Eart hq u a ke R i s k i n M u l ti fam ily Resident i a l Bu i l d i n g s Europe and Central Asia Region EAR TH Q UA K E RI SK I N M U L T I F AM I L Y R E S I D E N T I A L B U I L D I NGS EUR O P E A N D C E N TRA L A S I A R E G I O N Disclaimer This document is the product of work performed by World Bank staff and consultants. The findings, interpretations, and conclusions expressed in this document do not necessarily reflect the views of the Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown in any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Suggested Citation: World Bank Group. 2020. “Earthquake Risk in Multifamily Residential Buildings, Europe and Central Asia Region.” World Bank Group, Washington, DC. Acknowledgments This report was prepared as part of regional technical assistance titled “Strategies and Options for Scaling Up Disaster Risk Management in ECA Countries,” supported by a grant from the Global Facility for Disaster Reduction and Recovery. The World Bank team was led by Alanna Simpson and Maryia Markhvida. The analysis was prepared by a team of experts from Mott MacDonald in Bulgaria, led by Manya Deyanova, Joao Travanca, Stoyan Andreev, and Maria Christoskova, with project direction by Anton Andonov. Technical review was carried out by Pablo Camilo Heresi Venegas, and the writing was by Zahraa Saiyed of Scyma Consulting. The report was edited by Anne Himmelfarb and designed by Zahraa Saiyed (Scyma Consulting). Cover photos: (Top) “Boy in the Window.” A young boy outside a damaged home near Spitak, Armenia. © Razmik Zackaryan (Center) Nine-story precast-frame building collapse in Gyumri after the 1988 Spitak earthquake, Armenia. Credit: Wyllie, L. A., and J. R. Filson, 1989. ©Earthquake Spectra, EERI. Reprinted with permission. (Bottom) “Terremoto Albania [2019]” by Dipartimento Protezione Civile, licensed under CC BY 2.0 / Added color transparency. Back cover photo: “Relief.” The Bulvar district of Gyumri [Armenia] was built for earthquake victims. © Lilit Galstyan. Bottom-right photo: Image from above of buildings collapsed following the [2019 Albania] earthquake. “Terremoto Albania” by Dipartimento Protezione Civile, licensed under CC BY 2.0 / Added color transparency. ©The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A August 2020 CONTENTS iv EXECUTIVE SUMMARY 22 CITY EARTHQUAKE RISK PROFILE S ABBREVIATIONS 23 Understanding the City Earthquake Risk Profiles AAL average annual loss 6 INTRODUCTION 24 Tirana, Albania AAL% average annual loss ratio 7 MOTIVATION OF THIS STUDY 26 Durrës, Albania BLK block 28 Mostar, Bosnia and Herzegovina DFW dual frame-wall 7 Earthquake Risk in Multifamily Buildings 30 Sofia, Bulgaria ECA Europe and Central Asia 8 Cities Included in Risk Analysis 32 Plovdiv, Bulgaria GDP gross domestic product 8 Output of Earthquake Risk Analysis 34 Zagreb, Croatia GEM Global Earthquake Model 36 Rijeka, Croatia GNI gross national income 10 TYPES OF MULTIFAMILY 38 Budapest, Hungary GNIPC gross national income per capita BUILDINGS IN ECA 40 Chişinău, Moldova LPB large panel building 12 Unreinforced Masonry (URM) Buildings 42 Podgorica, Montenegro M magnitude 12 Reinforced Concrete Frame (RCF) Buildings 44 Skopje, North Macedonia RCF reinforced concrete frame 13 Dual Frame-wall (DFW) System Buildings 46 Bucharest, Romania RCW reinforced concrete wall 13 Precast Large Panel Buildings (LPB) 48 Iași, Romania SCS slab-column system 14 Block (BLK) Buildings 50 Belgrade, Serbia URM unreinforced masonry 14 Slab-column System (SCS) Buildings 52 Bratislava, Slovakia 15 Reinforced Concrete Wall (RCW) System Buildings 54 Ljubljana, Slovenia 56 Almaty, Kazakhstan 16 METHODOLOGY 58 Shymkent, Kazakhstan 60 Bishkek, Kyrgyz Republic Figure 1 Selected Historic Earthquakes in 16 Seismic Hazard Model 62 Osh, Kyrgyz Republic ECA Region 16 Exposure Model 64 Dushanbe, Tajikistan Figure 2 Earthquake Hazard Map in ECA 16 Vulnerability Model 66 Ashgabat, Turkmenistan Region, Global Earthquake Model, 17 Probabilistic Seismic Risk Analysis 68 Tashkent, Uzbekistan 2018 70 Yerevan, Armenia Figure 3 Cities in ECA Multifamily 18 OVERALL FINDINGS 72 Gyumri, Armenia Residential Building Earthquake Risk Study (Map) 19 Results Snapshot 74 Baku, Azerbaijan Figure 4 Building Construction Period of 19 Multifamily Building Vulnerabilities 76 Tbilisi, Georgia Multifamily Building Types in Study 20 RECOMMENDATIONS 78 FUTURE WORK Figure 5 Relative Vulnerability of Multifamily Building Types in Study 21 IMPACT OF STUDY 80 WORKS CITED Figure 6 Models Used in Probabilistic Seismic RIsk Analysis 85 CITY PROFILE SOURCES Figure 7 Using the City Profiles: Quick Facts Table 1 Cities Investigated in Study Table 2 Exposure Data Reliability Matrix Table 3 Vulnerability Data Reliability Matrix Table 4 Summary of Earthquake Risk Assessment Results Box 1 Terminology E XECUTIVE SU M M A RY INTRODUCTION cities investigated in study METHODOLOGY Multifamily residential buildings in earthquake-prone Country City A probabilistic seismic risk analysis of multifamily cities across Central Europe, Balkans, and Central Asia Albania Tirana, Durrës buildings was conducted in order to predict damage have proved to be vulnerable and have revealed several Bosnia & Herzegovina Mostar and losses due to earthquakes in the 27 cities. The risk structural deficiencies during previous earthquake Bulgaria Sofia, Plovdiv analysis used a combination of hazard, exposure, and events. Earthquakes in these regions have damaged vulnerability models, and was performed using an open Croatia Zagreb, Rijeka certain types of multifamily buildings beyond repair or source software called OpenQuake, developed by the Hungary Budapest have caused complete collapse, resulting in extensive Global Earthquake Model (GEM). Moldova Chișinău homelessness and many fatalities. Montenegro Podgorica Earthquake Hazard Model: The overall earthquake The vulnerability of some of these buildings can be North Macedonia Skopje hazard model is comprised of three existing regional attributed to the mass-produced nature of multifamily Romania Bucharest, Iași models (ESHM13, EMME14, EMCA15) and modifications residential buildings as well as lack of maintenance, Serbia Belgrade that are based on two local seismic hazard models oversight, and disinvestment over many decades. This Slovakia Bratislava (BIGSEES15, BUL18). The regional models are obtained study investigates earthquake risk of multifamily Slovenia Ljubljana from the Global Earthquake Model GEM2018 global buildings constructed before 2000 across 27 cities in Kazakhstan Almaty, Shymkent mosaic. 20 countries within Europe and Central Asia (ECA) Kyrgyz Republic Bishkek, Osh to better understand their behavior and potential Tajikistan Dushanbe Exposure Model: The building exposure model was losses when subjected to earthquakes, and to inform Turkmenistan Ashgabat developed by analyzing country-specific, cross-country, recommendations for mitigation. The table on the right Uzbekistan Tashkent and structural typology-specific information sources. For shows the cities and countries considered within this Armenia Yerevan, Gyumri each building type, many data sets were evaluated to study. The intended audience for this report includes Azerbaijan Baku understand the number of buildings, number of dwellings, government officials, policy and decision makers, Georgia Tbilisi and number of occupants, as well as the associated engineers, researchers and analysts, among other construction costs. The sources used include census data stakeholders and beneficiaries. Two supplementary (mainly 2011 and 2014 data), national energy efficiency technical documents provide additional insight into the multifamily building types in study programs, building inventory reports, past projects on inputs, methods, and results, and can be found in the and relative earthquake vulnerability risk assessment, international building databases, post- Works Cited section (Mott MacDonald 2020a, 2020b). Low Medium High earthquake reports, technical publications, and satellite images. It should be noted that the reliability of exposure Unreinforced unconfined TYPES OF MULTIFAMILY BUILDINGS IN ECA masonry UR M data varies from city to city, based on the information confined that was available. The relative reliability of the exposure The study evaluated seven broad classes of multifamily Reinforced data is shown in the exposure reliability matrix (Table 2) precast buildings found in the cities and countries under concrete RC F in the Methodology section. frame cast-in-place consideration. The buildings are categorized into classes based on their behavior when subjected to earthquake Dual-frame Vulnerability Model: The vulnerability models used in DFW dual frame-wall shaking, and according to the data available across wall system the risk analysis are based on existing damage and loss the different cities. Building classes are characterized models from various sources. Over 400 damage and loss by specific features that affect their earthquake Precast large models were reviewed and the most appropriate ones for panel LPB precast large panel performance: horizontal and vertical structural elements each city and building type were selected. The models and the material type, connections, building height, were used to quantify several earthquake consequences, construction quality, potential deterioration, and period Block BLK block including: direct financial losses from building damage, of construction. The diagram on the right shows the permanent relocation of occupants based on building seven building types and the acronyms used in this study, Slab-column SCS lift-slab repair time, and fatalities due to building damage. Similar as well as the relative vulnerability of major subtypes system post-stressed to the exposure data, vulnerability data varies by city. indicated as low, medium, and high. For example, the Reinforced The relative reliability of vulnerability models is shown sliding unreinforced masonry building type has two considered concrete wall RC W in the vulnerability reliability matrix (Table 3) in the system tunnel & standard subtypes: unconfined and confined masonry. Methodology section. iv RESULTS SUMMARY Results for each of the 27 cities are presented in the losses in many of the cities studied. RCF vulnerability is City Earthquake Risk Profiles section, which include an attested to by the immense damage and collapse of many earthquake history background, predicted estimated precast reinforced concrete frame buildings witnessed in losses per building type and impact on occupants, and the 1988 Spitak earthquake in Armenia. city-specific reliability and exposure information. The following are key findings from the risk analysis results RECOMMENDATIONS AND POTENTIAL APPLICATIONS of the 27 cities. Given the assumptions and limitations of this study (see Overall population exposure: This study finds that on the Methodology section for more details), the following average, approximately half of the population of the cities recommendations and applications are proposed for investigated reside in multifamily residential buildings seismic risk reduction of multifamily buildings in the constructed before 2000. Europe and Central Asia region. High population exposure: Bucharest has the most •  Prepare databases with building information relevant inhabitants living in pre-2000 multifamily housing, with to the earthquake risk assessment on a large scale. nearly 90% of the total city population residing in these •  Collect data on buildings’ current condition and level of multifamily buildings. This is followed by Tashkent and maintenance to understand their vulnerability. Belgrade. •  Develop an appropriate earthquake risk reduction Population in high-risk buildings: High-risk building types strategy and prioritization plan for investment in pre- are considered to be the top two building types within disaster risk reduction measures. each city that have the highest average annual loss ratios* •  Integrate earthquake risk reduction measures into due to building damage. By this definition, Bucharest has ongoing investment programs. the largest portion of its population residing in high-risk reinforced concrete frame and unreinforced masonry •  Update current codes and legislation related to safety buildings. The next two cities with a large segment of assessment and retrofit of existing buildings. its population residing in its respective high-risk building •  Increase public awareness about earthquake risk. types are Belgrade and Tashkent. • Design and promote the adoption of risk financing and Absolute building damage loss: Bucharest is expected risk transfer strategies by owners and residents. to have the highest total losses from future earthquakes, See the Recommendations section for further detail. with almost €200 million in average annual losses. This is followed by Yerevan and Tbilisi, which are expected to Results of this study offer government officials, decision have average annual losses of nearly €73 million and €60 makers, engineers, researchers, and stakeholders, million, respectively. among other users, a high-level understanding of earthquake risk to multifamily residential buildings Vulnerable building typologies: It is found that and are intended to be a starting point for addressing unreinforced masonry buildings (URM) are expected to seismic risk and prioritizing interventions. The scoping, experience the most damage out of all seven building procedural documentation, methodology, and results of classes in the considered cities. URMs contribute to this investigation offer an opportunity for further, more a significant portion of the direct financial losses, detailed studies and continued work towards earthquake number of fatalities, and number of people who will risk reduction. • be permanently displaced after an earthquake, even in regions with lower seismicity. Reinforced concrete frame buildings (RCF) are also large contributors to annual Image Credit: Precast large panel buildings in Bulgaria, by Maryia Markhvida. *High risk building types based on earthquake loss are determined by the expected average annual loss ratio, or average annual loss as a percentage of replacement value (AAL%). The replacement value includes only the multifamily buildings assessed in this study. See Box 1 for more information. v I NTR OD UC TI O N Much of the Europe and Central Asia (ECA) region Over the past several decades, mass-produced housing is exposed to earthquake hazard: from historical has suffered from lack of maintenance and oversight as earthquakes to today’s events, earthquakes have well as disinvestment, and as a result certain types of created dramatic consequences for ECA countries and multifamily buildings have become some of the most their populations. Many catastrophic earthquakes have vulnerable in the region. Moreover, these buildings occurred within the ECA region, with lasting impacts represent a significant portion of the building stock and including deaths, displacement of people, destruction of house a large segment of the population. Earthquake assets, impediments to business and economic growth, events in the past (see Figure 1) have demonstrated the and overall stunting of affected countries’ development fragility of multifamily residential buildings: many have for years and even decades. At least 20 countries across been damaged beyond repair or completely collapsed, the region have a 10–20 percent chance of experiencing causing extensive homelessness and many deaths. a major earthquake in the next 50 years (Mathema and Simpson 2018). Past regional earthquake events have This report investigates earthquake risk to multifamily also shown that a significant portion of the damaged buildings across 27 cities within Europe and Central Asia buildings are residential. (Table 1). The 27 City Earthquake Risk Profiles are based on a probabilistic seismic risk analysis that quantifies the The sociopolitical and economic transition of countries extent of the risks associated with multifamily residential in Central Europe, the Balkans, Central Asia, and the buildings. The buildings considered in this study include Caucasus since the 1990s has revealed the challenges of those designed and constructed before 2000; this focus ensuring safe housing in regions exposed to earthquake captures the typical pre-2000 construction practices perils. Many buildings in these regions exhibit inherent within the 20 countries described herein. This high-level structural vulnerabilities due to their mass-produced assessment seeks to provide meaningful information nature across both seismic and non-seismic regions and to governments and stakeholders on the underlying the limited knowledge of building earthquake behavior. earthquake risks to multifamily housing and supports Image Credit: Precast large panel buildings in Baku, Buildings have also been weakened by post-construction prioritization of risk reduction efforts and further studies by Maryia Markhvida. modifications, or have outlived their design life span. for resilient cities and communities. • figure 1. selected historic earthquakes in eca region Major Earthquakes Central Asia 2011 Fergana Valley M6.1 1885 Belovodskoje M6.9 2000 Balkanabat M7.0 1946 Kazandzhik M6.9 1985 Kayrakkum M5.9 1992 Suusamyr M7.5 1902 Andizhan M6.4 1948 Ashgabat M7.3 1966 Tashkent M5.1 1907 Karatag M7.4 1911 Almaty M7.7 1989 Gissar M5.3 1976 Gazli M.7.0 2015 Sarez M7.2 1949 Khait M7.4 2008 Nura M6.6 1984 Gazli M7.0 1400 1900 1950 2020 Banská Štiavnica M? 1443 Idrija M6.5 1511 Dubrovnik M7.2 1667 Garni M7.0 1679 Komárom M6.2 1763 Vlorë M6.6 1851 Zilina M? 1858 Ljubljana M6.1 1895 Chirpan M7.0 1928 Zangezur M6.3 1931 Friuli M6.5 1976 Vrancea M7.5 1977 Montenegro M6.9 1979 Vrancea M7.2 1986 Spitak M6.8 1988 Vrancea M7.0 1990 Vrancea M7.8 1940 Dunaharaszti M5.8 1956 Skopje M6.0 1963 Banja Luka M6.1 1969 Racha M7.0 1991 Bovec M5.8 1998 Baku M6.8 2000 Mamurras M6.4 2019 Zagreb M5.4 2020 Major Earthquakes Europe and South Caucasus NOTE: Earthquakes included for the ECA region are not exhaustive. See the City Earthquake Risk Profiles for more information on cities and countries affected by specific earthquakes. 6 M OTIVATIO N O F TH I S STU D Y EARTHQUAKE RISK IN MULTIFAMILY BUILDINGS have suffered deterioration, or have been subject to post- table 1. cities investigated in study construction modifications, putting their structural safety Multifamily buildings comprise 40–70 percent of the in question. This study finds that on average, half of the % city population in existing ECA housing stock (Mathema and Simpson 2018) population of all the cities investigated live in multifamily Country City multifamily buildings and house a large portion of the population in many buildings designed and constructed before 2000. cities within the region. From the 1950s to the 1990s, Albania Tirana 34% shortages of housing in Central Europe and Central Asia Past disasters like the 1988 Spitak earthquake in Armenia Durrës 27% were eased by mass construction of buildings, through demonstrate the tragic consequences of poor-quality Bosnia & Mostar 40% processes that followed standardized designs and design and construction. The earthquake killed over Herzegovina employed prefabrication of building components for 25,000 people, injured 130,000, and left over 500,000 Bulgaria Sofia 71% economic efficiency. people homeless (World Bank Group 2017). Many of the Plovdiv 69% fatalities were due to vulnerable multifamily buildings Croatia Zagreb 52% The multifamily housing built during this period was that were severely damaged or suffered complete Rijeka 54% often designed and constructed with low quality control, collapse. The town of Spitak was wholly destroyed, and Hungary Budapest 68% an omission that increased buildings’ vulnerability. Many cities like Gyumri are still recovering decades later. Moldova Chișinău 81% of the buildings have also been inadequately maintained, Montenegro 33% Podgorica North Macedonia Skopje 29% Romania Bucharest 89% Iași 71% Serbia Belgrade 74% Slovakia Bratislava 64% Slovenia Ljubljana 67% Kazakhstan Almaty 36% Shymkent 26% Kyrgyz Republic Bishkek 38% Osh 44% Tajikistan Dushanbe 54% Turkmenistan Ashgabat 39% Uzbekistan Tashkent 62% Armenia Yerevan 71% Gyumri 39% Nine-story precast-frame building collapse in Gyumri after Partial collapse of masonry building after 1988 Spitak Azerbaijan Baku 48% the 1988 Spitak earthquake, Armenia. Credit: Wyllie, L. A., earthquake, Armenia. Credit: WHE, n.d. (Report 202). Georgia Tbilisi 82% and J. R. Filson, 1989. ©Earthquake Spectra, EERI. Reprinted © World Housing Encyclopedia, EERI, and IAEE. with permission. Mu ltifam ily bui l di ngs compr i se M a ny of the b uild ing s ha v e 40 to 70 % of the exi sti ng b een ina d eq ua tely m a inta ined , Eu ro pe an d Centr al Asi a r egi on ha v e s uf fered d eteriora tion, h o u sin g stock. or ha v e b een s ub jec t to p ost- (M at hem a and Si m ps o n 2 0 1 8 ) construc tion m od if ica tions , p utting their struc tura l s a fety in q uestion. 7 CITIES INCLUDED IN RISK ANALYSIS OUTPUT OF EARTHQUAKE RISK ANALYSIS In response to the general seismicity of the region (Figure The probabilistic seismic risk analysis of regional 2) and the known consequences of past earthquakes, this multifamily residential housing stock evaluates buildings’ study carried out quantitative earthquake risk analysis structural characteristics, quantifies the population living for 27 of the largest and most risk-prone cities in 20 within these buildings, and investigates the performance countries across Europe and Central Asia that include of the buildings when subjected to regional earthquake mass constructed multifamily buildings. These 27 cities hazard. The results of the assessment provide information were chosen by ranking a combination of the earthquake on the distribution of losses (in euros) arising from hazard and city size,1 then selecting for each country one earthquake-induced building damage, loss of life, and Baltic Sea or two cities that are most susceptible to earthquake displacement of population. Information on results is in shaking and to building and population impacts (cities the Overall Findings section. • are shown in Figure 3). It should be noted that the cities selected through this ranking process do not represent an exhaustive list; many other risk-prone cities and areas exist in the region. figure 2. earthquake hazard map in eca region, global earthquake model, 2018 Bratislava slovakia Ljubljana slovenia Budapest Iași hungary romania Zagreb croatia Mostar Belgrade Bucharest Rijeka bosnia & serbia romania croatia herzegovina Sofia bulgaria Plovdiv Source: Pagani et al. 2018. © 2020 GEM Foundation and Partners. Licensed under CC BY-NC-SA 4.0 (CC BY-NC-SA Podgorica bulgaria 4.0). Note: PGA = peak ground acceleration. montenegro Durrës Skopje albania north macedonia Tirana albania M 1 The seismic hazard used in the risk index for the ranking process included peak ground acceleration for a 475-year return period. The size of city was based on the number of exposed buildings and the size of the affected population. See Technical Report 1 by Mott MacDonald (2020a) for more information. 8 figure �. C ITIES IN E C A M U LTI FA M I LY R E S I D E N T I A L B U I L D I N G E A R T H Q U AK E RISK STU DY Chişinău moldova Almaty kazakhstan Black Sea Shymkent Bishkek Tbilisi kazakhstan kyrgyz republic georgia Caspian Sea Gyumri Baku armenia azerbaijan Tashkent Osh uzbekistan kyrgyz republic Yerevan armenia Dushanbe Ashgabat tajikistan turkmenistan Mediterranean Sea 9 TYP ES OF M U LTI FA M I LY BUIL D I N G S I N E C A The scope of this study is limited to evaluation of the balconies and extension of rooms creating a cantilever earthquake risk associated with multifamily residential outside of the building. buildings constructed before 2000; single-family homes and nonresidential buildings are not included in the The following classification was developed specifically investigation. After 2000, privatized development created for typologies of multifamily housing in the considered varying building typologies with differing performance countries. Buildings are subdivided into seven classes during earthquakes. Because the data on these are according to their behavior when subjected to limited and the building types cannot be compared for earthquake shaking, and according to the data available performance among the countries and cities of interest, across different cities. Buildings are first classified at the post-2000 multifamily buildings are not included in this regional level, and then further categorized according study. to specific features that affect earthquake performance in each country. These features include horizontal and In neighborhoods across the studied cities, rapid vertical structural elements, connections, building urbanization and attempts to reduce housing shortages height, construction quality, potential deterioration, led to the construction of many buildings using standard and period of construction. It is important to note designs. In large urban centers such as the ones that the classification is not exhaustive but is generally considered in this study, more than half of the population compatible with other existing building taxonomies. The lives in apartments within such standardized multifamily seven building descriptions in this section indicate the buildings. These buildings range in height from 3 to 20 approximate construction period of the building type stories and contain 3 to more than 100 apartment units, within the 20 countries and 27 cities studied, along with and they represent a high concentration of risk due to significant earthquakes in which the building class faced the large percentage of the population they house. The varying degrees of damage, including total collapse.2 It risk is exacerbated due to shared ownership by residents can be assumed that multifamily buildings built between and lack of consensus in decision making around capital 1950 and 2000 were designed and constructed to meet investment, as well as by poor maintenance and inability low to mid code levels.3 of some lower-income households to contribute to building upgrades. Moreover, apartments in these The City Earthquake Risk Profiles include information buildings often did not match the needs of their owners on the distribution of the building types within the 27 and therefore post-construction modifications are very studied cities as well as the specific information related common, including removal of internal walls, addition of to the loss incurred by each type. 2 The technical reports accompanying this document use a slightly different nomenclature for the seven building classes. See Mott MacDonald (2020a, 2020b). 3 Low to mid code refers to the level of earthquake-resistant design concepts prescribed in the building code and can vary by construction period, Image Credit: “Damaged Building Ilicia [Zagreb, 2020]” by Franjo Tahy, country, and (at times) city. licensed under CC0 1.0. 