Damages of Non-Structural Components

Transcription

1 Building Damages 20

2 Damages of Non-Structural Components 21

3 Damages of Building Utilities 22

4 Loss Estimation Model Vulnerability Curve Loss Ratio Loss Amount = Replacement CostLoss Ratio Loss Ratio 20% JMA Seismic Intensity 6.4+ Ground Motion 23

5 Levels of Analysis Rigor Three Levels Level Statistical desk top analysis Replacement Value Location Construction Class Year Build # of Stories Occupancy Level Enhanced analysis based on engineering review of design drawings and calculations. Yields customized performance modeling Level Level 2 with inspection to determine as-built condition vs. original design. Yields customized performance modeling. Most rigorous, yet cost-effective assessment of risk. Combination of different levels 24

6 Level-3 Analysis Break down a Building to Components Constructional Element of a Building Beam, Columns Earthquake-resisting wall Exteriors Partition walls Ceilings Electrical equipment Pipes, Ducts etc. Structural Components: Damaged by Horizontal Force Beam, Columns Earthquake-resisting wall etc. Non-Structural Components: Damaged by Deformation Exteriors Partition walls Pipes, Ducts etc. Non-Structural Components: Damaged by Acceleration Ceilings Electrical equipment etc. 25

7 Response Analysis and Loss Estimation of Damaged Components 26

8 Various Analysis Levels for Loss Estimation Information Few Many Uncertainty Large It is important to select a right method to meet objectives by considering amount of information and cost Small Cost Low High Level One Desktop Level Two Expert Opinion Level Three Engineering ReviewInspection Sophisticate Analysis Rupture Source Model 3D Dynamic Analysis Non Linear Dynamic Analysis 27

9 Probability Distribution at 10% of Mean Loss (Secondary Uncertainty) Vulnerability Curve Median = 9.0% Loss Ratio Probability Density 50% Median = 3.8% Mean = 10% Ground Motion Level Three Analysis Level One Analysis 90 percentile = 17.6% 50% 90% 10% 90 percentile =29.7% 0% 5% 10% 15% 20% 25% 30% Loss Ratio 28

10 Portfolio Effect Change of Uncertainty by Summing Probability Density Mean Loss Ratio = 10% Probability density distribution at N=20 is almost equivalent with Level-3 analysis N=1 N=5 N=10 N=20 N=50 N=200 0% 5% 10% 15% 20% 25% 30% Loss Ratio 29

11 Portfolio Effect Risk Risk Concentration Tokyo Tokyo Chiba Chiba Yokohama Add to Loss axis direction Risk Risk Diversification Tokyo Tokyo Aomori Aomori Osaka Osaka Add to Probability axis direction 30

12 Business Interruption Model Building/ Facility Damage Occupancy Days of Downtime Lifeline Damage Lifeline Importance Factors Business Interruption Loss 31

13 Facility Restoration Function BI Model BI Loss Days of DowntimeIncome per Day Restored Function (%) Total Days of Downtime D 30 D 60 D 100 Days 32

14 Facility Restoration Function (Specific Occupancy Class) Loss Ratio 0-1% 1-10% 10-30% 30-60% % 100% 33

15 Casualty Model Exposure Data Geographic distribution of people, building type, time of day Simulated Catastrophic Events Building damage and collapse distributions by building type; population injured, entrapped, rescued, and injury distributions Casualties & Losses by event for Risk Curve Treatment costs Insurance claims settlements Fatal Permanent Permanent Total Partial-Major Permanent Partial-Minor Temporary Total Medical Only 34

16 Scenario Results Kobe, 1/17/1995, 5:46, M7.2 Niigataken-Chuetsu, 10/23/2004, 17:56, M6.8 Estimated Actual Modeled Estimated Actual Modeled Temporary Total 8,000 12,000 Serious 635 Temporary Total 42 5 PP-minor 2,400 2,800 PP-minor 45 PP-major 1,000 1,100 PP-major 15 Permanent Total Permanent Total 1 Fatalities 5,500 5,300 Fatalities 25 Fatalities 24 35

