Abstract on seismic risk simulation

Assessment of damage costs due to an earthquake is an important issue for the insurance industry, yet it currently lacks the level of detail that is required for realistic pricing of individual risks. Sophisticated numerical tools which are widely used to assess structural behavior in great detail (like  FEM) have developed to  a level where they may be used with confidence in seismic risk assessment and as such enable the insurance industry to perform more realistic assessments.
 
In this statistical approach we show a first proposal of how to integrate sophisticated 3D-FE modeling in insurance risk assessment. It is illustrated using masonry structures during the L’Aquila earthquake of April 6, 2006, Italy.
 
The process is broken down into six steps. The first step is to create multiple instances of a type of building (e.g. masonry) and calculate the transfer functions of each numerical 3D model.

The  second  step  implements  all  the  necessary  hazard  parameters  (PGA,  soil  type)  of  an earthquake region according to the generated locality of each instance. The "statistical process" will  then  use  the  power  spectra  of  local  stresses  to  assess  the  probability  of  a  crack  in  each finite  element.  This  can  be  related  to  the  rehabilitation  costs,  followed  by  the  creation  of  the vulnerability function for each building class as a function of the return period.

In the final step an integration of the vulnerability function will yield the insurance premium.
 
These results are compared with the values of the L’Aquila event from the Munich RE reinsurance for the insurance premium and the PML (Probable Maximum Loss)

Example for damage curves of four building variants for an earthquake region