10 11 UR M UNR E I N FO RC E D M A SO NR Y ( U R M ) B U I L D I N G S Construction period of URM in Relative Vulnerability countries studied: 1920-present Low Medium High Unreinforced masonry buildings have walls composed most vulnerable in an earthquake. The wall and facade unconfined of brick, stone, or cement blocks that lack steel elements of URMs can fall away from the building and are confined reinforcement bars embedded within the masonry. especially dangerous to people outside of the building. URM buildings in this study are further classified as Because of their poor performance in earthquakes “confined” or “unconfined.” Confined masonry buildings around the world, unconfined URM multifamily buildings have masonry walls that are confined by reinforced are no longer built in medium- to high-seismic-risk concrete tie beams and tie columns around the wall. regions in ECA. This typology also allows for relatively Unconfined masonry buildings have walls that do not easy structural modifications, such as ad hoc removal of include such reinforced concrete elements around the walls on the ground floor, which can greatly increase its Undamaged unreinforced masonry wall. They perform much worse than confined masonry vulnerability in the face of an earthquake. building, Armenia. buildings in earthquakes. Confined masonry buildings Credit: JICA 2012. are constructed by installing the masonry walls first and Past earthquakes in the investigated region in which subsequently installing concrete beam (horizontal) and unreinforced masonry buildings faced damage, partial column (vertical) elements. In this configuration, the collapse, or total collapse include the following: the masonry wall is better able to resist earthquake loading. It 1988 Spitak earthquake (Armenia), 1985 Kayrakkum should be noted that this study did not include reinforced earthquake (Tajikistan), 1992 Suusamyr earthquake (Kyrgyz Republic), 1976 Fruili earthquake and 1998 Bovec Damage to unreinforced masonry masonry buildings—those that include steel bars inside buildings after the 1988 Spitak the masonry wall—as a subcategory, since such buildings earthquake (Slovenia), and 1940, 1977, 1986, and 1990 earthquake, Armenia. are rarely found in the ECA region. Vrancea earthquakes (Romania, Bulgaria, Moldova). Credit: WHE, n.d. (Report 202). © World Housing Encyclopedia, URM buildings are among the building types that are EERI, and IAEE. RC F REINFO RC E D C O N C RE TE F R A M E ( R C F ) B U I L D I N G S Construction period of RCF in countries Relative Vulnerability studied: 1945-present Low Medium High Reinforced concrete frame buildings are those in which are often less vulnerable than those with welded on- precast more than half of the earthquake resistance is provided site connections because the welded connections are cast-in-place by reinforced concrete beams and columns connected prone to rusting and are often placed in critical stress as frames. This category is further classified into precast zones of the beam-column interface. For the cast-in- concrete frames and cast-in-place concrete frames. place subclass, the vulnerability is greatly influenced by Precast frames consist of prefabricated reinforced the year of construction, as lessons learned over time concrete elements joined together on the construction from major earthquakes have led to implementation of site by welding, or welding and grouting. Cast-in-place increasingly rigorous seismic design concepts. Following frames have elements that are constructed at the building the 1970s, enhanced engineering principles began to Existing cast-in-place reinforced site. As observed in the 1988 Spitak earthquake, precast be applied universally and improved the performance concrete frame buildings, Romania. RCF buildings’ weak connections led to building collapse of cast-in-place RCF buildings. Past earthquake damage Credit: WHE 2003. and other grave consequences. Precast RCF buildings shows that precast RCF buildings and the low-to-mid- pose some of the greatest risks and are much more code cast-in-place RCF buildings are very vulnerable to vulnerable than cast-in-place concrete frame buildings. earthquake shaking damage and collapse. Total collapse of prefabricated The main structural design deficiencies of RCF buildings Past earthquakes in the investigated region in which RCF reinforced concrete frame are weak or low-quality beam-column and frame-floor buildings faced damage, partial collapse, or total collapse buildings in foreground and minor connections, which make the frame structures especially include the following: the 1988 Spitak earthquake damage to large panel buildings in background, Gyumri, Armenia. vulnerable to earthquakes. In addition, insufficient (Armenia), in which nearly 70 percent of the precast RCF Credit: Peter Yanev. © EQE reinforcement and poor concrete confinement increase buildings collapsed and the remaining were demolished; Engineering. Reprinted with the vulnerability of these buildings. In precast frame and the 1940, 1977, 1986, and 1990 Vrancea earthquakes permission. buildings, those with grouted on-site connections (Romania, Bulgaria, Moldova). 12 DFW DUAL FRA M E -WA LL (D F W ) S YS T E M B U I L D I N G S Construction period of DFW in Relative Vulnerability countries studied: 1945-present Low Medium High Dual frame-wall systems have reinforced concrete frame directions to appropriately resist the seismic forces. If dual frame-wall and wall elements that work together to resist earthquake not carefully engineered and constructed, DFW buildings forces. Like other reinforced concrete structures, they can could suffer severe damage in earthquakes. be precast or cast-in-place; in some cases precast frames Past earthquakes in the investigated region in which DFW Dual frame-wall system, Kyrgyz are seen in combination with cast-in-place walls. As is Republic. the case for RCF buildings, precast concrete frame and system buildings experienced a range of damage levels, Credit: U. Begaliev, A. Duishev, wall elements are often more vulnerable to earthquake from no damage to total collapse, include the following: and R. Musakov, in OpenQuake, damage than are cast-in-place frame and wall elements. the 1977 Vrancea earthquake (Romania, Bulgaria, “Glossary for GEM Taxonomy,” Moldova), in which many buildings were damaged https://taxonomy.openquake.org/ The main seismic design deficiencies in this building and one collapsed completely; and the 1986 and 1990 terms/dual-frame-wall-system- category arise from weak connections and from Vrancea earthquakes, in which no damage to this building ldual licensed under CC BY 3.0. inadequate distribution of the frames and walls in the two class was observed. main perpendicular directions of the building. Frames and walls must be positioned carefully in two perpendicular Damage to cast-in-place frame building with reinforced concrete shear walls following the 1977 Vrancea earthquake, Bucharest, Romania. Credit: Georgescu and Pomonis 2012. LPB PREC A ST LA RG E P A N E L B U I L D I N G S ( L PB ) Construction period of LPB in countries Relative Vulnerability studied: 1950-1995 Low Medium High Large panel buildings are found in all of the investigated was reduced and buildings were damaged. In addition, precast large panel cities. Precast LPBs are built with precast reinforced the arrangement of interior walls in two perpendicular concrete panels that are assembled on the construction directions changes the behavior of LPBs when subjected site by steel welded connections, or grouted dowel and to earthquakes: LPBs with one long interior load-bearing welded connections. LPBs generally have floor-to-ceiling- wall are more vulnerable to earthquakes than those with height facade panels with additional steel reinforcement two or more load-bearing walls. As in URM buildings, around window openings. Included in the LPB category facade elements that are large panels may be especially is the uncommon typology of “precast boxes”—buildings dangerous to individuals outside the building during Large panel building construction built of prefabricated 3D components assembled on the an earthquake: if the panel connections are weak, the technique typical of Bratislava, construction site with welded or grouted connections. entire concrete panel could fall off the building due to Slovakia. Compared to other buildings in this study, LPBs have earthquake shaking. Credit: Stoychev 1976. faced much less damage from past earthquakes. Past earthquakes in the investigated region in which The main structural design deficiencies in large panel LPBs faced slight damage include the following: the buildings arise from the connections of the precast 1966 Tashkent earthquake (Uzbekistan), 1988 Spitak panels. Post-earthquake building surveys have shown earthquake (Armenia), 1985 Kayrakkum earthquake that some grouted connections did not have appropriate (Tajikistan), 1977 Vrancea earthquake (Romania, Bulgaria, quality control when being constructed; as a result, Moldova), and 1984 Gazli earthquake (Uzbekistan). Large panel building that sustained minor cracking, Gyumri, Armenia. the ability of the system to resist earthquake shaking Credit: Wyllie, L. A., and J. R. Filson, 1989. ©Earthquake Spectra, EERI. Reprinted with permission. 13 BLK BLOCK (BLK ) BU I LD I N G S Construction period of BLK in countries Relative Vulnerability studied: 1960-mid 1990s Low Medium High Block buildings (“крупноблочный” in Russian), or precast Past earthquakes in the investigated region in which block plane element structures, consist of reinforced concrete block buildings faced damage include the following: the precast elements assembled on the construction site by 1971 Kamchatka earthquake (Russian Federation); and welding or grouting connections. They are similar to LPBs the 1995 Neftegorsk earthquake (Russia), in which most except that their prefabricated elements are smaller; of the BLK buildings completely collapsed. instead of one large floor-to-ceiling precast facade panel, block buildings have four elements surrounding the window opening. Observation of building damage Block building in Baku, Azerbaijan, following past earthquakes indicates that this building 2019. typology is one of the most vulnerable among the types Credit: Maryia Markhvida, World considered in this study. Bank. The main structural design deficiencies in BLK buildings arise from the inadequate connections between the precast concrete elements. For these buildings (as for LPBs), earthquake shaking may affect weak or deteriorated facade element connections and cause portions of the building to fall on people outside the building. Damage to block building, Russia. Credit: WHE 2002a. SCS S LAB- C O LU M N SY STE M ( S C S ) B U I L D I N G S Construction period of SCS in countries Relative Vulnerability studied: 1957-1996 Low Medium High This group of buildings includes structural types in A past earthquake in the investigated region in which lift-slab which the steel reinforced concrete floors are directly SCS buildings experienced damage was the 1988 Spitak post-stressed connected to the columns. This typology is further earthquake (Armenia); in this event, lift slab SCS buildings classified as follows: some buildings have floors that are suffered severe damage to complete collapse. cast-in-place at the ground level and then lifted one by one to each floor level (lift slabs), and some buildings have steel reinforcement that is pulled in tension following the casting of the concrete on the construction site (post-stressed) to reduce cracking of the concrete Slab-column building under slab. Observations from past earthquakes and expected construction, Kyrgyz Republic. behavior of this typology suggest that the lift slab subclass Credit: WHE, n.d. (Report 39). may be more vulnerable than the post-stressing subclass. The main structural design deficiencies are in the connections between the concrete slabs and the columns, often with poorly executed construction joints. Damage to slab-column due to It has been found that the connections to the columns twisting at the top floors from the are often inadequately reinforced with steel, further 1988 Spitak earthquake, Gyumri, Armenia. weakening the joints at this location. Credit: Wyllie, L. A., and J. R. Filson, 1989. ©Earthquake Spectra, EERI. Reprinted with permission. 14 RC W R EINF O RC E D C O N C RE TE W A L L ( R C W ) S YS T E M Construction period of RCW in Relative Vulnerability countries studied: 1960-present Low Medium High This group includes buildings that have cast-in-place formwork are more vulnerable to earthquakes than those sliding concrete walls to provide earthquake resistance. RCW that use standard or tunnel formwork because buildings tunnel & standard buildings are further categorized by the kind of formwork employing sliding formwork typically have less steel used in construction—that is, the type of temporary reinforcement. These deficiencies are most commonly Reinforced concrete wall system molds into which concrete is poured to construct the noted in RCW buildings constructed before the 1990s. building under construction, Kyrgyz walls. Subtypes include buildings with standard, sliding, Republic. or tunnel formwork. Standard formwork is usually used A past earthquake in the investigated region in which Credit: Kanat Kanbolotov, in in RCW buildings that have varying geometries and cast-in-place RCW system buildings faced damage OpenQuake, “Glossary for GEM architectural plan layouts. Tunnel or sliding formwork is was the 1977 Vrancea earthquake (Romania, Bulgaria, Taxonomy,” https://taxonomy. used in buildings with established and repetitive designs, Moldova). This event caused superficial damage on the openquake.org/terms/wall-lwal most typically in buildings constructed between the mid- walls and cracks along the wall connections, and in more licensed under CC BY 3.0. 1960s and late 1990s. Observations of damage from severe cases concrete crushing in the shear walls. past earthquakes show that this building typology has performed better than others in this study. The main structural design deficiencies of the cast- in-place RCW system buildings are limited steel reinforcement in the concrete walls and poorly executed Damage to reinforced concrete construction joints. In addition, the very slender walls wall system building after 1977 in these buildings are susceptible to damage and failure Vrancea earthquake, Romania. during earthquake shaking. RCW buildings that use sliding Credit: WHE 2002b. SUMMARY OF MULTIFAMILY BUILDING TYPES historical design codes and standards, or a combination of both. See Overall Findings for more details. • Of the seven building types described, those with the highest relative risk from earthquake shaking include figure �. relative vulnerability of unreinforced masonry buildings (URM), precast frame multifamily building types in study O f t h e b u i l d i n g t y p e s st u d i e d , (RCF) buildings, block buildings (BLK), and slab-column Low Medium High u n r e i n fo r ce d ma s o n r y (U R M ) system buildings with lift slab construction (SCS). Overall, UR M unconfined b u i l d i n g s , p r e ca st co n cr ete f r a me building vulnerabilities arise from low construction confined (RCF ) b u i l d i n g s , b l o ck (BL K) b u i l d i n g s , quality associated with construction period, weak RC F precast a n d s l a b - co l u mn sy ste m (S C S ) figure �. building construction period cast-in-place b u i l d i n g s wi t h l i f t s l a b co n st r u ct i o n of multifamily building types in study h a v e t h e h i g h e st r e l a t i v e r i s k . DFW dual frame-wall unreinforced masonry (URM) reinforced concrete frame (RCF) LPB precast large panel dual frame-wall (DFW) system precast large panel (LPB) BLK block block (BLK) lift-slab slab-column system (SCS) SCS post-stressed reinforced concrete wall system (RCW) sliding RC W 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020 tunnel & standard present 15 M ETHOD OLO G Y The probabilistic seismic risk analysis of multifamily •  The development of the hazard model is intended only •  The exposure data are based on the latest census buildings in the 27 cities required development of to provide risk results at the regional scale and not at a statistics available for the countries and cities. In most hazard, exposure, and vulnerability models (see Figure single-building level. cases, the census data used are from 2011 and 2014. 6; key definitions are in Box 1). The following describes EXPOSURE MODEL •  Where construction costs for a city are unavailable, the process of addressing each of these components data from nearby cities are used. to output quantifiable information on the building and The building exposure model was developed through The reliability of exposure data varies across building human impacts from earthquakes in the ECA region. reviewing and collecting country-specific, cross-country, types and city to city, based on the available information; figure 6. models used in probabilistic and structural typology–specific information sources. For see the exposure data reliability matrix (Table 2) and City seismic risk analysis each building typology, many data sets were evaluated to Profile Sources. It should be noted that the exposure understand the number of buildings, number of dwellings, reliability index is relevant for high-level regional risk hazard model and number of occupants, as well as the associated analysis, whereas even the most highly reliable data are construction cost (exposure value). The sources used not sufficient or appropriate for building-specific analysis. probabilistic include censuses and national statistical institutions, exposure model seismic risk analysis national energy efficiency programs, building inventory VULNERABILITY MODEL vulnerability model reports, past projects on risk and exposure, international building databases, post-earthquake reports, technical The vulnerability models used in the risk analysis were publications, and satellite images. In most cases, the obtained by combining different loss models from number of buildings was deduced or projected from various sources, since models for specific building types SEISMIC HAZARD MODEL combined information on the number of dwellings in within each city or country were not readily available. The earthquake model consists of three regional models multifamily buildings, number of dwellings according to The types of consequences that were quantified in this and modifications based on two local seismic hazard the year of construction, buildings according to number study are direct financial losses from building damage, models for the seismic hazard computations in this of stories, damage to buildings after earthquakes, permanent relocation of occupants, and fatalities due to study. The regional models are obtained from the Global buildings renovated as part of ongoing energy efficiency building damage. Over 400 models using both empirical Earthquake Model GEM2018 global mosaic of seismic programs, statistics on the tallest buildings in the and analytical methods were reviewed for direct financial hazard models. They include the Euro-Mediterranean cities, and examples of past building inventories (Mott losses from building damage, and expert judgment Seismic Hazard Model 2013 (ESHM13) developed within MacDonald 2020a). was used to develop damage and loss models for the the SHARE Project (Woessner et al. 2015), the Earthquake various building typologies under consideration. For Model Middle East 2014 (EMME14), and the Earthquake Key assumptions and considerations for the exposure each building type, the vulnerability model considers the Model Central Asia 2015 (EMCA15). Two local seismic component are as follows: earthquake load resisting system, and is further classified hazard models used in the hazard component are the •  Exposure data and its reliability vary across cities. by vulnerabilities of connections, infills, facade elements, BIGSEES15 model, which includes the latest Romanian Where required, deduction and expert judgment were and cladding. The reliability of vulnerability data varies national seismic hazard model (Pavel et al. 2016) and used to complete the exposure information (see Table 2). across building types and city to city, based on the the BUL18 model for Bulgarian cities (Mott MacDonald available information; see the vulnerability data reliability •  Exposure data was aggregated at a single point in a city, matrix (Table 3). Vulnerability models produced from this 2020a). or in some cases, across several points within a city. approach also provide the basis for the relocation period, The assumptions and considerations for the hazard •  Asset values of the buildings under consideration are depending on the time needed to repair or replace the component are as follows: the replacement value, or construction cost, of each building. The HAZUS methodology developed in the USA •  Only strong ground motion is considered. Estimates building typology within each city. The replacement was used for repair time and fatality modeling (DHS, of secondary seismic hazards—such as liquefaction, value does not include removal of debris, design, quality FEMA n.d.). landslides, subsidence, or slope instability—are not assurance, project approval costs, or price increase For many of the buildings, the structural elements provided. surge due to high reconstruction demand following an and their connections will likely have different levels earthquake. •  National seismic hazard models are used only where of strength degradation throughout their lifetime. In regional models are inconsistent with detailed national •  Multifamily buildings are defined as residential buildings addition, some of the buildings may have undergone hazard models. Seismic hazard was calculated only at with more than two dwellings and three or more floors. It structural modifications—by extending and closing several locations (sometimes one location) throughout is assumed that the population within these buildings is balconies, cutting openings, or even removing walls to the city. proportional to the number of dwellings. modify the interior or to add entrances at the ground 16 box 1. terminology floor. Although statistical data on such post-construction structural changes are not available, these considerations Hazard: Hazard refers to the physical forces produced by a peril, such as ground shaking are indirectly included in the uncertainties within the induced by an earthquake. In this study, the hazard is strictly that of earthquake vulnerability models (Mott MacDonald 2020a). shaking and does not include soil liquefaction, subsidence, or slope instability. The assumptions and considerations for the vulnerability Exposure: Exposure refers to the location, characteristics, and value of assets such model are as follows: as people, buildings, critical facilities, and transport networks located in an area that •  The model for casualties is specific to indoor fatalities may be subject to a hazard event. In this study, only multifamily residential buildings and does not include outdoor fatalities caused by falling are considered and the value of the building is determined through the present day building elements. construction cost considering similar buildings today (i.e., building replacement value). Vulnerability: Vulnerability is the susceptibility of people and assets to the forces of a •  The fatalities are assessed at nighttime (2:00 a.m.) on a hazard event. For example, a building’s earthquake vulnerability depends on a variety weekend. It is assumed that 100 percent of the occupants of factors, including material and structural system, quality and year of construction, are inside their homes, which is considered a worst-case and height. Human earthquake vulnerability depends on whether it is a weekend or scenario in the vulnerability models. weekday when an earthquake occurs, and whether people are more or less likely to be home at different times in the day. For example, a nighttime earthquake has a PROBABILISTIC SEISMIC RISK ANALYSIS high likelihood of all residents being inside of their homes, whereas an afternoon The risk analysis combining the hazard, exposure, and earthquake may have fewer residents at home. vulnerability component was performed with open Risk: Disaster risk is a function of hazard, exposure, and vulnerability. Several risk source software OpenQuake,4 and the results are metrics are calculated in this study: average annual loss (AAL) and average annual presented in terms of damage and losses per building loss ratio (AAL%), as well as financial loss and number of lives lost at a specified return type and impacts on occupants. • period. The city risk profiles provide annualized loss values to compare between cities and regions, as well as risk information for earthquake impacts with a 100-year return table 2. exposure data reliability matrix period (i.e., this impact or greater impact occurs on average once in 100 years) and a Dushanbe Bucharest Podgorica Shymkent Bratislava Budapest Ashgabat Ljubljana Tashkent Belgrade Chișinău 475-year return period. Yerevan Bishkek Gyumri Plovdiv Mostar Almaty Zagreb Durrës Skopje Tirana Rijeka Tbilisi Sofia Baku Osh Average annual loss: AAL is the total loss when averaged over a very large number Iași URM of years. In this study, the AAL is provided in absolute monetary terms (euros) and RCF DFW is strictly expected loss as a result of multifamily building damage from earthquake LPB ground shaking. BLK SCS Average annual loss ratio: AAL% is the ratio of expected average annual loss RCW High-reliability data are found in publications, based on Low-reliability data reflect the number of buildings to buildings’ replacement value. The building replacement value is the cost of manual counting, or otherwise precisely derived. projected with other assumptions. constructing a new building in a given building typology. When the loss is normalized Medium-reliability data reflect the number of buildings Dark-colored cells indicate building types that were not over the replacement cost, it provides a useful loss-to-value metric on the relative projected from the national census data through present in the city under investigation. optimization processes. risk across cities and regions. Note that a higher AAL% indicates a more risk-prone portfolio of buildings. table �. vulnerability data reliability matrix Average Annual Loss (AAL,€) Average Annual Loss Ratio (AAL%) = Dushanbe Bucharest Podgorica Shymkent Bratislava Budapest Ashgabat Ljubljana Tashkent Belgrade Chișinău Yerevan Building Replacement Value (€) Bishkek Gyumri Plovdiv Mostar Almaty Zagreb Durrës Skopje Tirana Rijeka Tbilisi Sofia Baku Osh Iași Return period: The return period of an earthquake impact metric (e.g., loss) indicates URM RCF that on average that metric is exceeded once during that period. For example, a 100- DFW year loss indicates a loss that on average is exceeded once in 100 years. It is very LPB BLK important to understand that a loss with a return period of 100 years could occur in SCS any given year, irrespective of when that impact last occurred. The greater the return RCW High-reliability data have good correspondence with Low-reliability data indicate data that are not found period, the more intense and less frequent the event’s occurrence. This study presents respect to the building type and the country. for the specific building type or country; in this case, both 100-year return period losses and 475-year return period losses. models are developed based on expected vulnerability and employ expert engineering judgment. Probabilistic Seismic Risk Analysis: A probabilistic analysis considers the consequences Medium-reliability data have good correspondence Dark-colored cells indicate building types that were not of all possible damaging earthquakes affecting the region, considering their probability with respect to the building type, but not with respect present in the city under investigation. to the country; in this case, models developed for of occurrence. different countries are combined and applied. 4 OpenQuake was developed by the Global Earthquake Model (GEM); see the OpenQuake web page at https://www.globalquakemodel.org/openquake. 