median estimate of the hazard parameter (i.e., uncertainty in demand) and the variability of the median value of the damage state (i.e., uncertainty in capacity) for the hazard of interest. It is important to distinguish the demand and capacity components of uncertainty, since the demand uncertainty component used for evaluation of losses due to a deterministic (scenario) tsunami is not required for evaluation of probabilistic losses (using tsunami hazard functions that directly incorporate this uncertainty) or when comparing estimated losses with observed losses for which the values of hazard parameters are reasonably well known. Building Damage States Table 1 provides qualitative descriptions of Moderate, Extensive and Complete damage states due to tsunami for the structure, nonstructural components and building contents. These damage states are intentionally based on the same generic damage states as those of the Earthquake Model to permit combination of damage-state probabilities due to tsunami with damage-state probabilities due to earthquake (e.g., for evaluation of local tsunami damage and loss). While tsunami damage states are generically the same as those of the Earthquake Model, fewer damage states are required to adequately address tsunami losses. Slight damage is not required for tsunami, since it is difficult to distinguish from no damage, and of no practical significance to tsunami economic losses. Along similar lines, Moderate and Extensive states of damage are not used for all model building types and systems. In general, shorter (and lighter) model building types require fewer damage states to reliably calculate tsunami losses. Nonstructural systems and contents damage states are based solely on tsunami flood hazard (water depth based on median peak inundation height) assuming that if the building survives tsunami flow effects (e.g., is not washed away or otherwise does not sustain Complete damage to the structure), then damage and related losses to these systems are primarily a function of water depth. Conceptually, nonstructural systems and components located on fully inundated floors are considered to be ruined (i.e., 100 percent damage), and that only a few feet of water is required to significantly damage contents on a partially inundated floor. Of course, nonstructural systems and contents are also damaged by tsunami flow, but such damage is assumed to be adequately captured by damage due to inundation (e.g., since nonstructural systems and contents of fully inundated floors are assumed to be a complete loss). Additionally, to the extent that tsunami flow causes Complete damage to the structure, then nonstructural systems and contents are also assumed to have Complete damage. Thus, the probability of Complete structural damage (due to tsunami flow) is an important contributor to building damage and loss, particularly for model building types of shorter, lighter construction (consistent with observations of tsunami damage in past events).
Table 1 Qualitative Description of Building Damage States due to Tsunami Flow and Flood Model Building Type (Height/Weight) Low-Rise Light MBTs (W1, W2, S3, MH) Low-Rise - Other Mid-Rise - All High-Rise - All Damage States Moderate Extensive Complete Structure Damage due to Tsunami Flow Limited, localized damage to elements at lower floors. Diagonal cracks in shear walls, limited yielding of steel braces, cracking and hinging of flexural elements no or only minor permanent offsets (i.e., less than ½ inch per floor). Localized failure of elements at lower floors. Large diagonal cracks in shear walls, failure of steel braces, large flexural cracks/buckling of rebar, buckled flanges and connection failures large permanent offsets of lower stories. Localized erosion or scour, limited foundation settlement A significant portion of structural elements have exceeded their ultimate capacities; critical elements/connections have failed resulting in dangerous permanent offset, partial collapse, full collapse or building moved off foundation (e.g., washed away ). Extensive erosion or scour, substantial foundation settlement Nonstructural Components Ruined by Tsunami Flood (floors fully inundated, unless noted otherwise) Low-Rise - 1-Story Floor 1 (1/2 height) Floor 1 Low-Rise - 2-Story Floor 1 Floors 1-2 Mid-Rise - 5-Story 1 st Floor Floors 1-3 Floors 1-5 High-Rise - 12-Story 1 st Floor Floors 1-6 Floors 1-12 Low-Rise - 1-Story Contents Ruined by Tsunami Flood (floors fully inundated, unless noted otherwise) Floor 1 (3 feet) Low-Rise - 2-Story Floor 1 (3 feet) Floors 1 Floor 2 (3 ft.) Mid-Rise - 5-Story Floor 1 (3 feet) Floors 1 2, 3 (3 ft.) Floors 1 4, 5 (3 ft.) High-Rise - 12-Story Floor 1 (3 feet) Floors 1 5, 6 (3 ft.) Floors 1 11, 12 (3 ft.) Structure damage states are based solely on tsunami flow hazard. Conceptually, the structure is considered undamaged until lateral forces due to hydrodynamic loads, including the effects of debris impact, exceed the yield-force