Developing a framework for the seismic risk assessment of interdependent lifelines systems for the insurance sector Indranil Kongar University College London UC Lifelines Week Day 2 21st April 2015

Single system assessment 1.  Infrastructure system classification 2.  Select damage scales 3.  Determine appropriate hazard parameters 4.  Generate seismic demand fields 5.  Assign damage to system components 6.  Measure system performance

2

Risk = Hazard x Vulnerability x Exposure HAZARD

Interdepende Independent nt

Earthquake Direct physical damage

Direct physical damage

VULNERABILITY

Element System AA

Element System BB Indirect operational outage

EXPOSURE

Direct impact: Repair cost

Indirect impact: Business interruption

Direct impact: Repair cost

Indirect impact: Business interruption

3

Catastrophe Models Loss = f(Exposure, Hazard, Vulnerability) Risk = f(Σ Loss x Probability of hazard)

4

Infrastructure taxonomy Infrastructure system Electric power

Potable water

Components

Component attributes

Generation plants Substations Cables Wells Water treatment plants Pumping stations Storage tanks

Capacity, seismic design level Voltage, seismic design level Material, size Seismic design level Capacity, seismic design level Capacity, seismic design level Elevation, material, geometry, seismic design level

Pipelines Lift stations Treatment plants Pipelines Natural gas Pipelines Compressor stations Fuel Refineries Pumping stations Storage tanks Pipelines Telecommunication Central offices Cables s Highways Roadways Bridges Waste water

Tunnels Embankments

Material, joint type, age, diameter Capacity, seismic design level Capacity, seismic design level Material, joint type, age, diameter Material, joint type, age, diameter Capacity, seismic design level Capacity, seismic design level Capacity, seismic design level Elevation, material, geometry, seismic design level Material, joint type, age, diameter Seismic design level Material, size Importance level Structural system, material, age, geometry, seismic design level Construction method, geometry, local geology Height, soil type

5

6

Interdependency •  Two approaches to interdependency: a.  Model connections between components within a systems, or: b.  Model coupling between systems as a whole

System of systems

System coupling 7

Interdependency

Source: Reed et al. (2009)

Empirical Expert judgment

Source: Duenas-Osorio & Kwasinski (2012)

8

Case study: Christchurch Compare observed damage from September 2010 and February 2011 events with predictions from HAZUS

9

Buried cable fragility •  No existing fragility functions for buried cables •  Damage to cables observed from both September 2010 and February 2011 earthquakes so new functions can be derived

10

Observed liquefaction

11

Damage in non-liquefaction areas Typology

Ground shaking only

Minor liquefaction

Minor lateral spreading

Copper

0.08

0.23

0.44

Aluminium

0.02

0.02

0.32

PILCA

0.06

0.15

0.41

XLPE

0.01

0.00

0.20

Repair rates in zones where there was no observed ground deformation are lower than in areas where ground deformation was observed, indicating that ground deformation is a stronger driver of damage than ground shaking

12

Substations 1.0 0.9 0.8 0.7 0.6

0.4

New Min

0.3

HAZ Comp

0.2

New Comp

P (DS ≥ DSi)

0.5

HAZ Min

0.1 0.0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2.0

Peak ground acceleration (g)

13

Thanks for your attention

[email protected] www.epicentreonline.co.uk

14