17 table �. summary of earthquake risk assessment results Total Percent Top 2 Total Percent of city Average Average OVER ALL FI N D I N G S population of city high-risk population in population in annual annual in pre-2000 population building top 2 high-risk top 2 high-risk loss total loss ratio Although many uncertainties are associated with the City, Country multifamily in pre-2000 types* building types* building types* (million (AAL%)* housing multifamily euros)** earthquake risk analysis conducted in this study (including housing uncertainty in the composition of the building stock), the most obvious trend is the influence of the level of Tirana, Albania 164,900 34% RCF, URM 47,000 10% € 5.25 0.39% seismicity in the investigated regions in combination with Durrës, Albania 30,510 27% RCF, URM 7,000 6% € 1.28 0.37% the vulnerabilities of the buildings studied. This section provides results and trends based on earthquake losses Mostar, Bosnia & 44,070 39% URM, SCS 20,000 18% € 5.27 0.61% and commentary on pre-2000 multifamily buildings in Herzegovina the 27 cities studied. Sofia, Bulgaria 881,820 71% RCF, URM 260,000 21% € 41.10 0.26% Table 4 presents findings that include city-specific Plovdiv, Bulgaria 239,430 69% RCF, URM 65,000 20% € 6.55 0.21% population data of pre-2000 multifamily housing, as well Zagreb, Croatia 418,080 52% URM, RCF 280,000 35% € 48.80 0.44% as absolute and relative losses from the probabilistic Rijeka, Croatia 64,260 54% URM, RCF 38,000 32% € 4.75 0.40% seismic risk analysis. The highlighted cells indicate the highest three values (cities) for the result metric, where Budapest, Hungary 1,191,360 68% URM, RCF 440,000 25% € 48.80 0.13% the dark-colored cell is the largest value and the light- Chişinău, Moldova 555,660 81% URM, LPB 324,000 47% € 6.26 0.08% colored cell is the lowest. Podgorica, 49,830 33% RCF, URM 14,000 9% € 2.90 0.26% Montenegro Skopje, North 158,340 29% SCS, RCF 38,000 7% € 5.78 0.29% Fo r th e 27 c i ti es i n t h i s st u dy : Macedonia Bucharest, Romania 1,626,920 89% RCF, URM 1,100,000 61% € 192.00 0.42%  Total p o p u lat io n i n p re- 2 0 0 0 m u l ti fa m i l y Iași, Romania 205,900 71% URM, RCF 154,000 53% € 5.41 0.31% housin g is ap p rox i m ate l y 13.9 mi l l i on . Belgrade, Serbia 1,248,380 74% RCF, URM 550,000 33% € 33.60 0.10% Bratislava, Slovakia 271,360 64% RCF, URM 30,000 7% € 8.48 0.06%  On ave rage , h a lf of the popul ati on l i ves Ljubljana, Slovenia 187,600 67% RCF, URM 78,000 28% € 17.90 0.32% in p re -2000 mu lt i fa m i l y h o u s i n g . Almaty, Kazakhstan 670,680 36% BLK, SCS 13,000 1% € 31.50 0.38% Shymkent, 242,320 26% BLK, URM 100,000 11% € 3.04 0.19%  Total m u lt ifamily b u i l d i n g ex p o s u re va l u e Kazakhstan is ne arly € 238 bi l l i on . Bishkek, Kyrgyz 384,940 38% URM, RCF 143,000 14% € 4.28 0.20% Republic  Total avera ge an nu a l l o s s ** f ro m e a r t h q u a kes Osh, Kyrgyz Republic 113,080 44% URM, RCF 105,000 41% € 1.27 0.47% is ne arly € 680 mi l l i on . Dushanbe, Tajikistan 456,840 54% SCS, URM 67,000 8% € 3.21 0.39% Ashgabat, 322,920 39% URM, RCF 115,000 14% € 5.45 0.20%  Top t wo b u ild in g t y p es co nt r i b u ti n g to Turkmenistan tota l average a n n u a l l o s s a re Tashkent, Uzbekistan 1,543,800 62% SCS, URM 550,000 22% € 13.00 0.15% un reinforc ed masonr y (URM) a n d Yerevan, Armenia 776,030 71% URM, SCS 250,000 23% € 72.50 0.67% preca st la rge pane l (LPB) . Gyumri, Armenia 45,630 39% URM, RCF 43,000 37% € 8.17 0.70% Baku, Azerbaijan 1,110,240 48% BLK, URM 125,000 5% € 42.40 0.41%  Top t wo h igh -risk * b u i l d i n g t y p es a re Tbilisi, Georgia 883,140 82% BLK, SCS 300,000 28% € 59.50 0.89% un reinforc ed masonr y (URM) a n d *High risk building types based on earthquake loss are determined by the expected average annual loss ratio, or average annual loss as a percentage of re inforc ed co ncrete frame (RC F) . replacement value (AAL%). The replacement value includes only the multifamily buildings assessed in this study. See Box 1 for more information. **Average annual loss (AAL) is the total loss when averaged over a very large number of years, considering current building stock. In this study, the AAL 18 is provided in absolute monetary terms (euros) and is strictly expected loss as a result of multifamily building damage from earthquake ground shaking. RECOMME RES ULTS SNNDAAP SH TI OS ON T FO R FU R T HE R MULTIFAMILY BUILDING VULNERABILITIES The study is able to quantify the relative vulnerability of the seven typologies under consideration. The most Po p u latio n i n pr e-2000 mul ti fami l y b uild ing s vulnerable typology is unreinforced masonry buildings (URM), especially the unconfined masonry structures. During and after an earthquake, bricks can fall out of such buildings, leading to irreparable damage and even total BU C H A RE ST collapse. In addition, and as verified by the 1988 Spitak has the most inhabitants living in pre-2000 multifamily housing, followed by Tashkent and earthquake in Armenia and other countries, reinforced Belgrade. Nearly 90% of the total city population in Bucharest lives in these multifamily concrete frame buildings (RCF) contribute to much of buildings. annual loss in absolute terms, and in relative terms as a share of building value. Slab-column system buildings (SCS)—especially those in the lift-slab subcategory, which increase in vulnerability with earthquake intensity—can experience total collapse. Large panel buildings (LPB) are expected to be less likely to collapse than other types of buildings designed and constructed before 2000. Po p u latio n i n hi gh-r i sk mul ti fami l y b uild ing s However, the final losses and the investment needed to repair LPBs after an earthquake may be considerable. Similarly, the anticipated damage to reinforced concrete BU C H A RE ST ! wall system buildings (RCW) will occur in localized areas has the greatest number of its residents in its high-risk reinforced concrete frame and of the structural system, and these buildings are not unreinforced masonry buildings. High-risk building typologies are considered the top two expected to collapse completely. For SCS buildings, the building types in each city with the highest average annual loss ratios* from building damage. financial ramifications of repair may be substantial, just Similar to the previous findings, Belgrade and Tashkent follow Bucharest with the most as they are for LPBs. number of residents housed within their high-risk building types. Overall, unreinforced masonry buildings (URM) are expected to experience the most damage, contributing to a significant portion of the direct financial losses, number of fatalities, and number of people who will be permanently displaced after an earthquake. This remains true even in regions with lower seismicity; where A bso lu te an d r el ati v e l oss other building types are expected to experience light to moderate damage, the most severely damaged and collapsed buildings are of the masonry type. • € BU C H A RE ST is also expected to have the highest total losses from future earthquakes, with almost €200 million in average annual losses.** This is followed by Yerevan and Tbilisi, which are expected to have average annual losses of nearly €73 million and €60 million, respectively. O v e r a l l , u n r e i n fo r ce d ma s o n r y b u i l d i n g s (U R M ) a r e ex p e cte d TBI LI SI to ex p e r i e n ce t h e mo st d a ma g e , % is the city with the highest loss-to-value ratio when comparing the ratio of total euro loss co nt r i b u t i n g to a s i g n if ica nt (average annual loss) to the replacement value of all buildings within a city.* This is followed p o r t i o n o f t h e d i r e ct f i n a n cia l by Gyumri and Yerevan. While Bucharest has the highest total euro loss value, it has less than half of the relative losses (average annual loss ratio) as Tbilisi, indicating that Bucharest has a l o s s e s , n u mb e r o f fa ta l i t i e s , a n d larger replacement value of multifamily buildings than does Tbilisi. n u mb e r o f p e o p l e wh o wi l l b e p e r ma n e nt l y d i s p l a ce d a f te r a n e a r t h q u a ke . 19 RECOMME N D A TI O N S Considering all assumptions and limitations of this study (see Methodology), the following recommendations are proposed for consideration by governments, authorities, policy and decision makers, engineers, researchers, and analysts, among other possible stakeholders and beneficiaries of this study to further earthquake risk evaluation and reduction. Prepare databases with information relevant (but disaster risk reduction measures. This is a vital step seismic evaluation and retrofit of multifamily buildings. not restricted) to earthquake risk assessment on a for the safety of residents of multifamily buildings and In addition, the capacity of building departments should large scale. While socioeconomic parameters are often communities. Such strategies must include stakeholder be increased to provide improved oversight to existing well addressed by country-based censuses, adequate engagement and suitable government policies and buildings when they are assessed, renovated, or retrofit. information about the buildings—their structures, procedures with regulatory measures. In addition, structural deficiencies, and current conditions—is sparse this step should address investments, incentives, and Increase public awareness about earthquake risk. The and should be collected. It may be possible to collect contingencies; community awareness and engagement; risk awareness and engagement component should not building data in parallel to acquiring census statistics, and the creation of preparedness, response, and recovery be ignored—it is one of the most important aspects to with the goal of synchronous sharing of institutional plans and business operation plans. It is critical to consider in order to create a resilient society. To facilitate resources and capacity. The proposed database should be recognize the most vulnerable members of society when government policies and commitment to seismic safety in updated periodically to provide information to facilitate developing seismic risk reduction plans and strategies, countries and communities, it is imperative to continually any mitigation, response, or recovery efforts. as these groups are often disproportionately affected by educate the public about their earthquake risk and disasters and recover last, if at all. vulnerability. Proper modes of risk communication should Collect data on buildings’ current condition and level be initiated, and resources for earthquake preparedness of maintenance for earthquake retrofit. Such data Integrate earthquake risk reduction measures into should be provided at the very minimum. are generally unavailable in all cities studied. In some ongoing investment programs. In many of the studied countries, rapid visual assessment of residential buildings countries, energy efficiency programs and other funding Design and promote adoption of risk financing and risk has been initiated by programs related to living conditions programs for improvement of living conditions are transfer strategies by owners and residents. Awareness and not necessarily to address the seismic vulnerability currently under way. Such programs could be extended of and opportunities for risk financing and transfer of the structures. An understanding of the amount of to include seismic risk reduction measures and to provide mechanisms should be made available to residents, structural change and building degradation could greatly a range of possible retrofit solutions to owners and public homeowners, and public agencies. Earthquake insurance, increase understanding of building typologies’ overall oversight agencies. Since such programs often focus on financial incentives through premium subsidies and tax vulnerability. The collection of such data would require multifamily residential buildings, they offer a fitting breaks, as well as funding to retrofit unsafe building types the preparation of nationally approved guidelines opportunity to bridge energy efficiency with structural are some examples of financial instruments to reduce the and templates for on-site assessment and consistent safety, given proper planning and foresight. risk that inhabitants and owners of multifamily housing application. The following is a generalized four-step face. These mechanisms can provide compensation process for collecting data and systematically addressing Update current assessment codes and legislation for for property damage and indirect losses from damage, earthquake risk reduction of multifamily structures: existing buildings. It is likely that seismic building codes distribute the risk and burden of recovery between for existing structures as well as legislation will need public and private sectors, and encourage investment in 1. Building upon the risk analysis in this study, conduct to be updated to better address issues related to the earthquake risk reduction. • a more detailed risk assessment of multifamily building types across a country. 2. Initiate a rapid visual assessment program to identify critical structural design deficiencies in key building types, starting with the most vulnerable buildings found under step 1. 3. Prioritize and develop earthquake risk reduction intervention plans and develop design guidelines for standardized retrofit solutions. 4. Undertake detailed asset-level engineering evaluations to refine the understanding of structural design deficiencies to design building-specific retrofits. Image Credit: “Nalband Village.” Children make their way home from school, in the Develop an appropriate earthquake risk reduction epicenter of the 1988 earthquake [Spitak, strategy and prioritization plan for investment in pre- Armenia]. © Monica Hovhannesyan. 20 I M PACT OF TH I S STU D Y The strength of this study is the comprehensive approach through which data were collected, analyzed, and verified in order to conduct the earthquake risk assessment. A thorough research and review process was undertaken to characterize the multifamily residential building stock. This included investigating buildings’ structural deficiencies, identifying specific features of the urban territories, and determining how the building stock has been transformed through the years. For each of the studied cities, the process has led to a valuable database that can be utilized in future projects—from additional high-level studies to detailed seismic risk analyses for a single building type. In addition, data uncertainties and limitations are well documented and may inform future research to bolster any subsequent seismic risk assessments. Detailed information on the engineering methodology and earthquake risk results of multifamily housing in the ECA can be found in the technical engineering reports listed in the Works Cited section (Mott MacDonald 2020a, 2020b). • T h e stre n g th of thi s study i s th e co m prehensi ve appr oach th ro u g h whi ch data wer e co lle c te d , anal y zed, and ve rif ie d in or der to conduct the Image Credit: “ The Catholic Hospital in Plovdiv [Bulgaria, 1928], affected e a rth qu ake r i sk assessment. by the Chirpan earthquake” by unknown author. 21 C I TY EA RTHQ UA K E R I S K P ROFIL ES 22 UND ERS TA N D I N G TH E C I TY E A R T H Q U AKE R I S K PR O F I L E S The earthquake risk profiles for the ECA countries building-specific exposure and vulnerability data. The figure �. using the city profiles: quick facts presented here indicate the damage and loss values on diagram also shows the share of each building category as a per-city basis. Profiles are organized by region then by a percentage of total pre-2000 buildings in the city. Note •  All loss values are in present-day values (2019 €). country and city. Note that some countries have only one that not all building types are found in every city, and the city that has been studied. The elements included in the diagram reflects this with a null value. Highlighted in the •  Historic earthquake loss values are in 2019 €.* profile are explained below. Data Reliability and Building Exposure diagram are the •  The most current census data are listed. city’s two highest-risk building types. City and country profiles: The profile includes a Country * Zagreb loss values in 2020 € for March 2020 Snapshot with the 2018 gross domestic product (GDP) and Detailed losses: The Building Damage Losses section Zagreb earthquake. the most current information on the country’s population on the second page of each profile begins with a short count.5 In addition to relevant earthquake history of the description of the building types and specifies their city in the Summary section, the Earthquake Impacts relative vulnerability under Building Classification events—with the potential to far exceed the losses for Snapshot provides the expected average annual loss and and Vulnerability Range. All vulnerability information the 475-year return period—could take place within annual affected GDP from future earthquakes. The annual presented is approximate and depends on post- the city. The loss information includes damage from affected GDP is the percentage of the average loss over construction modifications, degradation, construction the event (in euros), the ratio of loss to total building the country’s current GDP, averaged over many years. quality, and year of construction. See Types of value in the city, and the number of fatalities expected Furthermore, information is provided on the number Multifamily Buildings in ECA for additional information. from an event occurring at 2:00 a.m. on a weekend. of people living within the city’s multifamily residential The Absolute and Relative Loss graph disaggregates Also presented is the percentage of the damage cost building stock. As a comparative metric for all regions in expected average annual losses per building type in the per capita (considering residents of these buildings) this study, the expected average annual loss ratio (AAL%) city. The graph compares the absolute average annual as a percentage of the country’s gross national income is included, with averages for all cities shown as horizontal loss in 2019 euros for each building type. This metric per capita (GNIPC) value.6 The gross national income bars, and regional averages indicated by vertical lines to considers both the vulnerability of the building type and measures annual income per number of people within understand the relative risk of the multifamily building how widespread that type is. The graph also provides the the country and is a comparable measurement of income portfolio. This is a metric for comparing loss-to-value average annual loss ratio, which is the ratio of the average across countries of different population sizes and living ratios, which can provide information for prioritizing annual loss to the exposure value (i.e., the replacement standards. If all costs of damage, and consequently seismic strengthening interventions. It is important to value) of that particular building type. This metric offers a repair, are borne by multifamily residential building understand that the expected average annual loss does relative risk comparison for all vulnerable buildings in the inhabitants, the ratio of expected loss to annual income not represent the impact of a single event, but an average city, where the higher ratios indicate building typologies serves to explain the degree of loss per resident.7 A high of all possible earthquake events. Thus, much larger considered higher risk-to-value for the city. The two loss-to-income ratio can create challenges to recovery if impacts could occur in the event of a rarer and more highest-risk building types that are found in the city and mitigation and strengthening measures are not planned intense earthquake. See Box 1 for more information on the associated consequences are listed below the graph. for and implemented. the terminology used in the profiles. The Earthquake Event–Based Losses section summarizes All expected annual and event-based loss values are Reliability and exposure: Every city profile contains a the losses expected from an event with a 100-year subject to limitation by the reliability of the input data on Data Reliability and Building Exposure diagram to visually return period and an event with a 475-year return hazard, exposure, and vulnerability. See the Methodology communicate the overall confidence level for city- and period. Note that other more hazardous earthquake section for more information. • T h e ex p e cte d a v e r a g e a n n u a l lo s s d o e s n o t r e p r e s e nt t h e 5 GDP data are from the World Bank Indicators Database, imp a ct o f a s i n g l e ev e nt , b u t https://data.worldbank.org/indicator. an a v e r a g e o f a l l p o s s i b l e 6 Data are for 2018 and are from the World Bank Indicators Database, https://data.worldbank.org/indicator. e a r t h q u a ke ev e nt s . Th u s , mu ch 7 The ratio explains how costs borne for repair after an earthquake can la r g e r i mp a ct s co u l d o ccu r i n impact annual savings for a resident. Damage (€) th e ev e nt o f a r a r e r a n d mo r e # of Multifamily Image Credit (left): Image from above of buildings collapsed following inte n s e e a r t h q u a ke . Building Residents the [2019 Albania] earthquake. “Terremoto Albania” by Dipartimento Gross National Income per Capita Protezione Civile, licensed under CC BY 2.0 / Added color transparency. 23 Tirana ALBANIA Summ ary Tirana is the capital and largest city in Albania, home to Nearly 50,000 people, or about 10 percent of the city’s approximately 17 percent of the total population. Albania population, reside in high-risk building types. has experienced many earthquakes of varying severity in the past. The most deadly was a series of earthquakes Compared to the data for all the buildings in this study, in 1851 (highest magnitude 6.6) near the towns of Vlorë the data used to conduct the risk analysis for buildings and Berat, causing the loss of 600 lives. Another large in Tirana average a low reliability; this rating is a function event was the 1967 magnitude 6.5 earthquake that led of the assumptions, uncertainties, and data available for to over 500 collapsed homes, 170 injuries, and nearly 20 the city (see Methodology). The diagram below indicates Tirana fatalities. More recently, in 2019, Albania was struck by a the reliability and confidence related to building stock series of damaging earthquakes, the strongest and most data (i.e., exposure data) as well as data on building deadly of which was a 6.4 magnitude event near the town vulnerability. These confidence and reliability levels of Mamurras in November of that year. This event caused should be considered when interpreting and applying the over 50 deaths, injured 2,000 people, and resulted in €1 risk metrics presented in the city profile. billion (2019 €) in damage. Approximately 1,500 buildings *At risk building types based on earthquake loss are were damaged in the capital city. determined by the expected average annual loss ratio, Based on the analysis, nearly one in every three people or average annual loss as a percentage of replacement of Tirana resides in pre-2000 multifamily residential value (AAL%). The expected loss results are predicted buildings. The two building types that are the most from models of future events subject to variable data at risk* are reinforced concrete frame (RCF) buildings availability and modeling assumptions. The replacement TIRANA FACTS and unreinforced masonry (URM) buildings. Along with value is the construction cost of all multifamily buildings in the city. Capital City Population: 485,000 (2019) GNI/capita: ~€4,000 reinforced concrete wall (RCW) buildings, these building types are also expected to cause the most fatalities. Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s € % RELIABILITY High EXPOSURE URM RCF LPB RCW Total building Building damage, Number of Medium 37% 11% 44% 8% damage % of total value fatalities Low DFW BLK SCS 100-year € 146 M 11% 360 0% 0% 0% 22% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 508 M 37% 2,400 Medium Return Period 75% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 24 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 12.8 Billion Euros* 1851 Vlorë / Berat M6.6 POPULATION 1967 Dibër M6.5 2.87 Million People* 2019 Mamurras M6.4 * 2018 estimate Buildin g D am age Lo s s es Ti rana BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €1.80 0.90% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €1.50 0.75% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €1.20 0.60% frames that provide earthquake resistance; RC F is € 5 . 2 5 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €0.90 0.45%  This amounts to 0.04% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €0.60 0.30%  Near ly 3 4 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.30 0.15% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that  Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Ti r ana Bucharest Iași Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Tirana. Bratislava Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~4 7 , 0 0 0 peop l e RC F 60% 70% 71% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 10% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Ti ran a URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 25 Durrës ALBANIA Summ ary Durrës is a major port city and is the second largest city types are also expected to cause the most fatalities. in Albania, home to approximately 4 percent of the total Approximately 7,000 people, or about 6 percent of the population. Albania has experienced many earthquakes city’s population, reside in high-risk building types. of varying severity in the past. The most deadly was a series of earthquakes in 1851 (highest magnitude 6.6) Compared to the data for all the buildings in this study, near the towns of Vlorë and Berat, causing the loss of the data used to conduct the risk analysis for buildings in 600 lives. Another large event was the 1967 magnitude Durrës average a low reliability; this rating is a function 6.5 earthquake that led to over 500 collapsed homes, of the assumptions, uncertainties, and data available for the city (see Methodology). The diagram below indicates Durrës 170 injuries, and nearly 20 fatalities. More recently, in 2019, Albania was struck by a series of damaging the reliability and confidence related to building stock earthquakes, the strongest and most deadly of which data (i.e., exposure data) as well as data on building was a 6.4 magnitude event near the town of Mamurras in vulnerability. These confidence and reliability levels November of that year. This event caused over 50 deaths, should be considered when interpreting and applying the injured 2,000 people, and resulted in €1 billion (2019 €) risk metrics presented in the city profile. in damage. In Durrës, at least three hotels, a residential *At risk building types based on earthquake loss are villa, and an apartment building collapsed. determined by the expected average annual loss ratio, Based on the analysis, about 27 percent of the or average annual loss as a percentage of replacement population of Durrës resides in pre-2000 multifamily value (AAL%). The expected loss results are predicted residential buildings. The buildings that are the most from models of future events subject to variable data D URRËS FACTS at risk* are reinforced concrete frame (RCF) buildings availability and modeling assumptions. The replacement value is the construction cost of all multifamily buildings Second Largest City Population: 113,000 (2011) GNI/capita: ~€4,000 and unreinforced masonry (URM) buildings. Along with reinforced concrete wall (RCW) buildings, these building in the city. Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s € % RELIABILITY High EXPOSURE URM RCF LPB RCW Total building Building damage, Number of Medium 29% 11% 54% 6% damage % of total value fatalities Low DFW BLK SCS 100-year € 35 M 10% 70 0% 0% 0% 28% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 125 M 36% 400 Medium Return Period 99% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 26 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 12.8 Billion Euros* 1851 Vlorë / Berat M6.6 POPULATION 1967 Dibër M6.5 2.87 Million People* 2019 Mamurras M6.4 * 2018 estimate Buildin g D am age Lo s s es Du rrë s BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €0.45 1.2% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €0.38 1.0% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €0.30 0.8% frames that provide earthquake resistance; RC F is € 1 . 