capacity of the structure. Structure damage increases with tsunami force until tsunami flow and debris forces exceed the ultimate-force capacity of the structure, and complete failure is assumed to occur. This approach focuses on the global damage to the structure, rather than on failure of individual elements. Hydrodynamic forces, can also cause localized damage to structural elements, including out-ofplane failure of walls, which could lead to progressive collapse of the building, and tsunami flow can also erode and scour the structure and compromise the foundation. While these are important modes of tsunami damage, quantification of building damage due to failure of individual structural elements, possible progressive collapse and loss of foundation integrity would require detailed structural information that is not available for generic model building types. Rather, tsunami damage functions use estimates of global building strength (which can be inferred from building type, height, age, etc.) to relate building damage states to tsunami flow and debris forces. Building Damage Due to Tsunami Inundation Building damage due to tsunami inundation is assumed to be similar to that caused by other
floods that have relatively slow water flow (e.g., riverine flooding). Building damage due to fast moving water flow is treated separately by damage functions that model damage due to hydrodynamic and related loads on the building (i.e., damage due to tsunami flow). Damage to nonstructural systems and contents due to tsunami inundation is related directly to the height of the water. Nonstructural systems and contents that are inundated are considered ruined (a total loss) and the damage state (Moderate, Extensive, or Complete) reflects the fraction of the nonstructural systems and contents in the building that are inundated. Consistent with the damage functions of the HAZUS Flood Model [5], contents which are primarily floor supported items are more vulnerable to water depth on a given floor than nonstructural components (which include ceilings, overhead lights, etc., as well as floor supported items. Hence, full-height inundation of a given floor is assumed necessary for 100 percent damage of nonstructural systems on that floor, whereas 3 feet of water on a given floor is assumed sufficient to cause 100 percent damage to building contents on that floor. Since damage is directly related to water depth, it is important to relate the elevation of building floors to the elevation of tsunami inundation, considering both the height, z, of the building s base above the sea level datum used to characterize tsunami inundation height, and the height of the first floor of the building above its base, h F. Figure 3 illustrates these parameters and their relationship to inundation height at building, R, inundation depth at building, H, and inundation depth relative to the first-floor of the building, H F. Figure 3 Schematic illustration of the relationship between inundation height at building location (R), inundation depth at building location (H), inundation depth relative to the first-floor (H F ), height of the first-floor above the base of the building (h F ), height of base of building above sea level datum (z), and model building height (h B ) above the first-floor level.
Building Damage Due to Tsunami Flow Building damage to the structure due to tsunami flow is assumed to be caused by hydrodynamic forces and debris impact forces. Tsunami flow forces also affect nonstructural components and contents (e.g., walls at the building s perimeter), but nonstructural and contents damage due to tsunami flow is assumed to be encompassed by tsunami flood damage functions (e.g., since walls affected by tsunami flow are also damaged by inundation). Further, and of most significance, nonstructural systems and contents of buildings with Complete structure damage due to tsunami flow are assumed to have Complete damage. The assumption of Complete building damage, if the structure sustains Complete damage, is consistent with observed damage due to tsunami (i.e., buildings whose structure failed were either collapsed or washed away). Development of building damage functions for tsunami flow utilizes an engineering approach that is based on the same concepts used for design of structures for tsunami lateral loads, such as those described in the Guidelines for Design of Structures for Vertical Evacuation from Tsunamis, FEMA P646 [6]. Few, if any, buildings have actually been designed for tsunami loads, but the design concepts provide a basis for characterizing the strength of model building types in terms of tsunami loads and parameters, namely hydrodynamic loads characterized by momentum flux. In addition to hydrodynamic forces, this approach also incorporates, in an approximate manner, additional lateral force due to debris impact forces. The simple, underlying notion of building damage functions for damage due to tsunami flow is to equate hydrodynamic