17 Estimated Loss of each Stochastic Events Mean loss 90 percentile loss 36

18 Event Curve Probability of Exceeding a Specific Loss Level Event # M M M M M7.20 Estimated Loss (billion Yen) Annual Probability of Event Occurrence 0.100% 0.142% 0.290% 0.177% 0.172% Cumulative Annual Probability of Exceedance 0.100% 0.241% 0.530% 0.706% 0.877% 37

19 Interpretation on Event Curve Probability of Exceeding a Specific Loss Level Annual Average Loss Annual Average Loss (AAL ΣEstimated LossAnnual Probability of Occurrence ΣEstimated LossReturn Period Expected Loss per Year At least, 8.2 billion is expected within the next 100-year period Worst case scenario 13.6 billion 38

20 Loss Distribution Reflects Potential Levels of Damage from a Single Event 90 percentile loss Annual Exceedance Probability Mean loss Estimated Loss 39

21 Risk Curve Exceedance Probability Curve with Secondary Uncertainty Annual Exceedance Probability Exceedance Probability Curve with Secondary Uncertainty L Estimated Loss 40

22 Event Curve vs. Risk Curve Annual Exceedance Probability Substantial Difference on on Annual Exceedance Probability Event Event Curve: Curve: Event EventOccurrence Risk Risk Curve: Curve: Loss LossOccurrence Estimated Loss (billion Yen) 41

23 Aggregation of Risks Market Risk Financial Risk CAT Risk Credit Risk Value Value Non-CAT Risk Value Correlation Aggregations Operation Risk Value Value Aggregated Risk Value Value 42

24 Event Curve vs. Risk Curve Annual Exceedance Probability Event Curve (Mean) Risk Curve with Small Uncertainty Risk Curve with Large Uncertainty Estimated Loss (billion Yen) 43

25 TCETail Conditional Expectation Annual Exceedance Probability Lp TCE Cruve = Ep / p Expected Loss (AAL): Ep Estimated Loss (billion Yen) 44

26 TCETail Conditional Expectation Annual Exceedance Probability Risk Curve with 2ndary Uncertainty Event Curve (Mean) Event Curve (90 percentile) TCE Curve Estimated Loss (million Yen) 45

27 Estimated Losses and Financial Statements Loss by Physical Damages Losses by facility damages Repairing Cost Dismantlement Cost, etc. Balance Sheet (Assets Side) Depletion of Cash Income Statement P/L Extraordinary Loss / Non-operating Expense Restoration Cost excluding the above (Asset Capitalization ) Loss of Booked Values Damages of Inventory Assets Losses by Business Interruptions Depletion of Cash New Asset (after recovery) Depletion of Fix Assets Depletion of Liquid Asset Depletion of Cash Depletion of Earnings Amortization Extraordinary Loss Extraordinary Loss Drop in Sales/Profit 46

28 Study on Risk Measures by Using Risk Curve of Equity Capital Annual Exceedance Probability Risk Curve at Present Acceptable Limit of Equity Capital Depletion Possible Occurrence with P% Probability P% P% of probability as the worst case scenario for the study 47 5 B Yen 7.5 B Yen Depletion of Equity Capital Require 2.5 billion yen to recover equity capital by insurance or others

29 Study on Risk Measures by Using Risk Curve of Equity Capital Annual Exceedance Probability Risk Curve at Present Acceptable Limit of Equity Capital Depletion Risk Curve after Mitigation Possible Occurrence with P% Probability Possible Occurrence with P% Probability P% P% of probability as the worst case scenario for the study 3 B Yen 5 B Yen Depletion of Equity Capital As 3 B Yen is in acceptable range, additional financial measure is not required. 48

30 Conclusions Assessment technology on Seismic Risk has already reached at a certain level of probabilistic approach. Utilizing existing models is more efficient way to build specific risk models for financial institutes First, a perspective model and then move to a detailed model Implementing uncertainties adequately is very important for risk modeling. Losses by earthquake will spread geographically and temporally Limiting seismic risk to operational risk is an appropriate way? 49