2 8 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €0.23 0.6%  This amounts to 0.01% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €0.15 0.4%  Near ly 2 7 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.08 0.2% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type  Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje H igh -R is k Bu i l d i n gs i n D u r r ë s Bucharest Iași Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Durrës. Bratislava Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~7 , 0 0 0 peo pl e l i ve RC F 50% 70% 58% Osh Dushanbe in these bui l di ngs , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 6% o f the to ta l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Durrës URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 27 Mostar BOSNIA & HERZEGOVINA Summ ary Mostar is the fifth largest city in Bosnia and Herzegovina Compared to the data for all the buildings in this study, and home to about 3 percent of the country’s population. the data used to conduct the risk analysis for buildings in It is notable as the location of one of the country’s most Mostar average a high reliability; this rating is a function visited landmarks—the Stari Most, or “Old Bridge,” built of the assumptions, uncertainties, and data available for in the 16th century. Bosnia and Herzegovina has been the city (see Methodology). The diagram below indicates impacted by earthquakes in the past, such as the 1943 the reliability and confidence related to building stock event that led to 19 fatalities. The Banja Luka earthquake data (i.e., exposure data) as well as data on building of 1969 was a strong magnitude 6.1 event that killed 15 vulnerability. These confidence and reliability levels people, left over 1,000 people injured, severely damaged should be considered when interpreting and applying the or completely destroyed 86,000 apartments, and resulted risk metrics presented in the city profile. in more than €50 million (2019 €) in damage. Mostar *At risk building types based on earthquake loss are Based on the analysis, nearly 40 percent of the population determined by the expected average annual loss ratio, of Mostar resides in pre-2000 multifamily residential or average annual loss as a percentage of replacement buildings. The buildings that are the most at risk* are value (AAL%). The expected loss results are predicted unreinforced masonry (URM) and slab-column system from models of future events subject to variable data (SCS) buildings. Along with reinforced concrete frame availability and modeling assumptions. The replacement (RCF) buildings, these building types are also expected value is the construction cost of all multifamily buildings to cause the most fatalities. Approximately 20,000 in the city. inhabitants, or about one in every five people in Mostar, MOSTAR FACTS reside in high-risk building types. Fifth Largest City Population: 113,000 (2013) GNI/capita: ~€5,000 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High URM RCF LPB SCS RCW € % RELIABILITY EXPOSURE 67% 9% 16% <1% 8% Total building Building damage, Number of Medium damage % of total value fatalities Low DFW BLK 100-year € 120 M 14% 200 0% 0% 55% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 380 M 44% 880 Medium Return Period 175% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 28 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 17.0 Billion Euros* 1962 Sarajevo M6.3 POPULATION 1969 Banja Luka M6.1 3.32 Million People* 2015 Tuzla M3.6 * 2018 estimate Buildin g D am age Lo s s es M o star BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €4.2 1.2% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €3.5 1.0% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €2.8 0.8% frames that provide earthquake resistance; RC F is € 5 . 2 7 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €2.1 0.6%  This amounts to 0.03% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €1.4 0.4%  Near ly 4 0 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.7 0.2% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel &  Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n M o s t a r Bucharest Iași Belgrade Masonry and slab-column system are the two highest-risk building types used for pre-2000 multifamily housing in Mostar. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d S la b - co lu mn Syste m (S CS ) b ui l di ngs Almaty co nt r ib u te to : Shymkent Central Asia Bishkek ~2 0 , 0 0 0 peop l e URM 77% 63% 61% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 18% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Mo sta r SCS n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 29 Sofia BULGARIA Summ ary Sofia is Bulgaria’s capital and largest city, home to nearly expected to cause the most fatalities. Approximately 20 percent of the country’s total population. Bulgaria 260,000 people, or 20 percent of the city’s population, has faced several large earthquakes in the past, the most reside in high-risk building types. damaging and deadly of which occurred in 1928 near Plovdiv. This magnitude 7.0 earthquake caused over 1,700 Compared to the data for all the buildings in this study, injuries and 120 deaths. In 1986, Bulgaria experienced the data used to conduct the risk analysis for buildings a magnitude 5.7 earthquake in the Veliko Turnovo- in Sofia averages a medium reliability; this rating is a Turgovishte region that caused nearly €11 million (2019 function of the assumptions, uncertainties, and data €) in damage; over 150 homes were destroyed and available for the city (see Methodology). The diagram Sofia approximately 7,000 suffered damage, leaving 80 percent below indicates the reliability and confidence related to of homes uninhabitable. The earthquake also injured 60 building stock data (i.e., exposure data) as well as data people and killed three. Over a century earlier, in 1858, on building vulnerability. These confidence and reliability Sofia faced a magnitude 6.2 earthquake that damaged levels should be considered when interpreting and nearly 80 percent of the city’s building stock and left applying the risk metrics presented in the city profile. many residents homeless. *At risk building types based on earthquake loss are Based on the analysis, over 70 percent of the population determined by the expected average annual loss ratio, of Sofia resides in pre-2000 multifamily residential or average annual loss as a percentage of replacement buildings. The buildings that are the most at risk* value (AAL%). The expected loss results are predicted are reinforced concrete frame (RCF) buildings and from models of future events subject to variable data SOFIA FACTS unreinforced masonry (URM) buildings. Along with large availability and modeling assumptions. The replacement value is the construction cost of all multifamily buildings Capital City Population: 1,242,000 (2017) GNI/capita: ~€7,500 panel buildings (LPBs), these building types are also in the city. Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High LPB RCW € % RELIABILITY EXPOSURE 49% 7% URM RCF DFW Total building Building damage, Number of Medium 25% 18% 1% damage % of total value fatalities Low BLK SCS 100-year € 900 M 6% 650 0% 0% 14% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 4.4 B 27% 5,600 Medium Return Period 66% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 30 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 55.2 Billion Euros* 1928 Chripan M7.0 POPULATION 1986 Strazhitsa M5.7 7.03 Million People* 2012 Pernik M5.6 * 2018 estimate Buildin g D am age Lo s s es So fi a BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €15.0 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €12.5 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €10.0 0.4% frames that provide earthquake resistance; RC F is € 4 1 . 1 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €7.5 0.3%  This amounts to 0.07% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €5.0 0.2%  Near ly 7 1 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €2.5 0.1% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent  Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n So f i a Bucharest Iași Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Sofia. Bratislava Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~2 6 0 , 0 0 0 p eop l e RC F 56% 76% 55% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 21% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f So fi a URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 31 Plovdiv BULGARIA Summ ary Plovdiv—once known as Philippoupole—is Bulgaria’s Compared to the data for all the buildings in this study, third largest city and home to nearly 5 percent of the the data used to conduct the risk analysis for buildings country’s total population. Bulgaria has faced several in Plovdiv average a medium reliability; this rating is a large earthquakes in the past, the most damaging and function of the assumptions, uncertainties, and data deadly of which occurred in 1928 near Plovdiv. This available for the city (see Methodology). The diagram magnitude 7.0 earthquake caused over 1,700 injuries and below indicates the reliability and confidence related to 120 deaths. In 1986, Bulgaria experienced a magnitude building stock data (i.e., exposure data) as well as data 5.7 earthquake in the Veliko Turnovo-Turgovishte region on building vulnerability. These confidence and reliability that caused €11 million (2019 €) in damage; over 150 levels should be considered when interpreting and homes were destroyed and approximately 7,000 suffered applying the risk metrics presented in the city profile. Plovdiv damage, leaving 80 percent of homes uninhabitable. This earthquake also left 60 people injured and killed three. *At risk building types based on earthquake loss are determined by the expected average annual loss ratio, Based on the analysis, over two-thirds of the population or average annual loss as a percentage of replacement of Plovdiv resides in pre-2000 multifamily residential value (AAL%). The expected loss results are predicted buildings. The buildings that are the most at risk* from models of future events subject to variable data are reinforced concrete frame (RCF) buildings and availability and modeling assumptions. The replacement unreinforced masonry (URM) buildings. These two value is the construction cost of all multifamily buildings building types are also expected to cause the most in the city. fatalities. Approximately 65,000 people, or 20 percent of PLOVD IV FACTS the city’s population, reside in high-risk building types. Third Largest City Population: 347,000 (2018) GNI/capita: ~€7,500 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High LPB RCW € % RELIABILITY EXPOSURE 51% 6% URM RCF DFW Total building Building damage, Number of Medium 29% 13% 1% damage % of total value fatalities Low BLK SCS 100-year € 140 M 4% 150 0% 0% 8% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 660 M 21% 1,100 Medium Return Period 37% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 32 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 55.2 Billion Euros* 1928 Chripan M7.0 POPULATION 1986 Strazhitsa M5.7 7.03 Million People* 2012 Pernik M5.6 * 2018 estimate Buildin g D am age Lo s s es P lo vd i v BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €2.4 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €2.0 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €1.6 0.4% frames that provide earthquake resistance; RC F is € 6 . 5 5 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €1.2 0.3%  This amounts to 0.01% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €0.8 0.2%  Near ly 6 9 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.4 0.1% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%)  Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n P l o v d iv Bucharest Iași Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Plovdiv. Bratislava Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~6 5 , 0 0 0 peop l e RC F 61% 80% 67% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 20% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Pl ovd i v URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 33 Zagreb CROATIA Summ ary Croatia’s capital and largest city, Zagreb is home to Compared to the data for all the buildings in this study, nearly 20 percent of the country’s total population. the data used to conduct the risk analysis for buildings Croatia has faced several large earthquakes in the past, in Zagreb average a medium reliability; this rating is a the most damaging of which occurred in Dubrovnik in function of the assumptions, uncertainties, and data 1667. This catastrophic magnitude 7.2 earthquake and available for the city (see Methodology). The diagram Zagreb tsunami killed 5,000 people, nearly one-third of the below indicates the reliability and confidence related to city’s population. More recently, in March 2020, Zagreb building stock data (i.e., exposure data) as well as data and experienced a magnitude 5.4 earthquake that resulted models on building vulnerability. These confidence and in €6.9 billion (2020 €) in damage to the housing sector, reliability levels should be considered when interpreting with approximately 24,000 residential buildings affected, and applying the risk metrics presented in the city profile. 29 people injured, and one person killed. *At risk building types based on earthquake loss are Based on the analysis, over half of the population of determined by the expected average annual loss ratio, Zagreb resides in pre-2000 multifamily residential or average annual loss as a percentage of replacement buildings. The buildings that are the most at risk* are value (AAL%). The expected loss results are predicted unreinforced masonry (URM) buildings and reinforced from models of future events subject to variable data concrete frame (RCF) buildings. These two building availability and modeling assumptions. The replacement types are also expected to cause the most fatalities. value is the construction cost of all multifamily buildings Approximately 280,000 people, over two-thirds of the in the city. city’s population, reside in high-risk building types. ZAG REB FACTS Capital City Population: 804,000 (2017) GNI/capita: ~€12,000 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High LPB € % RELIABILITY EXPOSURE 1% Medium URM RCF DFW RCW Total building Building damage, Number of 67% 18% 5% 9% damage % of total value fatalities Low BLK SCS 100-year € 1.3 B 12% 1,900 0% 0% 26% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 5.2 M 47% 9,500 Medium Return Period 104% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 34 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 51.6 Billion Euros* 1667 Dubrovnik M7.2 POPULATION 1962 Makarska M6.1 4.09 Million People* 2020 Zagreb M5.4 * 2018 estimate Buildin g D am age Lo s s es Zagre b BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €30 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €25 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €20 0.4% frames that provide earthquake resistance; RC F is € 4 8 . 8 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €15 0.3%  This amounts to 0.09% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €10 0.2%  Near ly 5 2 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €5 0.1% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of  Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Zagre b Bucharest Iași Belgrade Masonry and reinforced concrete frame are the two highest-risk building types used for pre-2000 multifamily housing in Zagreb. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d Re inforc e d Co n c rete Fra me (RC F) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~2 8 0 , 0 0 0 p eop l e URM 83% 77% 87% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 35% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Za greb RC F n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 35 R ijeka CROATIA Summ ary A major port city and Croatia’s third largest city, Rijeka is Compared to the data for all the buildings in this study, home to nearly 3 percent of Croatia’s total population. the data used to conduct the risk analysis for buildings Croatia has faced several large earthquakes in the past, in Rijeka average a medium reliability; this rating is a the most damaging of which occurred in Dubrovnik in function of the assumptions, uncertainties, and data 1667. This catastrophic magnitude 7.2 earthquake and available for the city (see Methodology). The diagram tsunami killed 5,000 people, nearly one-third of the below indicates the reliability and confidence related to city’s population. More recently, in March 2020, Zagreb building stock data (i.e., exposure data) as well as data Rijeka experienced a magnitude 5.4 earthquake that resulted on building vulnerability. These confidence and reliability in €6.9 billion (2020 €) in damage to the housing sector, levels should be considered when interpreting and with approximately 24,000 residential buildings affected, applying the risk metrics presented in the city profile. 29 people injured, and one person killed. *At risk building types based on earthquake loss are Based on the analysis, over half of the population determined by the expected average annual loss ratio, of Rijeka resides in pre-2000 multifamily residential or average annual loss as a percentage of replacement buildings. The buildings that are the most at risk* are value (AAL%). The expected loss results are predicted unreinforced masonry (URM) buildings and reinforced from models of future events subject to variable data concrete frame (RCF) buildings. These two building availability and modeling assumptions. The replacement types are also expected to cause the most fatalities. value is the construction cost of all multifamily buildings Approximately 38,000 people, or about one-third of the in the city. city’s population, reside in high-risk building types. RIJEKA FACTS Third Largest City Population: 119,000 (2017) GNI/capita: ~€12,000 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High LPB € % RELIABILITY EXPOSURE 2% Medium URM RCF DFW RCW Total building Building damage, Number of 71% 14% 4% 9% damage % of total value fatalities Low BLK SCS 100-year € 108 M 9% 240 0% 0% 14% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 450 M 40% 1,200 Medium Return Period 60% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 36 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 51.6 Billion Euros* 1667 Dubrovnik M7.2 POPULATION 1962 Makarska M6.1 4.09 Million People* 2020 Zagreb M5.4 * 2018 estimate Buildin g D am age Lo s s es Ri j e k a BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €3.0 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €2.5 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €2.0 0.4% frames that provide earthquake resistance; RC F is € 4 . 7 5 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €1.5 0.3%  This amounts to 0.01% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €1.0 0.2%  Near ly 5 4 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.5 0.1% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information.  Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Ri jek a Bucharest Iași Belgrade Masonry and reinforced concrete frame are the two highest-risk building types used for pre-2000 multifamily housing in Rijeka. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d Re inforc e d Co n c rete Fra me (RC F) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~3 8 , 0 0 0 peop l e URM 81% 82% 80% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 32% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f R i jeka RC F n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 37 Budapest HUNGARY Summ ary Budapest is the capital and largest city in Hungary, home Compared to the data for all the buildings in this study, to nearly 20 percent of the total population. Hungary has the data used to conduct the risk analysis for buildings in experienced several earthquakes in the past, including Budapest average a low reliability; this rating is a function a historic magnitude 6.2 event in 1763 that killed of the assumptions, uncertainties, and data available for 83 people. In 1956, the magnitude 5.8 Dunaharaszti the city (see Methodology). The diagram below indicates earthquake caused damage to over 3,000 buildings and the reliability and confidence related to building stock led to two deaths. data (i.e., exposure data) as well as data on building vulnerability. These confidence and reliability levels Budapest Based on the analysis, about 70 percent of the population should be considered when interpreting and applying the of Budapest resides in pre-2000 multifamily residential risk metrics presented in the city profile. buildings. The buildings that are the most at risk* are unreinforced masonry (URM) buildings and reinforced *At risk building types based on earthquake loss are concrete frame (RCF) buildings. These two building types determined by the expected average annual loss ratio, are also expected to cause the most fatalities. Nearly or average annual loss as a percentage of replacement 440,000 people, or about one in every four inhabitants, value (AAL%). The expected loss results are predicted reside in high-risk building types. from models of future events subject to variable data availability and modeling assumptions. The replacement value is the construction cost of all multifamily buildings in the city. BUD APEST FACTS Capital City Population: 1,752,000 (2017) GNI/capita: ~€12,500 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High € % RELIABILITY EXPOSURE URM RCF DFW LPB SCS RCW Total building Building damage, Number of Medium 65% 5% 5% 20% 2% 3% damage % of total value fatalities Low BLK 100-year € 1.1 B 3% 280 0% 7% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 5.0 B 13% 4,000 Medium Return Period 34% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 38 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 133.7 Billion Euros* 1763 Komárom M6.2 POPULATION 1956 Dunaharaszti M5.8 9.78 Million People* 2006 Beregdaróc M4.8 * 2018 estimate Buildin g D am age Lo s s es B u d ap e st BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €42 0.24% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €35 0.20% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €28 0.16% frames that provide earthquake resistance; RC F is € 4 8 . 8 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €21 0.12%  This amounts to 0.04% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €14 0.08%  Near ly 6 8 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €7 0.04% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka  Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Bu d a p e st Bucharest Iași Belgrade Masonry and reinforced concrete frame are the two highest-risk building types used for pre-2000 multifamily housing in Budapest. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d Re inforc e d Co n c rete Fra me (RC F) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~4 4 0 , 0 0 0 p eop l e URM 90% 80% 79% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 25% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Buda p est RC F n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 39 C hiş inău MOLDOVA Summ ary Chişinău is Moldova’s capital and largest city, home Compared to the data for all the buildings in this study, to nearly a quarter of the country’s total population. the data used to conduct the risk analysis for buildings in Moldova has been affected by devastating earthquakes Chişinău average a high reliability; this rating is a function in the past, the most deadly of which was the 1940 of the assumptions, uncertainties, and data available for magnitude 7.8 Vrancea earthquake that killed 78 people. the city (see Methodology). The diagram below indicates In 1986, another earthquake centered in Vrancea the reliability and confidence related to building stock of magnitude 7.2 caused over 550 injuries and two data (i.e., exposure data) as well as data on building deaths, damaged 55,000 homes, and left 12,500 people vulnerability. These confidence and reliability levels homeless. should be considered when interpreting and applying the Chişinău risk metrics presented in the city profile. Based on the analysis, over 80 percent of the population of Chişinău resides in pre-2000 multifamily residential *At risk building types based on earthquake loss are buildings. The buildings that are the most at risk* are determined by the expected average annual loss ratio, unreinforced masonry (URM) buildings and precast large or average annual loss as a percentage of replacement panel buildings (LPBs). Including reinforced concrete wall value (AAL%). The expected loss results are predicted (RCW), these building types are also expected to cause from models of future events subject to variable data the most fatalities. Over 324,000 people, or about one in availability and modeling assumptions. The replacement every two inhabitants, reside in high-risk building types. value is the construction cost of all multifamily buildings in the city. CH ISINAU FACTS Capital City Population: 686,000 (2017) GNI/capita: ~€2,500 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High URM LPB RCW € % RELIABILITY EXPOSURE 46% 29% 25% Total building Building damage, Number of Medium damage % of total value fatalities Low RCF 0% DFW 0% BLK 0% SCS 0% 100-year € 76 M 1% 100 Return Period 5% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 700 M 9% 940 Medium Return Period 50% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 40 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 9.70 Billion Euros* 1940 Vrancea M7.8 POPULATION 1986 Vrancea M7.2 2.71 Million People* 1990 Vrancea M7.0 * 2018 estimate Buildin g D am age Lo s s es C hi si nau BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €3.6 0.18% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €3.0 0.15% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €2.4 0.12% frames that provide earthquake resistance; RC F is € 6 . 2 6 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €1.8 0.09%  This amounts to 0.06% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €1.2 0.06%  Near ly 8 1 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.6 0.03% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest  Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n C h i s i n a u Bucharest Iași Belgrade Masonry and precast large panel are the two highest-risk building types used for pre-2000 multifamily housing in Chişinău. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d Pre ca st L a rge Pa n e l ( L PB ) b ui l di ngs Almaty co nt r ib u te to : Shymkent Central Asia Bishkek ~3 2 4 , 0 0 0 p eop l e URM 74% 87% 74% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 47% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f C hi şi n ă u LPB n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 41 Podgorica MONTENEGRO Summ ary Podgorica is Montenegro’s capital and largest city, home Compared to the data for all the buildings in this study, to nearly a quarter of the country’s total population. the data used to conduct the risk analysis for buildings Montenegro has faced several large earthquakes in the in Podgorica average a low reliability; this rating is a past, the most damaging and deadly of which occurred in function of the assumptions, uncertainties, and data 1979 near the Montenegrin coastline. The magnitude 6.9 available for the city (see Methodology). The diagram earthquake and tsunami caused nearly €9 billion (2019 below indicates the reliability and confidence related to €) in damage, injured 1,000 individuals, and killed 121. building stock data (i.e., exposure data) as well as data on building vulnerability. These confidence and reliability Based on the analysis, one-third of the population of levels should be considered when interpreting and Podgorica resides in pre-2000 multifamily residential applying the risk metrics presented in the city profile. buildings. The buildings that are the most at risk* Podgorica are reinforced concrete frame (RCF) buildings and *At risk building types based on earthquake loss are unreinforced masonry (URM) buildings. These two determined by the expected average annual loss ratio, building types are also expected to cause the most or average annual loss as a percentage of replacement fatalities. Approximately 14,000 people, or 9 percent of value (AAL%). The expected loss results are predicted the city’s population, reside in high-risk building types. from models of future events subject to variable data availability and modeling assumptions. The replacement value is the construction cost of all multifamily buildings in the city. POD G ORICA FACTS Capital City Population: 151,000 (2011) GNI/capita: ~€7,000 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High € % RELIABILITY EXPOSURE URM RCF DFW LPB RCW Total building Building damage, Number of Medium 26% 18% 13% 27% 16% damage % of total value fatalities Low BLK 0% SCS 0% 100-year € 61 M 6% 40 Return Period 17% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 276 M 25% 260 Medium Return Period 78% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 42 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 4.66 Billion Euros* 1968 Montenegro M5.3 POPULATION 1979 Montenegro M6.9 0.622 Million People* 2018 Plav M4.9 * 2018 estimate Buildin g D am age Lo s s es P o dgo ri ca BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €1.50 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €1.25 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €1.00 0.4% frames that provide earthquake resistance; RC F is € 2 . 