forces, incorporating the effects of impulsive and debris loads, with the lateral force (pushover) strength of model building types as defined by the properties of capacity curves of the HAZUS Earthquake Model. The Earthquake Model is a convenient source of the approximate lateral strength of the structural system of model building types considering model building type, height, and seismic design level. Lateral strength is an inherent property of the structural system, whether the building is designed for earthquake (or wind) loads, or not designed for lateral loads (even buildings not designed for lateral loads still have inherent lateral strength). The Earthquake Model provides lateral strength for buildings not designed for earthquake loads (i.e., Pre-Code buildings), as well as those designed for earthquake loads. Equating tsunami hydrodynamic forces and earthquake forces assumes parity in the building damage states which is reasonable, except for collapse. Earthquake ground motions are vibratory in nature, often intense, but peak forces are typically of short duration (i.e., a few seconds, at most, in a given direction). Hence, buildings can reach their full strength, but not necessarily displace far enough to collapse before earthquake force reverses direction. In contrast, peak tsunami flow force is sustained in a given direction for a relatively long period of time (i.e., several minutes), and buildings that have reached their full strength are much more likely to collapse (and possibly be washed away with the flow). Thus, the likelihood of collapse given Complete damage for tsunami flow forces is much higher than that of earthquake. Evaluation of Building Damage The probability of a given building damage state due tsunami is combined with the probability of the same building damage state due to earthquake using Boolean logic assuming damage states
are statistically independent, except as noted below: 1. Probability of Complete (Extensive) Damage The probability of Complete (Extensive) damage also includes the joint probability of Extensive (Moderate) Damage due to tsunami and Extensive (Moderate) damage due to earthquake based on the assumption that Extensive (Moderate) damage due to tsunami occurring to a building that already has Extensive (Moderate) damage due to earthquake would result in Complete (Extensive) damage to the building. This concept applies to structure, nonstructural and contents damage states. 2. Probability of Nonstructural and Contents Damage due to Complete Structure Damage The probability of nonstructural and contents damage also includes the probability of Complete structure damage, based on the assumption that nonstructural systems and contents are completely damaged in a building that sustains Complete damage to the structural system. This concept applies to all nonstructural and contents damage states. Reasonableness Check of Building Damage Functions Example values of loss ratios are developed from building damage functions (assuming nil earthquake damage and loss) and compared to observations of building damage from recent tsunamis. The loss ratio is defined as the cost of building damage repair or replacement divided by the full replacement value of the building or subsystem of interest. Estimated values of the loss ratio are compared to observed damage (rather than damage-state probabilities) since the loss ratio represents the combined effects of damage to the structural system (due to flow) and nonstructural and contents damage (due to flood). Observations of building damage typically mix structural and nonstructural damage in the same damage state (i.e., structural damage is not clearly distinguished from nonstructural damage), making it difficult to compare individual estimates of structural, nonstructural and contents damage with observed damage. Estimated values of the loss ratio are expressed in terms of the depth of water above the base of the building (H) since this is the hazard parameter commonly used by post-tsunami investigations to report and evaluate observed damage to buildings. Since building damage functions define the probability of structure damage in terms of tsunami flow (momentum flux), Equation (6-6) of FEMA P646 [6] is used to convert structure damage expressed in terms of momentum flux to structure damage expressed in terms of water depth (H), assuming for these examples that the height of the building above sea level datum (z) is 20 feet. Loss ratio calculations are based on economic loss rates of 100 percent loss for Complete damage, 50 percent loss for Extensive damage and 10 percent loss for Moderate damage. These rates apply to the structure, nonstructural systems and contents of the building. Total building economic loss is based on the assumption that the structure represents 17%, the nonstructural systems represent 50 percent, and contents represent 33% of total model building replacement value (i.e., replacement value including all contents). These fractions of total building replacement value are generally representative of residential and commercial buildings. Example Building Loss Curves Figures 4a 4d show example loss ratio curves for 1-story and 2-story wood (W1 and W2) model building types and reinforced concrete shear wall low-rise (C2L), mid-rise (C2M) and