9 0 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €0.75 0.3%  This amounts to 0.06% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €0.50 0.2%  Near ly 3 3 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.25 0.1% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău  Podgorica Skopje High -R is k Bu i l d i n gs i n P o d g or ica Bucharest Iași Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Podgorica. Bratislava Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~1 4 , 0 0 0 peop l e RC F 54% 79% 76% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 9% o f the to ta l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Po dgor i ca URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 43 Skopje NORTH MACEDONIA Summ ary Skopje is North Macedonia’s capital and largest city, home frame (RCF) buildings. These two building types are also to more than a quarter of the country’s total population. expected to cause the most fatalities. Approximately North Macedonia has faced several large earthquakes in 38,000 people, or 7 percent of the city’s population, the past, the most damaging and deadly of which was reside in high-risk building types. in 1963 in Skopje. The magnitude 6.0 Skopje earthquake caused €7.7 billion (2019 €) in damage, injured 3,300 Compared to the data for all the buildings in this study, Skopje individuals, and killed over 1,000. The earthquake the data used to conduct the risk analysis for buildings damaged 80 percent of Skopje’s building stock, including in Skopje average a medium reliability; this rating is a historic monuments, and left three of every four residents function of the assumptions, uncertainties, and data homeless. The extensive damage can be attributed to the available for the city (see Methodology). The diagram prevalence of non-engineered multistory brick buildings, below indicates the reliability and confidence related to many of which collapsed. Following this destructive building stock data (i.e., exposure data) as well as data earthquake, unconfined masonry buildings were no on building vulnerability. These confidence and reliability longer constructed in the country, and redevelopment levels should be considered when interpreting and of Skopje limited the density of inhabitants in central applying the risk metrics presented in the city profile. areas to reduce the population’s vulnerability to future *At risk building types based on earthquake loss are earthquakes. Historic earthquakes that destroyed the determined by the expected average annual loss ratio, city of Skopje occurred in 518 and 1555; other damaging or average annual loss as a percentage of replacement and fatal earthquakes occurred more recently in 1982 value (AAL%). The expected loss results are predicted SKOPJE FACTS and 1983. from models of future events subject to variable data availability and modeling assumptions. The replacement Capital City Population: 546,000 (2018) GNI/capita: ~€4,600 Based on the analysis, about 30 percent of the population of Skopje resides in pre-2000 multifamily residential value is the construction cost of all multifamily buildings buildings. The buildings that are the most at risk* are slab- in the city. column system (SCS) buildings and reinforced concrete Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s LPB € % RELIABILITY High EXPOSURE 28% URM RCF DFW SCS RCW Total building Building damage, Number of Medium 27% 17% 3% 6% 19% damage % of total value fatalities Low BLK 0% 100-year € 130 M 7% 110 Return Period 18% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 630 M 32% 1,100 Medium Return Period 86% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 44 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 10.7 Billion Euros* 1963 Skopje M6.0 POPULATION 1983 Skopje M4.7 2.08 Million People* 2016 Skopje M5.1 * 2018 estimate Buildin g D am age Lo s s es Sk o pj e BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €1.8 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €1.5 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €1.2 0.4% frames that provide earthquake resistance; RC F is € 5 . 7 8 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €0.9 0.3%  This amounts to 0.05% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €0.6 0.2%  Near ly 2 9 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.3 0.1% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica  Skopje High -R is k Bu i l d i n gs i n Sko pj e Bucharest Iași Belgrade Slab-column system and reinforced concrete frame are the two highest-risk building types used for pre-2000 multifamily housing in Skopje. Bratislava Ljubljana S la b - co lu mn Syste m (S CS ) a n d Re inforc e d Co n c rete Fra me ( RCF ) bui l di ngs Almaty co nt r ib u te to : Shymkent Central Asia Bishkek ~3 8 , 0 0 0 peop l e SCS 42% 51% 60% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 7% o f the to ta l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Sko p je RC F n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 45 Bucharest ROMANIA Summ ary Bucharest is Romania’s capital and largest city, home 1.1 million people, or about 60 percent of the city’s to nearly 10 percent of the country’s total population. population, reside in high-risk building types. Romania has experienced many severe earthquakes in the past, the most damaging and deadly of which was the Compared to the data for all the buildings in this study, 1977 Vrancea earthquake. This magnitude 7.5 earthquake the data used to conduct the risk analysis for buildings resulted in over 11,000 injuries, more than 1,500 fatalities in Bucharest average a low reliability; this rating is a (90 percent of them in Bucharest), and €5.5 billion (2019 function of the assumptions, uncertainties, and data €) in damage to the capital city. Romania experienced available for the city (see Methodology). The diagram a magnitude 7.8 earthquake in 1940, also centered in below indicates the reliability and confidence related to Vrancea, which resulted in approximately 600 deaths and building stock data (i.e., exposure data) as well as data 1,200 injuries in the country. In Bucharest, the 14-story for building vulnerability. These confidence and reliability Carlton Block residential building completely collapsed, levels should be considered when interpreting and applying the risk metrics presented in the city profile. Bucharest killing 140 residents and injuring nearly 90 people. Following the 1977 Vrancea earthquake, Romania *At risk building types based on earthquake loss are updated the building design code to improve seismic determined by the expected average annual loss ratio, requirements for reinforced concrete structures. or average annual loss as a percentage of replacement Based on the analysis, about 90 percent of the population value (AAL%). The expected loss results are predicted of Bucharest resides in pre-2000 multifamily residential from models of future events subject to variable data buildings. The buildings that are the most at risk* are availability and modeling assumptions. The replacement BUCH AREST FACTS reinforced concrete frame (RCF) and unreinforced value is the construction cost of all multifamily buildings in the city. Capital City Population: 1,828,000 (2018) GNI/capita: ~€10,000 masonry (URM) buildings. These two building types are also expected to cause the most fatalities. Approximately Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High € % RELIABILITY EXPOSURE URM RCF LPB RCW Total building Building damage, Number of Medium 30% 53% 12% 5% damage % of total value fatalities Low DFW BLK SCS 100-year € 5.3 B 11% 3,000 0% 0% 0% 34% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 18.9 B 41% 19,300 Medium Return Period 121% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 46 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 203.0 Billion Euros* 1940 Vrancea M7.8 POPULATION 1977 Vrancea M7.5 19.5 Million People* 1986 Vrancea M7.2 * 2018 estimate Buildin g D am age Lo s s es B u chare st BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €144 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €120 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €96 0.4% frames that provide earthquake resistance; RC F is € 1 9 2 m i l l i on . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €72 0.3%  This amounts to 0.10% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €48 0.2%  Near ly 8 9 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €24 0.1% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Bu c h a r e st  Bucharest Iași Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Bucharest. Bratislava Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~1 , 1 0 0 , 0 0 0 RC F 84% 82% 89% Osh Dushanbe peo pl e l i ve i n the s e o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat bui l di ngs, b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* Tashkent d is p la c e me n t i n t h e 61% Yerevan S. Caucasus or o f the tota l URM n ex t 5 0 ye a rs f ro m Gyumri po pul ati o n o f Buch a rest b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 47 Iași ROMANIA Summ ary Iași is Romania’s second largest city and is home to about or over half of the city’s population, reside in high-risk 1 percent of the country’s total population. Romania building types. has experienced many severe earthquakes in the past, the most damaging and deadly of which was the 1977 Compared to the data for all the buildings in this study, Vrancea earthquake. This magnitude 7.5 earthquake the data used to conduct the risk analysis for buildings resulted in over 11,000 injuries and more than 1,500 in Iași average a low reliability; this rating is a function of the assumptions, uncertainties, and data available for Iași fatalities, destroyed more than 10 percent of the dwellings in Iași, and resulted in €7.8 billion in damage. Romania the city (see Methodology). The diagram below indicates experienced a magnitude 7.8 earthquake in 1940, also the reliability and confidence related to building stock centered in Vrancea, which resulted in approximately 600 data (i.e., exposure data) as well as data on building deaths and 1,200 injuries in the country. Iași had many vulnerability. These confidence and reliability levels more injuries than fatalities from both events, mainly should be considered when interpreting and applying the from damage to mid-rise masonry buildings. Following risk metrics presented in the city profile. the 1977 Vrancea earthquake, Romania updated the *At risk building types based on earthquake loss are building design code to improve seismic requirements determined by the expected average annual loss ratio, for reinforced concrete structures. or average annual loss as a percentage of replacement Based on the analysis, over 70 percent of the population of value (AAL%). The expected loss results are predicted Iași resides in pre-2000 multifamily residential buildings. from models of future events subject to variable data The buildings that are the most at risk* are unreinforced availability and modeling assumptions. The replacement IASI FACTS masonry (URM) and reinforced concrete frame (RCF) value is the construction cost of all multifamily buildings in the city. Second Largest City Population: 290,000 (2011) GNI/capita: ~€10,000 buildings. These two building types are also expected to cause the most fatalities. Approximately 154,000 people, Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High € % RELIABILITY EXPOSURE URM RCF LPB RCW Total building Building damage, Number of Medium 59% 26% 14% 1% damage % of total value fatalities Low DFW BLK SCS 100-year € 140 M 8% 400 0% 0% 0% 7% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year € 570 M 32% 2,200 Medium Return Period 29% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 48 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 203.0 Billion Euros* 1940 Vrancea M7.8 POPULATION 1977 Vrancea M7.5 19.5 Million People* 1986 Vrancea M7.2 * 2018 estimate Buildin g D am age Lo s s es Iasi BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €3.0 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €2.5 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €2.0 0.4% frames that provide earthquake resistance; RC F is € 5 . 4 1 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and  This amounts to less €1.5 0.3% wall elements work together to resist earthquake DF W than 0 . 0 1 % of the forces. countr y ’s GDP. Precast Large Panel: built with precast reinforced €1.0 0.2% concrete panels with floor-to-ceiling height facade LP B  Near ly 7 1 % of the panels. city ’s population lives in €0.5 0.1% Block: built with precast concrete elements multifamily residential assembled on site; similar to LPB with smaller BLK buildings. facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n I as i Bucharest  Iași Belgrade Masonry and reinforced concrete frame are the two highest-risk building types used for pre-2000 multifamily housing in Iași. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d Re inforc e d Co n c rete Fra me (RC F) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~1 5 4 , 0 0 0 p eop l e URM 93% 83% 94% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 53% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Iași RC F n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 49 Belgrade SERBIA Summ ary Belgrade is Serbia’s capital and largest city, home to of the assumptions, uncertainties, and data available for nearly a quarter of the country’s total population. Serbia the city (see Methodology). The diagram below indicates has experienced several earthquakes of varying severity the reliability and confidence related to building stock in the past, the most recent of which occurred in 2010 data (i.e., exposure data) as well as data for building in Kraljevo. This magnitude 5.5 earthquake injured over vulnerability. These confidence and reliability levels 100, killed two, destroyed 1,000 homes, and damaged a should be considered when interpreting and applying the further 5,000. risk metrics presented in the city profile. Belgrade Based on the analysis, nearly three-quarters of the *At risk building types based on earthquake loss are population of Belgrade resides in pre-2000 multifamily determined by the expected average annual loss ratio, residential buildings. The buildings that are the most at or average annual loss as a percentage of replacement risk* are reinforced concrete frame (RCF) and unreinforced value (AAL%). The expected loss results are predicted masonry (URM) buildings. These two building types are from models of future events subject to variable data also expected to cause the most fatalities. Nearly 550,000 availability and modeling assumptions. The replacement inhabitants, or one in every three people, reside in high- value is the construction cost of all multifamily buildings risk building types. in the city. Compared to the data for all the buildings in this study, the data used to conduct the risk analysis for buildings in Belgrade average a low reliability; this rating is a function BELG RAD E FACTS Capital City Population: 1,687,000 (2011) GNI/capita: ~€5,400 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High € % RELIABILITY EXPOSURE URM RCF DFW LPB SCS RCW Total building Building damage, Number of Medium 80% 2% 2% 11% 3% 2% damage % of total value fatalities Low BLK 0% 100-year € 434 M 1% 240 Return Period 6% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 3.4 B 10% 2,000 Medium Return Period 50% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 50 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 42.9 Billion Euros* 1980 Kopaonik M5.8 POPULATION 1998 Mionica M5.5 6.93 Million People* 2010 Kraljevo M5.5 * 2018 estimate Buildin g D am age Lo s s es B e lgrad e BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €30 0.18% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €25 0.15% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €20 0.12% frames that provide earthquake resistance; RC F is € 3 3 . 6 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €15 0.09%  This amounts to 0.08% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €10 0.06%  Near ly 7 4 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €5 0.03% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Bel gra d e Bucharest Iași  Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Belgrade. Bratislava Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~5 5 0 , 0 0 0 p eop l e RC F 81% 59% 77% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 33% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Bel g ra d e URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 51 Bratislava SLOVAKIA Summ ary Bratislava is Slovakia’s capital and largest city, home in Bratislava average a medium reliability; this rating is to nearly 8 percent of the country’s total population. a function of the assumptions, uncertainties, and data Slovakia has faced damaging and deadly earthquakes available for the city (see Methodology). The diagram in the past; the country’s earliest recorded earthquake, below indicates the reliability and confidence related to in Banská Štiavnica in 1443, caused 30 deaths. In 2004, building stock data (i.e., exposure data) as well as data Slovakia experienced an earthquake in Slovenská L’upca for building vulnerability. These confidence and reliability that caused minor damage in the city. levels should be considered when interpreting and applying the risk metrics presented in the city profile. Based on the analysis, nearly two-thirds of the population of Bratislava resides in pre-2000 multifamily *At risk building types based on earthquake loss are residential buildings. The buildings that are the most determined by the expected average annual loss ratio, Bratislava at risk* are reinforced concrete frame (RCF) buildings or average annual loss as a percentage of replacement and unreinforced masonry (URM) buildings. These two value (AAL%). The expected loss results are predicted building types are also expected to cause the most from models of future events subject to variable data fatalities. Approximately 30,000 people, or about 7 availability and modeling assumptions. The replacement percent of the city’s population, reside in high-risk value is the construction cost of all multifamily buildings building types. in the city. Compared to the data for all the buildings in this study, the data used to conduct the risk analysis for buildings BRATISLAVA FACTS Capital City Population: 424,000 (2016) GNI/capita: ~€15,500 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High € % RELIABILITY EXPOSURE URM RCF LPB RCW Total building Building damage, Number of Medium 6% 14% 73% 7% damage % of total value fatalities Low DFW 0% BLK 0% SCS 0% 100-year € 102 M 1% 10 Return Period 2% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 830 M 6% 120 Medium Return Period 20% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 52 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 89.7 Billion Euros* 1443 Banská Štiavnica M-unknown POPULATION 1858 Zilina M-unknown 5.45 Million People* 2004 Slovenská L’upca M5.0 * 2018 estimate Buildin g D am age Lo s s es B rati slava BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €6 0.18% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €5 0.15% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €4 0.12% frames that provide earthquake resistance; RC F is € 8 . 4 8 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €3 0.09%  This amounts to 0.01% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €2 0.06%  Near ly 6 4 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €1 0.03% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Br at isl a v a Bucharest Iași Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Bratislava.  Bratislava Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~3 0 , 0 0 0 peop l e RC F 30% 84% 76% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 7% o f the to ta l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Brati sl ava URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 53 Ljubljana SLOVENIA Summ ary Ljubljana is Slovenia’s capital and largest city, home to Compared to the data for all the buildings in this study, nearly 14 percent of the country’s total population. the data used to conduct the risk analysis for buildings in Slovenia has faced several large earthquakes in the past, Ljubljana average a low reliability; this rating is a function the most severe of which occurred in Ljubljana in 1895. of the assumptions, uncertainties, and data available for This magnitude 6.1 event, sometimes called the Easter the city (see Methodology). The diagram below indicates Earthquake, killed seven people and damaged 10 percent the reliability and confidence related to building stock of all homes in the city. In 2004, Slovenia experienced a data (i.e., exposure data) as well as data for building magnitude 5.2 earthquake in the Bovec-Kobarid region vulnerability. These confidence and reliability levels that injured five people and killed one. This earthquake should be considered when interpreting and applying the Ljubljana was also felt in Ljubljana, but resulted in no recorded risk metrics presented in the city profile. damage or casualties. *At risk building types based on earthquake loss are Based on the analysis, about two-thirds of the population determined by the expected average annual loss ratio, of Ljubljana resides in pre-2000 multifamily residential or average annual loss as a percentage of replacement buildings. The buildings that are the most at risk* are value (AAL%). The expected loss results are predicted reinforced concrete frame (RCF) and unreinforced from models of future events subject to variable data masonry (URM) buildings. Along with reinforced concrete availability and modeling assumptions. The replacement wall (RCW), these building types are expected to cause value is the construction cost of all multifamily buildings the most fatalities. Approximately 78,000 people, or in the city. nearly 30 percent of the city’s population, reside in high- LJUBLJANA FACTS risk building types. Capital City Population: 280,000 (2016) GNI/capita: ~€20,800 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s € % RELIABILITY High EXPOSURE Medium DFW LPB RCW Total building Building damage, Number of 9% 3% 19% damage % of total value fatalities Low URM RCF 58% 11% BLK SCS 100-year € 400 M 7% 340 0% 0% 10% GNIPC people Return Period VULNERABILITY RELIABILITY Low 475-year €1.9 B 34% 2,000 Medium Return Period 47% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 54 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 45.7 Billion Euros* 1511 Idrija M6.5 POPULATION 1895 Ljubljana M6.1 2.07 Million People* 2004 Bovec-Kobarid M5.2 * 2018 estimate Buildin g D am age Lo s s es L j u blj ana BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €9.0 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €7.5 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €6.0 0.4% frames that provide earthquake resistance; RC F is € 1 7 . 9 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €4.5 0.3%  This amounts to 0.04% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €3.0 0.2%  Near ly 6 7 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €1.5 0.1% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Lju b l j a n a Bucharest Iași Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Ljubljana. Bratislava  Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~7 8 , 0 0 0 peop l e RC F 64% 77% 75% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 28% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Ljub l ja n a URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 55 Almaty KAZAKHSTAN Summ ary Almaty (formerly Alma-Ata) is Kazakhstan’s largest city, Compared to the data for all the buildings in this study, the country’s former capital, and home to 10 percent the data used to conduct the risk analysis for buildings in of the country’s total population. Kazakhstan has faced Almaty average a low reliability; this rating is a function several large earthquakes in the past, the most deadly of of the assumptions, uncertainties, and data available for which was the 1911 magnitude 7.7 earthquake in Almaty the city (see Methodology). The diagram below indicates that killed 450 people. Another event, the magnitude the reliability and confidence related to building stock 6.8 Zaysan earthquake in 1990, killed one individual, data (i.e., exposure data) as well as data for building destroyed 3,000 houses, and left 20,000 people homeless vulnerability. These confidence and reliability levels in the region. should be considered when interpreting and applying the risk metrics presented in the city profile. Based on the analysis, almost 40 percent the population Almaty of Almaty resides in pre-2000 multifamily residential *At risk building types based on earthquake loss are buildings. The buildings that are the most at risk* are determined by the expected average annual loss ratio, block (BLK) buildings and slab-column system (SCS) or average annual loss as a percentage of replacement buildings. Along with unreinforced masonry (URM) value (AAL%). The expected loss results are predicted buildings, these building types are also expected to cause from models of future events subject to variable data the most fatalities. Approximately 13,000 people, or less availability and modeling assumptions. The replacement than 1 percent of the city’s population, reside in high-risk value is the construction cost of all multifamily buildings building types. in the city. ALMATY FACTS Capital City Population: 1,863,000 (2019) GNI/capita: ~€6,800 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High LPB € % RELIABILITY EXPOSURE 50% URM RCF BLK SCS RCW Total building Building damage, Number of Medium 45% 3% <1% <1% <1% damage % of total value fatalities Low DFW 0% 100-year € 760 M 9% 1,600 Return Period 16% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 3.5 B 42% 15,000 Medium Return Period 75% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 56 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 152 Billion Euros* 1911 Almaty M7.7 POPULATION 1990 Zaysan M6.8 18.3 Million People* 2009 Tekeli M5.4 * 2018 estimate Buildin g D am age Lo s s es Almaty BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €18 3.0% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €15 2.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €12 2.0% frames that provide earthquake resistance; RC F is € 3 1 . 5 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €9 1.5%  This amounts to 0.02% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €6 1.0%  Near ly 3 6 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €3 0.5% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n A l m a t y Bucharest Iași Belgrade Block and slab-column system are the two highest-risk building types used for pre-2000 multifamily housing in Almaty. Bratislava Ljubljana Blo c k ( BL K ) a n d S la b - co lu mn Syste m (S CS ) b u ild in g s co nt r ib u te to :  Almaty Shymkent Central Asia Bishkek ~1 3 , 0 0 0 peop l e BLK 6% 60% 58% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or l e ss than 1% of d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus t he to tal po pul ati o n of SCS n ex t 5 0 ye a rs f ro m Gyumri A l m aty b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 57 Shymkent KAZAKHSTAN Summ ary Shymkent is Kazakhstan’s third largest city and home to of the assumptions, uncertainties, and data available for 5 percent of the country’s total population. Kazakhstan the city (see Methodology). The diagram below indicates has faced several large earthquakes in the past, the most the reliability and confidence related to building stock deadly of which was the 1911 magnitude 7.7 earthquake data (i.e., exposure data) as well as data for building in Almaty (formerly Alma-Ata) that killed 450 people. vulnerability. These confidence and reliability levels Another event, the magnitude 6.8 Zaysan earthquake in should be considered when interpreting and applying the 1990, killed one individual, destroyed 3,000 houses, and risk metrics presented in the city profile. left 20,000 people homeless in the region. *At risk building types based on earthquake loss are Based on the analysis, over a quarter of the population determined by the expected average annual loss ratio, of Shymkent resides in pre-2000 multifamily residential or average annual loss as a percentage of replacement buildings. The buildings that are the most at risk* are value (AAL%). The expected loss results are predicted Shymkent block (BLK) buildings and unreinforced masonry (URM) from models of future events subject to variable data buildings. These two building types are also expected to availability and modeling assumptions. The replacement cause the most fatalities. Approximately 100,000 people, value is the construction cost of all multifamily buildings or about 11 percent of the city’s population, reside in in the city. high-risk building types. Compared to the data for all the buildings in this study, the data used to conduct the risk analysis for buildings in SH YMKENT FACTS Shymkent average a low reliability; this rating is a function Third Largest City Population: 932,000 (2017) GNI/capita: ~€6,800 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High LPB € % RELIABILITY EXPOSURE 45% URM RCF BLK RCW Total building Building damage, Number of Medium 44% 2% 4% 5% damage % of total value fatalities Low DFW 0% SCS 0% 100-year € 67 M 4% 340 Return Period 4% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 290 M 18% 2,300 Medium Return Period 18% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 58 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 152 Billion Euros* 1911 Almaty M7.7 POPULATION 1990 Zaysan M6.8 18.3 Million People* 2009 Tekeli M5.4 * 2018 estimate Buildin g D am age Lo s s es Shy mk e nt BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €3.0 1.2% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €2.5 1.0% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €2.0 0.8% frames that provide earthquake resistance; RC F is € 3 . 0 4 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and  This amounts to less €1.5 0.6% wall elements work together to resist earthquake DF W than 0 . 0 1 % of the forces. countr y ’s GDP. Precast Large Panel: built with precast reinforced €1.0 0.4% concrete panels with floor-to-ceiling height facade LP B  Near ly 2 6 % of the panels. city ’s population lives in €0.5 0.2% Block: built with precast concrete elements multifamily residential assembled on site; similar to LPB with smaller BLK buildings. facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Sh y m k e n t Bucharest Iași Belgrade Block and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Shymkent. Bratislava Ljubljana Blo c k ( BL K ) a n d U n re info rc e d Ma s o n r y (U RM) b u ild in g s cont r i bute to : Almaty  Shymkent Central Asia Bishkek ~1 0 0 , 0 0 0 p eop l e BLK 86% 81% 86% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 11% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Shy m kent URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 59 Bishkek KYRGYZ REPUBLIC Summ ary Bishkek is the capital and largest city in the Kyrgyz Compared to the data for all the buildings in this study, Republic, home to nearly 17 percent of the total the data used to conduct the risk analysis for buildings population. The Kyrgyz Republic has faced several large in Bishkek average a medium reliability; this rating is a earthquakes in the past, the most damaging of which was function of the assumptions, uncertainties, and data the Suusamyr earthquake that occurred in Toluk in 1992. available for the city (see Methodology). The diagram This magnitude 7.5 event killed 75 people, destroyed over below indicates the reliability and confidence related to 8,000 houses, and caused damage totaling €217 million building stock data (i.e., exposure data) as well as the Bishkek (2019 €); in Bishkek, it collapsed about 95 percent of the data on building vulnerability. These confidence and single-story masonry homes. More recently, in 2008, reliability levels should be considered when interpreting the country experienced a magnitude 6.6 earthquake and applying the risk metrics presented in the city profile. centered in the town of Nura. This earthquake killed 74 people and injured 140. *At risk building types based on earthquake loss are determined by the expected average annual loss ratio, Based on the analysis, almost 40 percent of the population or average annual loss as a percentage of replacement of Bishkek resides in pre-2000 multifamily residential value (AAL%). The expected loss results are predicted buildings. The buildings that are the most at risk* are from models of future events subject to variable data reinforced concrete frame (RCF) and unreinforced availability and modeling assumptions. The replacement masonry (URM) buildings. These two building types are value is the construction cost of all multifamily buildings also expected to cause the most fatalities. Approximately in the city. 143,000 people, or about 14 percent of the population of BISH KEK FACTS Bishkek, reside in high-risk building types. Capital City Population: 1,013,000 (2019) GNI/capita: ~€1,000 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High URM LPB € % RELIABILITY EXPOSURE 50% 38% RCF SCS RCW Total building Building damage, Number of Medium 9% <1% 2% damage % of total value fatalities Low DFW 0% BLK 0% 100-year € 83 M 4% 350 Return Period 21% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 400 M 19% 3,000 Medium Return Period 101% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 60 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 6.9 Billion Euros* 1885 Belovodskoje M6.9 POPULATION 1992 Suusamyr M7.5 6.3 Million People* 2008 Nura M6.6 * 2018 estimate Buildin g D am age Lo s s es B i shk e k BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €3.0 0.48% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €2.5 0.40% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €2.0 0.32% frames that provide earthquake resistance; RC F is € 4 . 2 8 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €1.5 0.24%  This amounts to 0.06% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €1.0 0.16%  Near ly 3 8 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.5 0.08% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Bi s h k e k Bucharest Iași Belgrade Reinforced concrete frame and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Bishkek. Bratislava Ljubljana Re inforc e d Co n c rete Fra me (RCF) a n d U n re info rc e d Ma s on r y (URM ) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia  Bishkek ~1 4 3 , 0 0 0 p eop l e RC F 77% 61% 60% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 14% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Bi sh kek URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. Image credit: “Bishkek, Kyrgyzstan” by Ninara, licensed under CC BY 2.0 / Lowered opacity from original. 61 Osh KYRGYZ REPUBLIC Summ ary Osh, the second largest city in the Kyrgyz Republic, is Compared to the data for all the buildings in this study, considered the capital of the south and is home to over the data used to conduct the risk analysis for buildings in 4 percent of the total population. The Kyrgyz Republic Osh average a medium reliability; this rating is a function has faced several large earthquakes in the past, the most of the assumptions, uncertainties, and data available for damaging of which was the Suusamyr earthquake that the city (see Methodology). The diagram below indicates occurred in Toluk in 1992. This magnitude 7.5 event killed the reliability and confidence related to building stock 75 people, destroyed over 8,000 houses, and caused data (i.e., exposure data) as well as data for building damage of €217 million (2019 €) . More recently, in 2008, vulnerability. These confidence and reliability levels the country experienced a magnitude 6.6 earthquake should be considered when interpreting and applying the centered in the town of Nura. That earthquake killed 74 risk metrics presented in the city profile. Osh people and injured 140. *At risk building types based on earthquake loss are Based on the analysis, almost 45 percent of the determined by the expected average annual loss ratio, population of Osh resides in pre-2000 multifamily or average annual loss as a percentage of replacement residential buildings. The buildings that are the most value (AAL%). The expected loss results are predicted at risk* are unreinforced masonry (URM) buildings and from models of future events subject to variable data reinforced concrete frame (RCF) buildings. These two availability and modeling assumptions. The replacement building types are also expected to cause the most value is the construction cost of all multifamily buildings fatalities. Approximately 105,000 people, or about two in in the city. every five inhabitants in Osh, reside in high-risk building OSH FACTS types. Second Largest City Population: 257,000 (2017) GNI/capita: ~€1,000 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s € % RELIABILITY High EXPOSURE URM RCF LPB RCW Total building Building damage, Number of Medium 54% 1% 42% 3% damage % of total value fatalities Low DFW 0% BLK 0% SCS 0% 100-year € 28 M 10% 430 Return Period 4% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 95 M 35% 2,000 Medium Return Period 12% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 62 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 6.9 Billion Euros* 1885 Belovodskoje M6.9 POPULATION 1992 Suusamyr M7.5 6.3 Million People* 2008 Nura M6.6 * 2018 estimate Buildin g D am age Lo s s es Osh BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €1.2 1.2% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €1.0 1.0% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €0.8 0.8% frames that provide earthquake resistance; RC F is € 1 . 2 7 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €0.6 0.6%  This amounts to 0.02% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €0.4 0.4%  Near ly 4 4 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.2 0.2% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n O s h Bucharest Iași Belgrade Masonry and reinforced concrete frame are the two highest-risk building types used for pre-2000 multifamily housing in Osh. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d Re inforc e d Co n c rete Fra me (RC F) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~1 0 5 , 0 0 0 p eop l e URM 73% 78% 76%  Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 41% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Osh RC F n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 63 Dushanbe TA JIKISTAN Summ ary Dushanbe is Tajikistan’s capital and largest city, home Compared to the data for all the buildings in this study, to nearly 10 percent of the country’s total population. the data used to conduct the risk analysis for buildings Tajikistan has been affected by many catastrophic in Dushanbe average a medium reliability; this rating is earthquakes, the most deadly of which was the 1907 a function of the assumptions, uncertainties, and data magnitude 7.4 event in Karatag that led to the loss of available for the city (see Methodology). The diagram 12,000 lives. In 1985, a magnitude 5.9 earthquake near below indicates the reliability and confidence related to the Kayrakkum-Gafurov region caused 29 deaths, 80 building stock data (i.e., exposure data) as well as data injuries, and €428 million (2019 €) in damage. More for building vulnerability. These confidence and reliability recently, in 2015, a magnitude 7.2 earthquake located levels should be considered when interpreting and Dushanbe near Sarez Lake killed two, injured dozens, destroyed applying the risk metrics presented in the city profile. 500 homes, and resulted in nearly €5 million (2019 €) in damage. *At risk building types based on earthquake loss are determined by the expected average annual loss ratio, Based on the analysis, more than half of the population or average annual loss as a percentage of replacement of Dushanbe resides in pre-2000 multifamily residential value (AAL%). The expected loss results are predicted buildings. The buildings that are the most at risk* are slab- from models of future events subject to variable data column system (SCS) buildings and unreinforced masonry availability and modeling assumptions. The replacement (URM) buildings. Along with reinforced concrete frame value is the construction cost of all multifamily buildings (RCF) buildings, these building types are also expected to in the city. cause the most fatalities. Approximately 67,000 people, D USH ANBE FACTS or about 8 percent of the city’s population, reside in high- Capital City Population: 846,000 (2019) GNI/capita: ~€900 risk building types. Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High RCW € % RELIABILITY EXPOSURE 2% URM RCF LPB SCS Total building Building damage, Number of Medium 18% 1% 79% <1% damage % of total value fatalities Low DFW 0% BLK 0% 100-year € 81 M 10% 560 Return Period 21% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 273 M 33% 4,600 Medium Return Period 70% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 64 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 6.4 Billion Euros* 1907 Karatag M7.4 POPULATION 1985 Kayrakkum-Gafurov M5.9 9.1 Million People* 2015 Sarez M7.2 * 2018 estimate Buildin g D am age Lo s s es Du shanbe BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €1.8 1.50% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €1.5 1.25% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €1.2 1.00% frames that provide earthquake resistance; RC F is € 3 . 2 1 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €0.9 0.75%  This amounts to 0.05% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €0.6 0.50%  Near ly 5 4 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €0.3 0.25% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n D u s ha n b e Bucharest Iași Belgrade Slab-column system and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Dushanbe. Bratislava Ljubljana S la b - co lu mn Syste m (S CS ) a n d U n re info rc e d Ma s o n r y (U RM) bui l di ngs Almaty co nt r ib u te to : Shymkent Central Asia Bishkek ~6 7 , 0 0 0 peop l e SCS 40% 58% 39% Osh Dushanbe  live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 8% o f the to ta l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Dus h a n b e URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 65 Ashgabat TURKMENISTAN Summ ary Ashgabat is the capital and largest city of Turkmenistan, Compared to the data for all the buildings in this study, home to nearly 15 percent of the total population. the data used to conduct the risk analysis for buildings Turkmenistan has experienced devastating earthquakes in Ashgabat average the highest reliability; this rating is in the past, including a catastrophic 7.3 magnitude event a function of the assumptions, uncertainties, and data in 1948 centered in Ashgabat. The Ashgabat earthquake available for the city (see Methodology). The diagram killed 110,000 people, caused the collapse of many brick below indicates the reliability and confidence related to buildings, and resulted in over €240 million (2019 €) in building stock data (i.e., exposure data) as well as the damage. More recently, in 2000, a magnitude 7.0 event data for building vulnerability. These confidence and located near the city of Balkanabat led to the deaths of reliability levels should be considered when interpreting 11 individuals. and applying the risk metrics presented in the city profile. Ashgabat Based on the analysis, almost 40 percent of the *At risk building types based on earthquake loss are population of Ashgabat resides in pre-2000 multifamily determined by the expected average annual loss ratio, residential buildings. The buildings that are the most or average annual loss as a percentage of replacement at risk* are unreinforced masonry (URM) buildings and value (AAL%). The expected loss results are predicted reinforced concrete frame (RCF) buildings. These two from models of future events subject to variable data building types are also expected to cause the most availability and modeling assumptions. The replacement fatalities. Nearly 115,000 people, or about 14 percent of value is the construction cost of all multifamily buildings the city’s inhabitants, reside in high-risk building types. in the city. ASH G ABAT FACTS Capital City Population: 828,000 (2019) GNI/capita: ~€5,700 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High URM RCF LPB RCW € % RELIABILITY EXPOSURE 53% 2% 45% <1% Total building Building damage, Number of Medium damage % of total value fatalities Low DFW 0% BLK 0% SCS 0% 100-year € 100 M 10% 570 Return Period 6% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 600 M 22% 2,800 Medium Return Period 33% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 66 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 34.5 Billion Euros* 1946 Kazandzhik M6.9 POPULATION 1948 Ashgabat M7.3 5.9 Million People* 2000 Balkanabat M7.0 * 2018 estimate Buildin g D am age Lo s s es Ashgabat BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €6.0 0.6% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €5.0 0.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €4.0 0.4% frames that provide earthquake resistance; RC F is € 5 . 4 5 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €3.0 0.3%  This amounts to 0.02% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €2.0 0.2%  Near ly 3 9 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €1.0 0.1% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n A s h g a b a t Bucharest Iași Belgrade Masonry and reinforced concrete frame are the two highest-risk building types used for pre-2000 multifamily housing in Ashgabat. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d Re inforc e d Co n c rete Fra me (RC F) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~1 1 5 , 0 0 0 p eop l e URM 82% 91% 83% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n  Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 14% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Ashga b at RC F n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 67 Tashkent UZBEKISTAN Summ ary Tashkent is Uzbekistan’s capital and largest city, home over one in every five people in Tashkent, reside in high- to about 8 percent of the country’s total population. risk building types. Uzbekistan has been impacted by catastrophic earthquakes in the 20th century, the most deadly of Compared to the data for all the buildings in this study, which was the magnitude 6.4 Andijan earthquake in the data used to conduct the risk analysis for buildings 1902 that caused 5,000 deaths. In 1966, a magnitude 5.1 in Tashkent average a medium reliability; this rating is earthquake in Tashkent destroyed 80 percent of the city, a function of the assumptions, uncertainties, and data with total damage estimated at over €2 billion (2019 €). available for the city (see Methodology). The diagram Tashkent Masonry buildings were hard hit—many were damaged below indicates the reliability and confidence related to or destroyed. More recently, in 2011, the magnitude 6.1 building stock data (i.e., exposure data) as well as data Fergana Valley earthquake killed 14 people and injured for building vulnerability. These confidence and reliability 86 in Uzbekistan. After the 1966 earthquake, the very levels should be considered when interpreting and vulnerable unconfined masonry building type was no applying the risk metrics presented in the city profile. longer built. *At risk building types based on earthquake loss are Based on the analysis, over 60 percent of the population determined by the expected average annual loss ratio, of Tashkent resides in pre-2000 multifamily residential or average annual loss as a percentage of replacement buildings. The buildings that are the most at risk* are slab- value (AAL%). The expected loss results are predicted column system (SCS) buildings and unreinforced masonry from models of future events subject to variable data (URM) buildings. Along with reinforced concrete frame availability and modeling assumptions. The replacement TASH KENT FACTS (RCF) buildings, these building types are also expected to value is the construction cost of all multifamily buildings in the city. Capital City Population: 2,490,000 (2019) GNI/capita: ~€1,700 cause the most fatalities. Approximately 550,000 people, Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High RCW € % RELIABILITY EXPOSURE 2% Medium URM RCF LPB SCS Total building Building damage, Number of 67% 3% 28% <1% damage % of total value fatalities Low DFW 0% BLK 0% 100-year € 200 M 2% 100 Return Period 7% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 1.5 B 17% 3,000 Medium Return Period 56% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 68 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 42.7 Billion Euros* 1902 Andizhan M6.4 POPULATION 1966 Tashkent M5.1 33.0 Million People* 2011 Fergana Valley M6.1 * 2018 estimate Buildin g D am age Lo s s es Tashk e nt BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €9.0 0.36% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €7.5 0.30% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €6.0 0.24% frames that provide earthquake resistance; RC F is € 1 3 . 0 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €4.5 0.18%  This amounts to 0.03% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €3.0 0.12%  Near ly 6 2 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €1.5 0.06% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Tas hk e n t Bucharest Iași Belgrade Slab-column system and masonry are the two highest-risk building types used for pre-2000 multifamily housing in Tashkent. Bratislava Ljubljana S la b - co lu mn Syste m (S CS ) a n d U n re info rc e d Ma s o n r y (U RM) bui l di ngs Almaty co nt r ib u te to : Shymkent Central Asia Bishkek ~5 5 0 , 0 0 0 p eop l e SCS 54% 64% 43% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 22% o f the tota l d is p la c e me n t i n t h e  Tashkent Yerevan S. Caucasus po pul ati o n o f Tash kent URM n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 69 Yerevan ARMENIA Summ ary Yerevan is the capital and largest city of Armenia, home 250,000 people, or about one in every four inhabitants of to over 35 percent of the country’s total population. Yerevan, reside in high-risk building types. Armenia has faced large earthquakes in the past, the most catastrophic of which was the 1988 magnitude 6.8 Compared to the data for all the buildings in this study, earthquake in Spitak. This earthquake caused up to 25,000 the data used to conduct the risk analysis for buildings in deaths, injured 130,000 people, and left 500,000 people Yerevan average a high reliability; this rating is a function homeless, with damage totaling over €31 billion (2019 €). of the assumptions, uncertainties, and data available for It affected 40 percent of the population of Armenia, with the city (see Methodology). The diagram below indicates Yerevan nearly 95 percent of Spitak itself completely destroyed. the reliability and confidence related to building stock Precast reinforced concrete masonry buildings as well as data (i.e., exposure data) as well as data on building slab-column system buildings suffered severe damage vulnerability. These confidence and reliability levels or total collapse. The event significantly affected the should be considered when interpreting and applying the multifamily residential building stock and design practice, risk metrics presented in the city profile. and this research found that precast reinforced concrete *At risk building types based on earthquake loss are structures were no longer constructed in Gyumri and determined by the expected average annual loss ratio, Yerevan after 1988. or average annual loss as a percentage of replacement Based on the analysis, over 70 percent of the population value (AAL%). The expected loss results are predicted of Yerevan resides in pre-2000 multifamily residential from models of future events subject to variable data buildings. The buildings that are the most at risk* are availability and modeling assumptions. The replacement YEREVAN FACTS unreinforced masonry (URM) buildings and slab-column value is the construction cost of all multifamily buildings in the city. Capital City Population: 1,093,000 (2019) GNI/capita: ~€3,600 system (SCS) buildings. These two building types are also expected to cause the most fatalities. Approximately Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s URM RCF LPB SCS € % RELIABILITY High EXPOSURE 48% 22% 28% 2% Total building Building damage, Number of Medium damage % of total value fatalities Low DFW 0% BLK 0% RCW 0% 100-year € 1.8 B 16% 3,000 Return Period 64% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 5.0 B 46% 10,600 Medium Return Period 180% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 70 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 10.5 Billion Euros* 1679 Garnì M7.0 POPULATION 1931 Zangezur M6.3 3.0 Million People* 1988 Spitak M6.8 * 2018 estimate Buildin g D am age Lo s s es Y e re van BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €54 1.50% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €45 1.25% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €36 1.00% frames that provide earthquake resistance; RC F is € 7 2 . 5 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €27 0.75%  This amounts to 0.69% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €18 0.50%  Near ly 7 1 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €9 0.25% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Y er ev a n Bucharest Iași Belgrade Masonry and slab-column system are the two highest-risk building types used for pre-2000 multifamily housing in Yerevan. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d S la b - co lu mn Syste m (S CS ) b ui l di ngs Almaty co nt r ib u te to : Shymkent Central Asia Bishkek ~2 5 0 , 0 0 0 p eop l e URM 77% 75% 77% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 23% o f the tota l d is p la c e me n t i n t h e  Tashkent Yerevan S. Caucasus po pul ati o n o f Ye reva n SCS n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 71 Gyumri ARMENIA Summ ary Gyumri is Armenia’s second largest city and home to 4 Approximately 43,000 people, or about 40 percent of the percent of the country’s total population. Armenia has inhabitants of Gyumri, reside in high-risk building types. faced large earthquakes in the past, the most catastrophic of which was the 1988 magnitude 6.8 earthquake in Compared to the data for all the buildings in this study, Gyumri Spitak. This earthquake caused up to 25,000 deaths, the data used to conduct the risk analysis for buildings in injured 130,000 people, and left 500,000 people Gyumri average a high reliability; this rating is a function homeless, with damage totaling over €31 billion (2019 €). of the assumptions, uncertainties, and data available for It affected 40 percent of the population of Armenia, with the city (see Methodology). The diagram below indicates nearly 95 percent of Spitak itself completely destroyed. the reliability and confidence related to building stock Precast reinforced concrete masonry buildings as well as data (i.e., exposure data) as well as data for building slab-column system buildings suffered severe damage vulnerability. These confidence and reliability levels or total collapse. The event significantly affected the should be considered when interpreting and applying the multifamily residential building stock and design practice, risk metrics presented in the city profile. and this research found that precast reinforced concrete *At risk building types based on earthquake loss are structures were no longer constructed in Gyumri and determined by the expected average annual loss ratio, Yerevan after 1988. or average annual loss as a percentage of replacement Based on the analysis, almost 40 percent of the population value (AAL%). The expected loss results are predicted of Gyumri resides in pre-2000 multifamily residential from models of future events subject to variable data buildings. The buildings that are the most at risk* are availability and modeling assumptions. The replacement G YUMRI FACTS unreinforced masonry (URM) buildings and reinforced value is the construction cost of all multifamily buildings in the city. Second Largest City Population: 117,000 (2019) GNI/capita: ~€3,600 concrete frame (RCF) buildings. These two building types are also expected to cause the most fatalities. Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s RCF LPB € % RELIABILITY High EXPOSURE 1% 3% URM Total building Building damage, Number of Medium 96% damage % of total value fatalities Low DFW 0% BLK 0% SCS 0% RCW 0% 100-year € 242 M 21% 320 Return Period 147% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 755 M 65% 1,300 Medium Return Period 460% GNIPC people High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 72 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 10.5 Billion Euros* 1679 Garnì M7.0 POPULATION 1931 Zangezur M6.3 3.0 Million People* 1988 Spitak M6.8 * 2018 estimate Buildin g D am age Lo s s es Gy u mri BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €9.0 0.90% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €7.5 0.75% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €6.0 0.60% frames that provide earthquake resistance; RC F is € 8 . 1 7 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €4.5 0.45%  This amounts to 0.08% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €3.0 0.30%  Near ly 3 9 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €1.5 0.15% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n G y u m r i Bucharest Iași Belgrade Masonry and reinforced concrete frame are the two highest-risk building types used for pre-2000 multifamily housing in Gyumri. Bratislava Ljubljana U n re info rc e d Ma s o n r y (U RM) a n d Re inforc e d Co n c rete Fra me (RC F) Almaty b u ild in g s co nt r ib u te to : Shymkent Central Asia Bishkek ~4 3 , 0 0 0 peop l e URM 98% 92% 95% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 37% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Gy um r i RC F n ex t 5 0 ye a rs f ro m  Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. Image credit: “Holy Saviour Church Gyumri” by Armenak Margarian, licensed under CC BY 4.0 / Lowered opacity from original. 73 Baku AZERBAIJAN Summ ary Baku is Azerbaijan’s capital and largest city, home to Compared to the data for all the buildings in this study, nearly a quarter of the country’s total population. the data used to conduct the risk analysis for buildings Azerbaijan has experienced many severe earthquakes in in Baku average a high reliability; this rating is a function the past, including the damaging 2000 Baku earthquake, of the assumptions, uncertainties, and data available for a magnitude 6.8 event in which 31 people lost their the city (see Methodology). The diagram below indicates lives and 430 were injured. In 1999, a magnitude 5.4 the reliability and confidence related to building stock earthquake in the Agdas-Ucar-Ağalı area led to one data (i.e., exposure data) as well as data on building death, 18 injuries, and €7 million (2019 €) in damage. vulnerability. These confidence and reliability levels Baku should be considered when interpreting and applying the Based on the analysis, almost half of the population risk metrics presented in the city profile. of Baku resides in pre-2000 multifamily residential buildings. The buildings that are the most at risk* are *At risk building types based on earthquake loss are reinforced block (BLK) buildings and slab-column system determined by the expected average annual loss ratio, (SCS) buildings. Along with, unreinforced masonry (URM) or average annual loss as a percentage of replacement buildings, these building types are also expected to cause value (AAL%). The expected loss results are predicted the most fatalities. Approximately 125,000 people, or from models of future events subject to variable data about five percent of the inhabitants of Baku, reside in availability and modeling assumptions. The replacement high-risk building types. value is the construction cost of all multifamily buildings in the city. BAKU FACTS Capital City Population: 2,313,000 (2019) GNI/capita: ~€3,400 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High URM RCF LPB BLK SCS RCW € % RELIABILITY EXPOSURE 49% 4% 29% 4% 4% 10% Total building Building damage, Number of Medium damage % of total value fatalities Low DFW 100-year € 1.0 B 10% 2,800 0% Return Period 27% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 3.5 B 34% 11,400 Return Period 93% GNIPC people Medium High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 74 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 40.0 Billion Euros* 1999 Agdas M5.4 POPULATION 2000 Baku M6.8 10.0 Million People* 2012 Zaqatala M5.7 * 2018 estimate Buildin g D am age Lo s s es B ak u BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €24 3.0% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €20 2.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €16 2.0% frames that provide earthquake resistance; RC F is € 4 2 . 4 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €12 1.5%  This amounts to 0.11% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €8 1.0%  Near ly 4 8 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €4 0.5% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Baku Bucharest Iași Belgrade Block and slab-column system are the two highest-risk building types used for pre-2000 multifamily housing in Baku. Bratislava Ljubljana Blo c k ( BL K ) a n d S la b - colu mn Syste m (S CS ) b u ild in g s co nt r ib u te to : Almaty Shymkent Central Asia Bishkek ~1 2 5 , 0 0 0 p eop l e BLK 31% 70% 49% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 5% o f the to ta l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Ba ku SCS n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e  Baku to ea r t h q u a ke s Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 75 Tbilisi GEORGIA Summ ary Tbilisi is Georgia’s capital and largest city, home to over Compared to the data for all the buildings in this study, a quarter of the country’s total population. Georgia has the data used to conduct the risk analysis for buildings been affected by many earthquakes of varying severity in Tbilisi average a low reliability; this rating is a function in the past, the most notable and damaging of which was of the assumptions, uncertainties, and data available for the magnitude 7.0 Racha earthquake in 1991—the most the city (see Methodology). The diagram below indicates powerful earthquake ever recorded in the Caucasus. the reliability and confidence related to building stock The event caused 270 deaths, injured 1,000 people, left data (i.e., exposure data) as well as data on building 67,000 people homeless, and caused nearly €3 billion vulnerability. These confidence and reliability levels (2019 €) in damage. In 2002, a magnitude 4.3 earthquake should be considered when interpreting and applying the in Tbilisi killed five, injured over 50, and damaged or risk metrics presented in the city profile. destroyed 2,400 buildings. Tbilisi *At risk building types based on earthquake loss are Based on the analysis, over 80 percent of the population determined by the expected average annual loss ratio, of Tbilisi resides in pre-2000 multifamily residential or average annual loss as a percentage of replacement buildings. The buildings that are the most at risk* are value (AAL%). The expected loss results are predicted block (BLK) buildings and slab-column system (SCS) from models of future events subject to variable data buildings. Along with unreinforced masonry (URM) availability and modeling assumptions. The replacement buildings, these building types are also expected to cause value is the construction cost of all multifamily buildings the most fatalities. Approximately 300,000 people, or in the city. nearly 30 percent of the population of Tbilisi, reside in TBILISI FACTS high-risk building types. Capital City Population: 1,077,000 (2019) GNI/capita: ~€3,800 Data Relia b i l i t y an d Bu i l d i ng E xp osu r e E a r t h q uak e E ve nt-B ase d L o sse s High € % RELIABILITY EXPOSURE RCW Total building Building damage, Number of Medium damage % of total value fatalities 10% Low URM RCF LPB BLK SCS 9% 9% DFW 40% 30% 2% 100-year € 1.4 B 21% 6,600 0% Return Period 43% GNIPC people VULNERABILITY RELIABILITY Low 475-year € 3.2 B 48% 19,500 Return Period 97% GNIPC people Medium High  A 100-year return period event is one where the damage and loss have approximately a 40% probability of being exceeded in 50 Building exposure data is for pre-2000 multifamily buildings. years; it is considered a more frequent event. A 475-year return period event is one where the damage and loss have approximately a  Data on building stock and population (exposure) and their susceptibility to earthquake losses 10% probability of being exceeded in 50 years; it is considered a less frequent event, more damaging event. (vulnerability) for the city is conveyed with vertical bars labeled by the building typology acronym.  GNIPC is gross national income per capita. Percentage of GNIPC is the losses from damage per resident as a percentage of GNIPC.  The diagram also indicates the percentage of total building stock that a typology comprises  The fatalities presented are those that are expected to occur at nighttime (2:00 a.m.) on a weekend. in the city denoted as a percentage below the building typology acronym. Highlighted building  Note: Event-based losses are subject to limitation by the reliability of the input data. The chosen return periods do not suggest that types are those that are considered highest risk in the city. more hazardous and infrequent earthquake events are not probable in this city, as higher return period events may occur at any point. 76 Country Snapshot GDP SIGNIFICANT EARTHQUAKES Earthquake Risk Profile 15.0 Billion Euros* 1991 Racha M7.0 POPULATION 2002 Tbilisi M4.3 3.7 Million People* 2009 Oni M6.0 * 2018 estimate Buildin g D am age Lo s s es Tbi li si BUILDING CLASSIFICATION & VULNERABILITY RANGE ABSOLUTE AND RELATIVE LOSS EARTHQUAKE IMPACTS Description Type Vulnerability €48 3.0% SNAPSHOT PERCENTAGE OF REPLACEMENT VALUE AVERAGE ANNUAL LOSS (€, millions) Lo. Med. Hi. Unreinforced Masonry: built with masonry walls that lack steel reinforcement; includes confined URM unconfined  Average annual loss €40 2.5% from ear thquake damage AVERAGE ANNUAL LOSS AS and unconfined walls. confined Reinforced Concrete Frame: reinforced concrete to multifamily residences precast €32 2.0% frames that provide earthquake resistance; RC F is € 5 9 . 5 m i l l i o n . includes precast and cast-in-place frames. cast-in-place Dual Frame-wall: reinforced concrete frame and €24 1.5%  This amounts to 0.40% wall elements work together to resist earthquake DF W of the countr y ’s GDP. forces. Precast Large Panel: built with precast reinforced €16 1.0%  Near ly 8 2 % of the concrete panels with floor-to-ceiling height facade LP B city ’s population lives in panels. multifamily residential €8 0.5% Block: built with precast concrete elements buildings. assembled on site; similar to LPB with smaller BLK facade elements. €0 0.0% LOSS COMPARISON Slab-column System: reinforced concrete floor slabs directly connected to concrete columns; SCS lift-slab UR M RC F DFW LPB BLK SCS RC W TO REGION includes lift slabs and post-stressed slabs. post-stressed AAL% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% Reinforced Concrete Wall: cast-in-place walls that Tirana sliding provide earthquake resistance; includes standard, RC W Building Type Durrës sliding, and tunnel formwork. tunnel & Mostar standard Average annual loss Average annual loss as a percent Sofia Note that the relative vulnerabilities are approximate and depend on post- Central Europe & Balkans (€, millions) of replacement value (AAL%) Plovdiv construction modifications, degradation, construction quality, and year of Zagreb construction. See Types of Multifamily Buildings in ECA for more information. Rijeka Budapest Chişinău Podgorica Skopje High -R is k Bu i l d i n gs i n Tb i l i si Bucharest Iași Belgrade Block and slab-column system are the two highest-risk building types used for pre-2000 multifamily housing in Tbilisi. Bratislava Ljubljana Blo c k ( BL K ) a n d S la b - co lu mn Syste m (S CS ) b u ild in g s co nt r ib u te to : Almaty Shymkent Central Asia Bishkek ~3 0 0 , 0 0 0 p eop l e BLK 80% 64% 55% Osh Dushanbe live i n these bui l d i n g s , o f to ta l lo s s f ro m o f a l l p e r ma n e nt o f fata lit ie s i n Ashgabat b u i l d i n g d a m a ge re s id e nt ia l m u l t i fa m i l y buildings* or 28% o f the tota l d is p la c e me n t i n t h e Tashkent Yerevan S. Caucasus po pul ati o n o f Tbi li s i SCS n ex t 5 0 ye a rs f ro m Gyumri b u i l d i n g d a m a ge d u e Baku to ea r t h q u a ke s  Tbilisi AAL% Central Europe & Balkans (0.24%) *This is an average number of fatalities AAL% Central Asia (0.25%) when assessed over many years. AAL% South Caucasus (0.62%) References: See City Profile Sources for information on building types, classification, and exposure data sources. 77 FUTURE W O RK The current study can be used as the basis for more •  Expansion of the study. The sole focus of this study detailed risk assessment of the 27 cities presented. was pre-2000 multifamily residential buildings, as these Several areas of further research include: house a large percentage of the population within all cities in the region. The study could expand to consider •  Development and refinement of the asset data. the entirety of the residential building stock in these Researching and compiling accurate data on the number cities. In addition, investigating other cities in the ECA of buildings, dwellings, and occupants (exposure) region could provide a more holistic understanding of per structural group, as well as information on the earthquake risk. current condition of buildings that may influence their vulnerability, will improve the reliability of the exposure • Inclusion of socioeconomic factors. Future work should models used in this study and increase the accuracy of also include aspects of demographics, poverty, education, the assessment results. and human vulnerability to comprehensively understand the indirect losses from earthquakes. The socioeconomic •  Development and refinement of the damage and layer provides an understanding of critical rehabilitation loss data. Additional research on and validation of the and recovery factors that are not captured by building damage and loss (vulnerability) models for the various damage consequences. • building types could increase the accuracy of the assessment results. Image Credit: “Looking out over Bishkek, Kyrgyzstan [2011]” by United Nations Development Programme in Europe and Central Asia, licensed under CC BY-NC-SA 2.0. 78 79 WOR K S CIT E D Abazaj, A. 2019. “Prefabrication and Modular Warning System for Armenia.” In Zschau, Jochen, and Colliers International. 2014 “Georgia - Residential Construction Dwellings in Albania.” International Journal Andreas N. Küppers (Eds.), Early Warning Systems for Market Report.” n.p. of Scientific & Engineering Research 10, no. 8: 811–819. Natural Disaster Reduction, page(s): 487–494. Springer- Verlag Berlin Heidelberg. doi: 10.1007/978-3-642- Croatian Design Society. “Bogdan Budimirov.” http:// Abolmasov, B., Jovanovski, M., Ferić, P., and Mihalić, 55903-7. dizajn.hr/blog/bogdan-budimirov-1928-2019/#. S. 2011. “Losses Due to Historical Earthquakes in the Balkan Region: Overview of Publicly Available Data.” Borozan, J. 2019a. Exposure Model for Residential Croatian Professional Firefighters Association. http:// Geofizika 28, no. 1: 161–81. Building Stock of Serbia, Personal Correspondence. www.upvh.hr/. October 2019: Serbian Association for Earthquake Crowley, H., Rodrigues, E. D., and Despotaki, E. V. 2018. Ademović, N. and Hadzima-Nyarko, M. 2018. “Seismic Engineering (SAEE/SUZI). Vulnerability, Damage and Strengthening of Masonry “D26.2 Methods for Developing European Residential Structures in the Balkans with a Focus on Bosnia and Borozan, J. 2019b. “Exposure Model for Serbia.” On Exposure Models.” n.p.: Eucentre. Herzegovina.” Thessaloniki: n.p. Behalf of the SUZI-SAEE Working Group for Seismic Risk, Csoknyai, T., Hrabovszky-Horváth, S., Seprődi-Egeresi, Belgrade: SERA workshop. M. and Szendrő, G. 2014. “Lakóépület Tipológia Ademović, N. 2019. “Seismic Behavior of Typical Masonry Buildings from Bosnia and Herzegovina.” Botici, A. A., Ungureanu, V., Ciutina, A., Botici, A., Magyarországon. National Typology Of Residential Belgrade: SERA Workshop. Dubinä, D., and Fülöp, L. 2013. “Sustainable Retrofitting Buildings in Hungary.” n.p.: Episcope, Tabula. Solutions for Precast Concrete Residential Buildings: Cutia, E. 2018. “Characterization of the Seismic Zone of Agency for Statistics of Bosnia and Herzegovina (BHAS). Architectural and Structural Aspects.” Romanian-Finnish http://www.bhas.ba/ Chişinău.” Buletinul Institutului Politehnic din lasi. Sectia Seminar on Opportunities in Sustainably Retrofitting the Constructii, Arhitectura 64, no. 2: 17–23. Agency for Statistics Under President of The Republic of Large Panel Reinforced Concrete Building Stock, page(s): Tajikistan. 2020. https://www.stat.tj/en. 23–42. Department of Homeland Security, Federal Emergency Management Agency. n.d. Hazus® - MH 2.1 Technical Alcaz, V., Isicico, E., and Ghinsari, V. 2011. “Riscul Seismic Bouzarovski, S., Salukvadze, J., and Gentile, M. 2011. Manual. Washington, D.C.: Department of Homeland în Teritoriul Oraşului Chişinău.” Akademos Seismologie 4, “A Socially Resilient Urban Transition? The Contested Security, Federal Emergency Management Agency. no. 23: 80–85. Landscapes of Apartment Building Extensions in Two Post-Communist Cities.” Urban Studies 48, no. 13: Dolšek, M. 2019. “Research Related to Seismic Risk in Alfirević, Đ. and Alfirević, S. 2015. “Urban Housing 2689–2714. Slovenia.” n.p.: SERA workshop. Experiments in Yugoslavia 1948-1970.” Spatium 1, no. 34: 1–9. Carydis, P. G., and Vougioukas, E. A. 1988. “The Tirana, Duzs. 2013. “Procjena Ugroţenosti Republike Hrvatske Albania, Earthquake of January 9, 1988.” n.p.: EERI Od Prirodnih i Tehniĉko-Tehnoloških Katastrofa i Velikih Aliaj, B. 2019. “Housing Models in Albania between Special Earthquake Report. Nesreća.” Zagreb: n.p. 1945-1999.” ENHR/MRI Conference, Athens: n.p. Central Bureau of Statistics, Republic of Croatia. 2015. Emporis. 2020 “Tallest buildings in Rijeka.” https://www. Aliaj, S., Koçiu, S., Muço, B., and Sulstarova, E. 2010. https://www.dzs.hr/. emporis.com/statistics/tallest-buildings/city/100717/ “Seismicity, Seismotectonics, and Seismic Hazard rijeka-croatia. Assessment in Albania.” Tirana: Albanian Academy of Central Census Bureau of Bosnia and Herzegovina. 2013. Sciences. “Census - Dwellings, Housing, and Housing Conditions.” Episcope, Tabula. “BA Bosnia and Herzegovina - Country Agency for Statistics of Bosnia and Herzegovina. https:// Page.” https://episcope.eu/building-typology/country/ Arnautović-Aksić, D., Burazor, M., Delalić, N., Gajić, D., www.popis.gov.ba/popis2013/knjige.php?id=5. ba/. Gvero, P., Kadrić, D., and Salihović, E. 2016. “Typology of Residential Buildings in Bosnia and Herzegovina.” Central Statistical Office, Budapest. https://www.ksh. European Union Building Database. https://ec.europa. Sarajevo: n.p. hu/. eu/energy/en/eu-buildings-database. Atalić, J., Novak, M., and M, U. 2019. “Development of City Kvira. 2018. “Soviet Buildings in Tbilisi - Interesting Folić, R., Laban, M., and Milanko, V. 2011. “Reliability Exposure Model for Croatia.” SERA Balkans Seismic Risk Statistics about Their Number.” http://www.city.kvira. and Sustainability Analysis of Large Panel Residential Workshop, Belgrade. ge/. Buildings in Sofia, Skopje and Novi Sad.” Facta universitatis-series: Architecture and Civil Engineering 9, Balassanian, S.Y., Martirossyan, A., Arzoumanian, V. City of Ljubljana. https://www.ljubljana.si/en. no. 1: 161–176. 2003. “Project of Creation of an Earthquake Early City of Rijeka. Rijeka.hr/en/. 80 W O RKS CITE D (C O N TI N UE D ) Fülöp, L., Ungureanu, V., Ott, W., Bolliger, R., and “National Building Typology as A Source for an Adequate in New Belgrade.” Znanstveni Èasopis Za Arhitekturu I Jakob, M. 2013. “Cost Effectiveness of Energy Retrofit Rehabilitation Policy.” Contemporary Materials 2, no. 4: Urbanizam 1, no. 49. Solutions: Results from Generic Calculations with a 137–144. Reference Building in Romania.” Proceedings from Ministry of National Development (MND), Non-Profit Romanian-Finnish Seminar on Opportunities in Infrastructure Department ECA. 2006. “Multi-Apartment Limited Liability Company for Quality Control and Sustainably Retrofitting the Large Panel Reinforced Housing in Azerbaijan: Issues Note.” Housing and Innovation in Building (ÉMI). 2015. “National Building Concrete Building Stock, page(s): 93–104. Communal Services in the South Caucasus. Energy Performance Strategy.” Budapest: MNS/ÉMI. Georgescu, E.S. and Pomonis, A 2008. “The Romanian Institute of Statistics, Tirana, Albania. http://www.instat. Ministry of National Economy of the Republic of Earthquake of March 4, 1977 Revisited: New Insights gov.al/en. Kazakhstan Statistics Committee. 2020. https://stat.gov. into its Territorial, Economic and Social Impacts and kz/. Institutul Național de Statistică (INSSE). 2016. “Populaţia Their Bearing on the Preparedness for the Future.” României pe localitati.” INSSE. https://insse.ro/cms/ Mott MacDonald. 2019. “Component 2—Exposure and In Proceedings of the 14th World Conference on sites/default/files/field/publicatii/populatia_romaniei_ Vulnerability Data. Seismic PRA of pre-1990 Multi- Earthquake Engineering, Beijing, page(s): 12–17. https:// pe_localitati_la_1ianuarie2016_1.pdf. family Structures in Bulgaria and ECA Region (Selection www.iitk.ac.in/nicee/wcee/article/14_10-0013.PDF. #1249310).” May 2019. Japan International Cooperation Agency (JICA). 2009. Georgescu, E.S. and Pomonis, A. 2012. “Building “Study on Earthquake Risk Management in the City of Mott MacDonald. 2020a. “Seismic Risk in Multi-family Damage vs. Territorial Casualty Patterns During the Almaty, Republi of Kazkhstan, Final Report.” JICA, Oyo Buildings in the Europe Central Asia Region: Technical Vrancea (Romania) Earthquakes of 1940 and 1977.” International Corporation, Nippon Koei Co., LTD., Aero Report 1.” June 3, 2020. In Proceedings of the 15th World Conference on Asahi Corporation. Earthquake Engineering, page(s): 24–28. http://www. Mott MacDonald. 2020b. “Seismic Risk in Multi-family iitk.ac.in/nicee/wcee/article/WCEE2012_2123.pdf. Japan International Cooperation Agency (JICA). 2012. Buildings in the Europe Central Asia Region—Seismic “The Project for Seismic Risk Assessment and Risk Risk Analysis: Technical Report 2.” June 3, 2020. Giardini, D., Danciu, L., Erdik, M., Şeşetyan, K., Tümsa, Management Planning in the Republic of Armenia.” M.B.D., Akkar, S., Gülen, L. and Zare, M. 2018. “Seismic JICA, Volume: 1. https://libopac.jica.go.jp/images/ Mrduljaš, M., and Vladimir K. eds. 2012. Unfinished Hazard Map of the Middle East.” Bulletin of Earthquake report/1000008242.pdf. Modernisations: Between Utopia and Pragmatism: Engineering 16, no. 8: 3567–3570. Architecture and Urban Planning in the Former Kalman Šipoš, T. and Hadzima-Nyarko, M. 2018. “Seismic Yugoslavia and the Successor States. UHA/CCA. Government of Croatia. 2020. “The Croatia Earthquake – Risk of Croatian Cities Based on Building’s Vulnerability.” Rapid Damage and Needs Assessment 2020.” Tehnički Vjesnik 25, no.4: 1088-1094. Muntean, D. and Ungureanu, V. 2017. “Adaptive Design of Large Prefabricated Concrete Panels Collective Horváth, S. E., Szalay. Z. 2012. “Decision-Making Case King, S., Khalturin, V., and Tucker, B., eds. 1996. Seismic Housing.” n.p.: International Conference on Building, Study for Retrofit of High-Rise Concrete Buildings Hazard and Building Vulnerability in Post-Soviet Central Architecture and Urbanism, page(s): 81—90. Based on Life Cycle Assessment Scenarios.” n.p.: Asian Republics. Springer Science and Business Media, International Symposium on Life Cycle Assessment and Volume: 52. Muntean, D., Ungureanu, V., Petran, I. and Georgescu, Construction—Civil Engineering and Buildings, page(s): M. 2017. “Large Prefabricated Concrete Panels Collective 116–124. Korak. “Remetinečki gaj in Novi Zagreb.” 2014. https:// Dwellings from the 1970s: Context and Improvements.” korak.com.hr/. September 30, 2014. IOP Conference Series: Materials Science and Hrabovszky-Horváth, S. 2015. “The Climate Strategy Engineering 245, no. 5. Aspects of the Energy Efficient Refurbishment of Precast Mathema, A., and Simpson, A. 2018. “On Shaky Ground: Concrete Buildings – Summary of Phd Dissertation.” n.p.: Housing in Europe and Central Asia.” August 23, 2018. National Bureau of Statistics, Moldova. http://statbank. BUTE, Department of Building Constructions. World Bank Blogs: Eurasian Perspectives. statistica.md/. Ibragimov, R.S., Nurtaev, B.S. and Khakimov Sh., A. 2000. Mavlyanova, N., Inagamov, R., Rakhmatullaev, H., and National Geophysical Data Center/World Data Service “Methodology of Assessment and Mitigation of Urban Tolipova, N. 2004. “Seismic Code of Uzbekistan.” In 13th (NGDC/WDS). NCEI/WDS Global Significant Earthquake Seismic Risk, Taking into Account Variability of Seismic World Conference on Earthquake Engineering, Volume: Database. NOAA National Centers for Environmental Hazard Parameters.” Proceedings of 12WCEE, Aukland, 1611. Information. doi:10.7289/V5TD9V7K. New Zealand, paper 1009. Mecanov, D. 2015. “Prefabricated Construction System National Geospatial Data Fund, Moldova. https:// Ignjatović, D., Ignjatović, N.Ć. and Popović, M.J. 2013. Jugomont from Zagreb ‘Horseshoe’ Building in Block 28 geoportal.md/. 81 WOR K S CIT E D (C O N TI N UE D ) National Institute of Statistics, Romania. 2020. http:// Dushanbe by the Use of Ambient Vibration Analysis.” of Large Housing Estates in Budapest - For Sustainable www.insse.ro/cms/en/. Annals of Geophysics 58, no. 1: 1–13. Development.” n.p. National Statistical Committee of the Kyrgyz Republic. Petrovic, G., IMS Institute. n.d. “IMS Building Samoilov, K. 2004. Architecture of The Kazakstan of the 2015. http://stat.kg/en/. Technology - Precast Prestressed Concrete Skeleton in 20th- Century: The Development of Architectural-Artistic Contemporary Buildings.” n.p. Forms. Moscow-Almaty: M-ARi’design. National Statistics Office of Georgia. 2013. “General Population Census 2014.” http://census.ge/. Petrovski, J. T. 2004. “Damaging Effects of July 26, 1963 Shendova, V., Apostolska, R. and Vitanova, M. 2019. Skopje Earthquake.” Middle East Seismological Forum, “Structural Classification of Building and Bridge Assets in National Statistics Office of Georgia. 2020. https://www. Cyber Journal of Geoscience 2. R.N Macedonia.” Belgrade: SERA workshop. geostat.ge/en/. PopulationStat. 