high-rise (C2H) model building types with either High Code (HC) or Pre-Code (PC) strength. Loss ratio curves are based on building damage function that include total uncertainty (i.e., functions include tsunami hazard uncertainty, as well as, building capacity uncertainty) or building damage functions that exclude hazard uncertainty. In all cases, the height of the firstfloor above the base of the building (h F ) is assumed to be 3 feet, corresponding to a height of 23 feet above sea level datum (z + h F ), and loss ratio curves reflect tsunami flow loads with debris. 1.0 1.0 0.9 0.9 Loss Ratio (fraction of replacement value) 0.8 0.7 0.6 0.5 1-Story Wood (W1 - HC) 0.4 2-Story Wood) (W2 - HC) 0.3 2-Story Concrete (C2L - HC) 5-Story Concrete (C2M - HC) 0.2 12-Story Concrete (C2H - HC) 0.1 Elevation of Building Base Elevation of First Floor 0.0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Median Peak Inundation Depth, R, (feet) Loss Ratio (fraction of replacement value) 0.8 0.7 0.6 0.5 1-Story Wood (W1 - PC) 0.4 2-Story Wood) (W2 - PC) 0.3 2-Story Concrete (C2L - PC) 5-Story Concrete (C2M - PC) 0.2 12-Story Concrete (C2H - PC) 0.1 Elevation of Building Base Elevation of First Floor 0.0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Median Peak Inundation Depth, R, (feet) a. High Code strength w/hazard uncertainty b. Pre-Code strength w/hazard uncertainty 1.0 1.0 0.9 0.9 Loss Ratio (fraction of replacement value) 0.8 0.7 0.6 0.5 1-Story Wood (W1 - HC) 0.4 2-Story Wood) (W2 - HC) 0.3 2-Story Concrete (C2L - HC) 5-Story Concrete (C2M - HC) 0.2 12-Story Concrete (C2H - HC) 0.1 Elevation of Building Base Elevation of First Floor 0.0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Median Peak Inundation Depth, R, (feet) Loss Ratio (fraction of replacement value) 0.8 0.7 0.6 0.5 1-Story Wood (W1 - PC) 0.4 2-Story Wood) (W2 - PC) 0.3 2-Story Concrete (C2L - PC) 5-Story Concrete (C2M - PC) 0.2 12-Story Concrete (C2K - PC) 0.1 Elevation of Building Base Elevation of First Floor 0.0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Median Peak Inundation Depth, R, (feet) c. High Code strength w/o hazard uncertainty d. Pre-Code strength w/o hazard uncertainty Figure 4 Plots of example loss ratio curves for 1-story and 2-story wood (W1, W2) and reinforced concrete shear wall low-rise (C2L), mid-rise (C2M) and high-rise (C2H) model building types with either High Code (HC) or Pre-Code (PC) strength and either including hazard uncertainty or excluding hazard uncertainty Figures 4a 4d illustrate the significant difference in the vulnerability of shorter, weaker (lighter) buildings (W1 and W2) to tsunami hazard, since these building types are both less stable against flow loads as well as more readily, extensively damaged by inundation than taller, stronger (heavier) buildings (e.g., C2L, C2M and C2H). It may be noted that the loss ratio curves of W1 and W2 building types also tend to be steeper when such are based on damage functions that exclude hazard uncertainty. It may also be noted that loss ratio curves based on damage functions that include hazard uncertainty can have appreciable non-zero loss ratios for median peak inundation heights less than the height of the base of the building above sea level datum (i.e., z = 20 feet in this example), reflecting the possibility that the peak inundation height has a 50 percent probability of being greater than the median value and therefore could be large
enough to exceed the height of the base of the building above sea level datum. Comparison of Estimated Building Damage with Observed Building Damage Table 2 compares water depths (H) corresponding to an 85 percent loss ratio (referred to as estimated damage ) with water depths corresponding to observed damage characterized as the initiation of partial failure of buildings in the 2004 Indian Ocean tsunami [7]and the median collapse of buildings in the 2009 Samoa tsunami [8] and the 2011 Tohoku tsunami [9]. Comparisons are made for HAZUS model building types of comparable construction to the building types damaged in these tsunamis. These model building types include, light-frame wood (W1) and timber (W2), low-rise unreinforced masonry (URML), low-rise reinforcedconcrete moment frames (C1L), low-rise reinforced-concrete shear walls (C2L), low-rise reinforced-concrete moment frames with masonry infill (C3L), and low-rise steel frames with cast-in-place concrete shear walls (S4L). Estimated damage is based on a relatively large (85 percent) loss ratio representative of buildings with a 50 percent (median) or greater probability of Complete structure damage. For comparison with observed damage (for which water depth