2020. https://populationstat.com/ Simaku, G. 2014. “Albania - National Building Typology, Nikolic, J. 2018. “Building ‘With the Systems’ vs. Building azerbaijan/baku. Energy Performance and Saving Potential.” n.p. ‘in The System’ of IMS Open Technology of Prefabricated Construction: Challenges for New ‘Infill’ Industry for Prosinečki, J. 2015. “Stambena Kriza i industrijalizacija Simaku, G. 2017. “Albanian Building Stock Typology Massive Housing Retrofitting. Energies 11, no. 5. građevinarstva u Jugoslaviji. Montažna gradnja u and Energy Building Code in Progress Towards National Zagrebu 1950-ih i 1960-ih – montažni sistemi poduzeća Calculation Methodology of Performance on Heating Nikolic, J. 2016. “Refurbishment scenarios for post-war Jugomont.” Diplomski Rad, Zagreb: Sveučilište u and Cooling.” European Journal of Multidisciplinary industrialized housing in Beograd.” PhD Thesis. UPC, Zagrebu. Studies 2, no.5: 13–35. Departament de Tecnologia de l’Arquitectura. Available at: http://hdl.handle.net/2117/96268. Rakušček, A., Zavrl, M. and Stegnar, G. 2012. “IEE Socioeconomic Data and Applications Center (SEDAC). TABULA - Typology Approach for Building Stock Energy 2020. “Population Density Grid, v1 (1990, 1995, 2000).” Open Data Bratislava. 2019. “Infrastructure, Assessment.” National Scientific Report - Slovenia, Center for International Earth Science Information Construction and Housing. https://opendata.bratislava. Ljubljana: Gradbeni Institut ZRMK. Network, Earth Institute, Columbia University. https:// sk/dataset/category/infrastruktura/. sedac.ciesin.columbia.edu/data/set/grump-v1- Regional Environmental Center, Austrian Development population-density/maps/2?facets=region:europe/ OpenDRI. 2020. “Kyrgyz Republic.” Data Sharing Corporation. 2015. “The Typology of the Residential Platform: http://geonode.mes.kg/. Building Stock in Albania and the Modelling of its Low- Stanfield, D., Childress, M., Dervishi, A. and Korra, L. Радовановиh, С. and Петрониjевиh, М. 2009. “Building Carbon Transformation – Albania.” Support for low- 1999. “Emerging Real Estate Markets in Metropolitan Types and Vulnerability to Ground Shaking in Serbia.” emission development in South Eastern Europe, n.p. Tirana, Albania.” Land Tenure Center, University of Banja Luka, Oct. 2009, Proceedings from International Wisconsin–Madison. Regional Environmental Center (REC), Austrian Conference on Earthquake Engineering. Development Corporation (ADC). 2015. “The Typology State Committee on Statistics of Turkmenistan. 2020. Pagani, M., J. Garcia-Pelaez, R. Gee, K. Johnson, V. Poggi, of the Residential Building Stock in Montenegro and http://www.stat.gov.tm/. R. Styron, G. Weatherill, M. Simionato, D. Viganò, L. the Modelling of its Low-Carbon Transformation – Montenegro.” Support for low-emission development in State Statistical Committee of the Republic of Danciu, and D. Monelli. 2018. “Global Earthquake Model Azerbaijan. 2019a. “Construction in Azerbaijan - Seismic Hazard Map.” Version 2018.8. December 2018 South Eastern Europe, n.p. Statistical Yearbook.” Baku: SSC. (accessed May 11, 2020). https://maps.openquake.org/ Republic of Slovenia, Statistical Office. https://pxweb. map/global-seismic-hazard-map/. stat.si/SiStatDb/pxweb/en/10_Dem_soc/ State Statistical Committee of the Republic of Azerbaijan. 2019b. “Demographic Indicators in Pavel, F., Vacareanu, R., Douglas, J., Radulian, M., Regional Environmental Center (REC), Austrian Azerbaijan - Statistical Yearbook.” Baku: SSC. Cioflan, C., and Barbat, A. 2016. “An Updated Development Corporation (ADC). 2015. “The Typology Probabilistic Seismic Hazard Assessment for Romania of the Residential Building Stock in Serbia and the State Statistical Office, Republic of North Macedonia. and Comparison with the Approach and Outcomes of Modelling of its Low-Carbon Transformation – Serbia.” http://www.stat.gov.mk/. the SHARE project.” Pure and Applied Geophysics 173, Support for Low-Emission Development in South Eastern no. 6: 1881–1905. Statistical Office of Montenegro. https://www.monstat. Europe, n.p. org/. Petrovic, B., Bindi, D., Pilz, M., Serio, M., Orunbaev, S., RS Architects. 2006. “How We Really Live In Panel Niyazov, J., Hakimov, F., Yasunov, P., Begaliev, U., and Statistical Office of the Republic of Serbia. 2020. https:// Blocks. Case Study on the Conditions and Potentials www.stat.gov.rs/en-US/. Parolai, S. 2015. “Building Monitoring in Bishkek and 82 W O RKS CITE D (C O N TI N UE D ) Stoychev, G. 1976. Panel Residential Buildings: United Nations Economic Commission for Europe WHE (World Housing Encyclopedia). 2003. “Medium/ Architectural and Design Solutions. State Publishing (UNECE). https://www.unece.org High Rise Moment Resisting Reinforced Concrete Frame House Slovakia. Buildings, Romania: Report 97.” Earthquake Engineering United Nations Economic Commission for Europe Research Institute and International Association for Statistical Committee of the Republic of Armenia. (UNECE). 2010. “Country Profiles on the Housing Sector: Earthquake Engineering. http://db.world-housing.net/ 2011. “The Results of the 2011 Population Census Azerbaijan.” United Nations, New York and Geneva. building/97/. of the Republic of Armenia.” https://armstat.am/ en/?nid=82&id=1512/. United Nations Economic Commission for Europe WHE (World Housing Encyclopedia). n.d. “Block of Flats (UNECE). 2011. “Country Profiles on the Housing Sector: with 11 Floors Out of Cast-in-Situ Concrete, Gliding Statistical Committee of the Republic of Armenia. 2017. Tajikistan.” United Nations, New York and Geneva. Frameworks: Report 87.” Earthquake Engineering https://www.armstat.am/en/ Research Institute and International Association for University of Sarajevo, Faculty of Architecture. http:// The State Committee of the Republic of Uzbekistan on af.unsa.ba/. Earthquake Engineering. http://db.world-housing.net/ Statistics. 2020. https://stat.uz/en/. generate_pdf/87/. U.S. Geological Survey. Earthquake Catalog (accessed The State Statistical Committee of the Republic of May 11, 2020). United States Geological Survey. https:// WHE (World Housing Encyclopedia). n.d. “Building of Azerbaijan. https://www.stat.gov.az/?lang=en. earthquake.usgs.gov/earthquakes/search/. the Modern Movement - Reinforced Concrete Frame Designed for Gravity Loads with No Commercial Ground The World of Teoalida. 2020. https://www.teoalida.com/ Vacareanu, R., Radoi, R., Negulescu, C. and Aldea, Floor, Romania: Report 96.” Earthquake Engineering A. 2004. “Seismic Vulnerability of RC Buildings in Research Institute and International Association for Teržan, V. 2011. “Ilija Arnautović je Projektiral Bucharest, Romania.” Vancouver, Canada: Proceedings Earthquake Engineering. http://db.world-housing.net/ Stanovanjske Objekte, v Katerih še Danes Živi Nekaj Več from 13th World Conference on Earthquake Engineering. pdf_view/96/. Kot 57 Tisoč Ljudi.” Priloga Bivanje. Woessner, J., Danciu L., D. Giardini and the SHARE WHE (World Housing Encyclopedia). n.d. “Buildings Todut, C., Dan, D. and Stoian, V. 2015. “Numerical and consortium. 2015. “The 2013 European Seismic Hazard with Cast In-Situ Load-Bearing Reinforced Concrete Experimental Investigation on Seismically Damaged Model: Key Components and Results.” Bulletin of Walls, Kyrgyzstan: Report 40.” Earthquake Engineering Reinforced Concrete Wall Panels Retrofitted With FRP Earthquake Engineering 13, 3553–3596. doi:10.1007/ Research Institute and International Association for Composites.” Composite Structures, Volume: 119, s10518-015-9795-1 Earthquake Engineering. http://db.world-housing.net/ page(s) 648–665. Issue published: January 2015. doi: generate_pdf/40/. 10.1016/j.compstruct.2014.09.047 WHE (World Housing Encyclopedia). 2002a. “Large Concrete Block Walls with Reinforced Concrete Floors WHE (World Housing Encyclopedia). n.d. “Buildings United Nations. 2002. “Country profiles on the housing and Roofs, Russia: Report 54.” Earthquake Engineering with Hollow Clay Tile Load-Bearing Walls and sector - Albania.” United Nations, New York and Geneva. Research Institute and International Association for Precast Concrete Floor Slabs, Kyrgyzstan: Report United Nations Development Programme (UNDP), Earthquake Engineering. http://www.world-housing.net/ 34.” Earthquake Engineering Research Institute and United Nations Industrial Development Organization WHEReports/wh100019.pdf. International Association for Earthquake Engineering. (UNIDO). 1985. “Building Construction Under Seismic http://db.world-housing.net/generate_pdf/34/. WHE (World Housing Encyclopedia). 2002b. “Precast, Conditions in The Balkan Region, Volume: 2 - Design Prestressed Concrete Frame Structure with concrete WHE (World Housing Encyclopedia). n.d. “Buildings and Construction of Prefabricated Reinforced Concrete Shear Walls, Serbia: Report 68.” Earthquake Engineering Protected with ‘Disengaging Reserve Elements,’ Russian Building Systems. Vienna: UNIDO. Research Institute and International Association for Federation: Report 77.” Earthquake Engineering United Nations Development Programme (UNDP), Earthquake Engineering. http://www.world-housing.net/ Research Institute and International Association for Global Environment Facility (GEF). 2013. “Improving WHEReports/wh100043.pdf. Earthquake Engineering. https://db.world-housing.net/ Energy Efficiency in Buildings - Armenia.” UNDP, New generate_pdf/77/. WHE (World Housing Encyclopedia). 2002c. “RC York. Structural Wall Building: Moment Frame with In- WHE (World Housing Encyclopedia). n.d. “Large United Nations Development Programme (UNDP), situ Shear Walls, Romania: Report 78.” Earthquake Panel Buildings with Two Interior Longitudinal Walls, Global Environment Facility (GEF). 2015. “Improving Engineering Research Institute and International Kazakhstan: Report 32.” Earthquake Engineering Energy Efficiency in the Residential Building Sector of Association for Earthquake Engineering. https:// Research Institute and International Association for Turkmenistan.” UNDP, New York. db.world-housing.net/building/78/. Earthquake Engineering. https://db.world-housing.net/ generate_pdf/32/. 83 WOR K S CIT E D (C O N TI N UE D ) WHE (World Housing Encyclopedia). n.d. “PC Frame Association for Earthquake Engineering. http://db.world- Kyrgyz Republic: Seismic Risk Reduction Strategy.” World Buildings, Armenia: Report 202.” Earthquake Engineering housing.net/building/71/. Bank Group, Washington, D.C. Research Institute and International Association for Earthquake Engineering. https://db.world-housing.net/ WHE (World Housing Encyclopedia). n.d. “Reinforced World Bank Group. 2017a. “Disaster Risk Finance building/202/. Concrete Frame Buildings without Beams, Kyrgyzstan: Country Note: Armenia.” World Bank Group, Report 39.” Earthquake Engineering Research Washington, D.C. WHE (World Housing Encyclopedia). n.d. “PC Large Panel Institute and International Association for Earthquake Building, Armenia: Report 203.” Earthquake Engineering Engineering. http://db.world-housing.net/building/39/. World Bank Group. 2017b. “Kyrgyz Republic: Measuring Research Institute and International Association for Seismic Risk.” World Bank Group, Washington D.C. Earthquake Engineering. http://db.world-housing.net/ WHE (World Housing Encyclopedia). n.d. “Small Concrete Block Masonry Walls with Concrete Floors World Bank Group. 2018. “Moldova: Earthquake generate_pdf/203/. Analysis. Country Disaster Risk Profile, Eastern Eruipe and Roofs, Russia: Report 53.” Earthquake Engineering WHE (World Housing Encyclopedia). n.d. “Precast Research Institute and International Association for and Central Asia Office.” World Bank Group, Washington Concrete Panel Apartment Buildings, Romania: Report Earthquake Engineering. https://db.world-housing.net/ D.C. 83.” Earthquake Engineering Research Institute and generate_pdf/53/. Wyllie, L. A. and Filson, J. R. 1989. “Performance of International Association for Earthquake Engineering. Engineered Structures. Earthquake Spectra, Armenia http://db.world-housing.net/pdf_view/83/. WHE (World Housing Encyclopedia). n.d. “Unreinforced Masonry Building, Slovenia: Report 73.” Earthquake Earthquake Reconnaissance Report,” Volume: 5 issue: 1_ WHE (World Housing Encyclopedia). n.d. “Precast Engineering Research Institute and International suppl, page(s): 70–92. Issue published: August 1, 1989. Reinforced Concrete Frame Building with Cruciform Association for Earthquake Engineering. http://db.world- Zarecor, Kimberly Elman. 2011. Manufacturing a and Linear-Beam Elements, Kyrgyzstan: Report 33.” housing.net/pdf_view/73/. Socialist Modernity: Housing in Czechoslovakia, 1945– Earthquake Engineering Research Institute and 1960. University of Pittsburgh Press. International Association for Earthquake Engineering. Wieland, M., Pittore, M., Parolai, S., Zschau, J., http://db.world-housing.net/generate_pdf/33/. Moldobekov, B. and Begaliev, U. 2012a. “Estimating Zorić, A. 2019. “Exposure Model for Serbia: Case Study Building Inventory for Rapid Seismic Vulnerability of Kragujevac.” Belgrade: SERA workshop. WHE (World Housing Encyclopedia). n.d. “Precast Assessment in Bishkek, Kyrgyzstan: Towards an Reinforced Concrete Frame Panel System of Seria IIS- Integrated Approach Based on Multi-Source Imaging.” 04, Uzbekistan: Report 66.” Earthquake Engineering Soil Dynamics and Earthquake Engineering, Volume: 36, Research Institute and International Association for page(s): 70–83. Issue Published: May 2012. Earthquake Engineering. http://db.world-housing.net/ building/66/. Wieland, M., Pittore, M., Parolai, S. and Zschau, J. 2012b. “Exposure Estimation from Multi-Resolution WHE (World Housing Encyclopedia). n.d. “Prefabricated Optical Satellite Imagery for Seismic Risk Assessment.” Concrete Panel Buildings with Monolithic Panel ISPRS International Journal of Geo-Information 1, no. 1: Joints (Seria 105), Kyrgyzstan: Report 38.” Earthquake 69–88. Engineering Research Institute and International Association for Earthquake Engineering. http://www. Wieland, M., Pittore, M., Parolai, S., Begaliev, U., db.world-housing.net/generate_pdf/38/. Yasunov, P., Tyagunov, S., Moldobekov, B., Saidiy, S., Ilyasov, I. and Abakanov, T. 2015. “A Multiscale Exposure WHE (World Housing Encyclopedia). n.d. “Reinforced Model for Seismic Risk Assessment in Central Asia.” Masonry, Armenia: Report 204.” Earthquake Engineering Seismiological Research Letters 86, no. 1: 210–222. Research Institute and International Association for Earthquake Engineering. http://db.world-housing.net/ World Bank Group. 2014. “Financial Protection Against generate_pdf/204/. Natural Disasters: From Products to Comprehensive Strategies.” World Bank Group, Washington, D.C. WHE (World Housing Encyclopedia). n.d. “Reinforced Concrete Frame Buildings with Diagonal Bracings and World Bank Group. 2016a. “Great Baku Housing Sector Brick Infill Walls, Romania: Report 71.” Earthquake Diagnostic.” World Bank Group, Washington, D.C. Engineering Research Institute and International World Bank Group. 2016b. “Measuring Seismic Risk in 84 CI TY P ROFILE SO URC E S Tirana, Albania Database (2018 data). References for Building Types, Classification, and Exposure Abolmasov et al. 2011; Aliaj et al. 2010; Mott MacDonald 2020a, 2020b; NCEI/WDS Global Significant Earthquake References for Building Types, Classification, and Exposure Central Statistics Office Budapest; Csoknyai et al.2014; Database; United States Geological Survey Earthquake European Union Building Database; Horváth and Szalay Mott MacDonald 2019. 2012; Hrabovszky-Horváth 2015; Ministry of National Catalog; World Bank Indicators Database (2018 data). Plovdiv, Bulgaria Development, Non-Profit Limited Liability Company References for Building Types, Classification, and Exposure Mott MacDonald 2020a, 2020b; NCEI/WDS Global for Quality Control and Innovation in Building 2015; Significant Earthquake Database; United States Geological RS Architects 2006; Stoychev 1976; United Nations Abazaj 2019; Abolmasov 2011; Aliaj 2019; Aliaj et al. Development Programme, United Nations Industrial 2010; Carydis et al. 1988; Crowley et al. 2018; Institute Survey Earthquake Catalog; World Bank Indicators Database (2018 data). Development Organization 1985. of Statistics, Tirana, Albania; Regional Environmental Center, Austrian Development Corporation 2015; Simaku References for Building Types, Classification, and Exposure Chişinău, Moldova 2014; Simaku 2017; Stanfield et al. 1999; United Nations Georgescu and Pomonis 2012; Mott MacDonald 2002. Mott MacDonald 2019. 2020a, 2020b; NCEI/WDS Global Significant Earthquake Durrës, Albania Zagreb, Croatia Database; United States Geological Survey Earthquake Abolmasov et al. 2011; Aliaj et al. 2010; Mott MacDonald Mott MacDonald 2020a, 2020b; NCEI/WDS Global Catalog; World Bank Indicators Database (2018 data). 2020,a 2020b; NCEI/WDS Global Significant Earthquake Significant Earthquake Database; Government of Croatia, References for Building Types, Classification, and Exposure Database; United States Geological Survey Earthquake 2020; United States Geological Survey Earthquake Catalog; World Bank Indicators Database (2018 data). Catalog; World Bank Indicators Database (2018 data). Alcaz et al. 2011; Cutia 2018; National Bureau of Statistics, Moldova; National Geospatial Data Fund, Moldova; References for Building Types, Classification, and Exposure References for Building Types, Classification, and Exposure United Nations Economic Commission for Europe; World Abazaj 2019; Abolmasov 2011; Aliaj 2019; Aliaj et al. Atalić et al. 2019; Central Bureau of Statistics, Republic Bank Group 2018. 2010; Carydis et al. 1988; Crowley et al. 2018; Institute of Croatia 2015; Croatian Design Society; Duzs 2013; Podgorica, Montenegro of Statistics, Tirana, Albania; Regional Environmental European Union Building Database; Emporis 2020; Mott MacDonald 2020a, 2020b; NCEI/WDS Global Center, Austrian Development Corporation 2015; Simaku Kalman Šipoš and Hadzima-Nyarko 2018; Korak 2014; Significant Earthquake Database; World Bank Indicators 2014; Simaku 2017; Stanfield et al. 1999; United Nations Mecanov 2015; Prosinečki 2015. Database (2018 data). 2002. Rijeka, Croatia References for Building Types, Classification, and Exposure Mostar, Bosnia and Herzegovina Abolmasov et al. 2011; Mott MacDonald 2020a, 2020b; City of Rijeka; Mott MacDonald 2020a, 2020b; NCEI/WDS Regional Environmental Center, Austrian Development NCEI/WDS Global Significant Earthquake Database; Global Significant Earthquake Database; Government Corporation 2015; Statistical Office of Montenegro. United States Geological Survey Earthquake Catalog; of Croatia, 2020; United States Geological Survey World Bank Indicators Database (2018 data). Earthquake Catalog; World Bank Indicators Database Skopje, North Macedonia (2018 data). Mott MacDonald 2020a, 2020b; NCEI/WDS Global References for Building Types, Classification, and Exposure Significant Earthquake Database; Petrovski 2004; World References for Building Types, Classification, and Exposure Bank Indicators Database (2018 data). Ademović and Hadzima-Nyarko 2018; Ademović 2019; Agency for Statistics of Bosnia and Herzegovina; Atalić et al. 2019; Central Bureau of Statistics, Republic References for Building Types, Classification, and Exposure Arnautović-Aksić, et al. 2016; Central Census Bureau of Croatia 2015; Croatian Design Society; Duzs 2013; of Bosnia and Herzegovina 2013; Croatian Professional European Union Building Database; Emporis 2020; Folić et al. 2011; Shendova et al. 2019; State Statistical Firefighters Association; Crowley et al. 2018; Episcope, Kalman Šipoš and Hadzima-Nyarko 2018; Korak 2014; Office, Republic of North Macedonia. Tabula BA Bosnia and Herzegovina; Petrovic, IMS Mecanov 2015; Prosinečki 2015. Bucharest, Romania Institute; University of Sarajevo, Faculty of Architecture. Budapest, Hungary Georgescu and Pomonis 2008; Mott MacDonald Sofia, Bulgaria Mott MacDonald 2020a, 2020b; NCEI/WDS Global 2020a, 2020b; NCEI/WDS Global Significant Earthquake Mott MacDonald 2020a, 2020b; NCEI/WDS Global Significant Earthquake Database; United States Geological Database; United States Geological Survey Earthquake Significant Earthquake Database; United States Geological Survey Earthquake Catalog; World Bank Indicators Catalog; World Bank Indicators Database (2018 data). Survey Earthquake Catalog; World Bank Indicators Database (2018 data). 85 C ITY P ROF I LE SO U RC E S (C O NT I N U E D ) References for Building Types, Classification, and Exposure References for Building Types, Classification, and Exposure References for Building Types, Classification, and Exposure Botici et al. 2013; Fülöp et al. 2013; Muntean and Open Data Bratislava 2019; Stoychev 1976; Zarecor 2011. National Statistical Committee of the Kyrgyz Republic Ungureanu 2017; Muntean et al. 2017; National Institute 2015; OpenDRI 2020; Wieland et al. 2012a; Wieland et al. of Statistics Romania 2020; Socioeconomic Data and Ljubljana, Slovenia 2012b; Wieland et al. 2015; World Housing Encyclopedia Applications Center 2020; The World of Teoalida 2020; City of Ljubljana; Mott MacDonald 2020a, 2020b; NCEI/ Reports 40, 34, 77, 33, 38, 39; World Bank Group 2016b; Todut et al. 2015; Vacareanu, et al. 2004; World Housing WDS Global Significant Earthquake Database; United World Bank Group 2017b. Encyclopedia Reports 78 (2002c), 97 (2003), 87, 96, 83, States Geological Survey Earthquake Catalog; World Bank 71. Indicators Database (2018 data). Osh, Kyrgyz Republic Mott MacDonald 2020a, 2020b; NCEI/WDS Global Iași, Romania References for Building Types, Classification, and Exposure Significant Earthquake Database; United States Geological Georgescu and Pomonis 2008, 2012; Institutul Național Dolšek 2019; Rakušček et al. 2012; Republic of Survey Earthquake Catalog; World Bank Indicators de Statistică 2016; Mott MacDonald 2020a, 2020b; NCEI/ Slovenia Statistical Office; Teržan 2011; World Housing Database (2018 data). WDS Global Significant Earthquake Database; United Encyclopedia Report 73. States Geological Survey Earthquake Catalog; World Bank References for Building Types, Classification, and Exposure Indicators Database (2018 data). Almaty, Kazakhstan National Statistical Committee of the Kyrgyz Republic Ministry of National Economy of the Republic of 2015; OpenDRI 2020; Wieland et al. 2012a; Wieland et al. References for Building Types, Classification, and Exposure Kazakhstan Statistics Committee; Mott MacDonald 2012b; Wieland et al. 2015; World Housing Encyclopedia Botici et al. 2013; Fülöp et al. 2013; Muntean and 2020a, 2020b; NCEI/WDS Global Significant Earthquake Reports 40, 34, 77, 33, 38, 39; World Bank Group 2016b; Ungureanu 2017; Muntean et al. 2017; National Institute Database; United States Geological Survey Earthquake World Bank Group 2017b. of Statistics Romania 2020; Socioeconomic Data and Catalog; World Bank Indicators Database (2018 data). Applications Center 2020; The World of Teoalida 2020; Dushanbe, Tajikistan References for Building Types, Classification, and Exposure Mott MacDonald 2020a, 2020b; NCEI/WDS Global Todut et al. 2015; Vacareanu, et al. 2004; World Housing Encyclopedia Reports 78 (2002c), 97 (2003), 87, 96, 83, Ministry of National Economy of the Republic of Significant Earthquake Database; United States Geological 71. Kazakhstan Statistics Committee 2020; King et al. 1996; Survey Earthquake Catalog; World Bank Indicators Japan International Cooperation Agency 2009; Samoilov Database (2018 data). Belgrade, Serbia 2004; World Housing Encyclopedia Report 32. Georgescu and Pomonis 2012; Mott MacDonald References for Building Types, Classification, and Exposure 2020a, 2020b; NCEI/WDS Global Significant Earthquake Shymkent, Kazakhstan Agency for Statistics Under President of The Republic Database; United States Geological Survey Earthquake Ministry of National Economy of the Republic of of Tajikistan 2020; Petrovic et al. 2015; United Nations Catalog; World Bank Indicators Database (2018 data). Kazakhstan Statistics Committee; Mott MacDonald Economic Commission for Europe 2011. 2020a, 2020b; NCEI/WDS Global Significant Earthquake References for Building Types, Classification, and Exposure Database; United States Geological Survey Earthquake Ashgabat, Turkmenistan Alfirević and Alfirević 2015; Borozan 2019a, Borozan Catalog; World Bank Indicators Database (2018 data). Mott MacDonald 2020a, 2020b; NCEI/WDS Global 2019b; Ignjatović et al. 2013; Mecanov 2015; Mrduljaš Significant Earthquake Database; United States Geological References for Building Types, Classification, and Exposure Survey Earthquake Catalog; World Bank Indicators and Vladimir 2012; Nikolic 2016; Nikolic 2018; Радовановиh and Петрониjевиh 2009; Petrovic, IMS Ministry of National Economy of the Republic of Database (2018 data). Institute; Regional Environmental Center, Austrian Kazakhstan Statistics Committee 2020; King et al. 1996; References for Building Types, Classification, and Exposure Development Corporation 2015; Statistical Office of the Japan International Cooperation Agency 2009; Samoilov Republic of Serbia 2020; World Housing Encyclopedia 2004; World Housing Encyclopedia Report 32. King et al. 1996; State Committee on Statistics of Report 68; Zorić 2019. Turkmenistan 2020; United Nations Development Bishkek, Kyrgyz Republic Programme, Global Environment Facility 2015. Bratislava, Slovakia Mott MacDonald 2020a, 2020b; NCEI/WDS Global Mott MacDonald 2020a, 2020b; NCEI/WDS Global Significant Earthquake Database; United States Geological Tashkent, Uzbekistan Significant Earthquake Database; United States Geological Survey Earthquake Catalog; World Bank Indicators Mott MacDonald 2020a, 2020b; NCEI/WDS Global Survey Earthquake Catalog; World Bank Indicators Database (2018 data). Significant Earthquake Database; United States Geological Database (2018 data). Survey Earthquake Catalog; World Bank Indicators Database (2018 data). 86 CI TY P ROFILE SO URC E S (C O N T I N U E D ) References for Building Types, Classification, and Exposure Republic of Azerbaijan 2019b; World Bank Group 2016a. Ibragimov et al. 2000; Mavlyanova et al. 2004; The State Tbilisi, Georgia Committee of the Republic of Uzbekistan on Statistics Mott MacDonald 2020a, 2020b; NCEI/WDS Global 2020; World Housing Encyclopedia Report 66. Significant Earthquake Database; United States Geological Survey Earthquake Catalog; World Bank Indicators Yerevan, Armenia Database (2018 data). Balassanian et al. 2003; Mott MacDonald 2020a, 2020b; NCEI/WDS Global Significant Earthquake Database; References for Building Types, Classification, and Exposure United States Geological Survey Earthquake Catalog; Bouzarovski et al. 2011; City Kvira 2018; Colliers World Bank Indicators Database (2018 data). International 2014; Giardini et al. 2018; National References for Building Types, Classification, and Exposure Statistics Office of Georgia 2013; National Statistics Office of Georgia 2020; World Housing Encyclopedia Reports 54 Japan International Cooperation Agency 2012; King et al. (2002a),77, 53. • 1996; Statistical Committee of the Republic of Armenia 2011; Statistical Committee of the Republic of Armenia 2017; United Nations Development Programme Global Environment Facility 2013; World Housing Encyclopedia Reports 202, 203, 204. Gyumri, Armenia Balassanian et al. 2003; Mott MacDonald 2020a, 2020b; NCEI/WDS Global Significant Earthquake Database; United States Geological Survey Earthquake Catalog; World Bank Indicators Database (2018 data); Zschau and Küppers 2003. References for Building Types, Classification, and Exposure Japan International Cooperation Agency 2012; King et al. 1996; Statistical Committee of the Republic of Armenia 2011; Statistical Committee of the Republic of Armenia 2017; United Nations Development Programme Global Environment Facility 2013; World Housing Encyclopedia Reports 202, 203, 204. Baku, Azerbaijan Mott MacDonald 2020a, 2020b; NCEI/WDS Global Significant Earthquake Database; United States Geological Survey Earthquake Catalog; World Bank Indicators Database (2018 data). References for Building Types, Classification, and Exposure Infrastructure Department ECA 2006; PopulationStat 2020; The State Statistical Committee of the Republic of Azerbaijan; United Nations Economic Commission for Europe 2010; State Statistical Committee of the Republic of Azerbaijan 2019a; State Statistical Committee of the 87 This page is intentionally left blank. ©The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Image Credit: “Damaged building in Đorđićeva [Zagreb, 2020]” Washington, D.C. 20433, U.S.A by Franjo Tahy, licensed under CC0 1.0. August 2020 88 Eart hq u a ke Ri s k i n Mu l t i fa m i l y Resid e nti a l Bu i l d i n gs E urope an d C ent ra l A s i a Reg i o n