is reasonably well known), estimated damage (loss ratios) is based on building damage functions excluding hazard uncertainty. Estimated damage is reported in Table 2 for model building types representing both newer, stronger (High Code) construction and older, weaker (Pre-Code) construction, the latter being more appropriate for comparison with observed building damage (other than that of modern Japanese buildings). Table 2 Comparison of water depth (H), in feet, of estimated damage corresponding to an 85 percent loss ratio and observed damage representing initiation of partial failure of buildings in the 2004 Indian Ocean tsunami [7] and median collapse damage of buildings in the 2009 Samoa tsunami [8] and 2011 Tohoku tsunami [9] Model Building Type Name No. of Stories Estimated Damage 85 Percent Loss Ratio (w/o hazard uncertainty) High-Code Strength Pre-Code Strength 2004 Indian Ocean Tsunami (Tinti et al., 2010) Observed Damage 2009 Samoa Tsunami (Reese et al., 2011) 2011 Tohoku Tsunami (Suppasri et al., 2012) W1 1 9.5 ft 6.5 ft 8.5 ft 5.3 ft 1&2 13.5 ft W2 2 17.5 ft 9 ft 15.9 ft URML 1 12.5 ft 13 ft 8.2 ft C1L 2 33 ft 22.5 ft 22 ft C2L 2 35.5 ft 24 ft 24 ft C3L 2 22.5 ft 19.5 ft S4L 2 35.5 ft 23 ft 31 ft It may be noted that the water depths of observed damage to similar construction vary significantly between the three tsunamis, indicating an imperfect understanding of actual building performance. Nonetheless these data are still useful for checking the reasonableness of
damage estimated using the new building damage functions. As shown in Table 2, water depths of estimated damage (based on Pre-Code strength) compare well with the water depths of observed damage to buildings in the 2004 Indian Ocean and 2009 Samoa tsunamis. For example, the estimated-damage water depth of 6.5 feet for light frame wood (W1) buildings, the most common building type falls, in between the observed-damage water depths of 5.3 feet (2009 Samoa tsunami) and 8.5 feet (Indian Ocean tsunami). Similarly, water depths of estimated damage to W1 and W2 building (based on High Code strength) bound those of observed damage to similar construction in the 2011 Tohoku tsunami. Conclusion New building damage and loss functions are developed for incorporation into a future Tsunamis Model of the HAZUS Loss Estimation Technology. These new functions consider damage due to both tsunami inundation (flood) and tsunami lateral force (flow) hazards. While structured to mesh with the model building types, damage-states and other HAZUS-specific methods and data, the underlying approach and concepts of the new building damage and loss functions are not limited to HAZUS and could be adapted for use with other tsunami loss estimation technologies and research programs. References 1. Atkins, 2013. Tsunami Methodology Development, prepared by Atkins Global (Atlanta) for the Federal Emergency Management Agency, Washington, D.C., June, 2013. 2. Graf, W. P., Y. Lee, and R. T. Eguchi, 2013. New Lifelines Damage and Loss Functions for Tsunami, Proceedings of the 10 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014. 3. FEMA, 2002. Earthquake Loss Estimation Methodology, HAZUS99-MR1, Advanced Engineering Building Module, Technical and User s Manual, prepared by National Institute of Building Sciences (NIBS) for the FEMA (Washington, D.C.: NIBS). 4. FEMA, 2011a. Multi-hazard Loss Estimation Methodology: Earthquake Model, HAZUS-MH MR4 Technical Manual, prepared by the National Institute of Building Sciences (NIBS) for the Federal Emergency Management Agency (Washington, D.C.: NIBS). 5. FEMA, 2011b. Multi-hazard Loss Estimation Methodology: Flood Model, HAZUS-MH MR4 Technical Manual, prepared by the National Institute of Building Sciences (NIBS) for the Federal Emergency Management Agency (Washington, D.C.: NIBS). 6. FEMA, 2012. Guidelines for Design of Structures for Vertical Evacuation from Tsunamis, FEMA P646, April 2012, FEMA (Washington, D.C: FEMA). 7. Tinti, S., R. Tonini, L. Bressan, A. Armigliato, A. Gargi, R. Guillande, N. Valencia and S. Scheer, 2011. "Handbook of Tsunami Hazard and Damage Scenarios," SCHEMA Project, JRC Scientific and Technical Reports, EUR 24691 EN, 2011 (Joint Research Centre, Institute for the Protection and Security of the Citizen, Bologna, Italy). 8. Reese, Stefan, Brendon A. Bradley, Jochen Bind, Graeme Smart, William Power, James Sturman, 2011. "Empirical building fragilities from observed damage in the 2009 South Pacific tsunami," Earth-Science Reviews, Elsevier (National Institute of Water and Atmospheric Research). 9. Suppasri, Anawat, Erick Mas, Shunichi Koshimura, Kentaro Imai, Kenji Harada, Fumihiko Imahura, 2012. Developing Tsunami Fragility Curves from the Surveyed Data of the 2011 Great East Japan Tsunami in Sendai and Ishinomaki Plains, Coastal Engineering Committee, Japan Society of Civil Engineers, Coastal Engineering Journal, Vol. 54, No. 1, March 24, 2012, (World Scientific Publishing Company: www.worldscientific.com).