Skip to main content

Probabilistic Assessment of Reduction and Transfer of Natural Earthquake Risk

Periodic Reporting for period 1 - PARTNER (Probabilistic Assessment of Reduction and Transfer of Natural Earthquake Risk)

Reporting period: 2017-01-01 to 2018-12-31

Seismic risk management in fast-growing populated seismic areas is a challenge for governments and communities in case of catastrophic events. Seismic resilience options are considered to mitigate risk, by either reducing the risk through vibration-control designs of risky assets, or by transferring the risk into the capital markets, through disaster financing options or (re)insurers. Both strategies require a good understanding of the seismic risk, which can be achieved only through thorough ground-motion uncertainty quantification and propagation to structural response, and accurate estimates of damage, cost and downtime estimates of the affected assets. The current proposal presents a complete framework to analyze the effects of resiliency measures on the overall risk of communities by (1) developing novel probabilistic model to characterize local seismic hazard by using site-specific records; (2) characterizing the exposed assets accurately using complex systems and dropping the linearity assumption of system's behaviour under seismic loads; (3) developing novel and efficient seismic vulnerability models consistent with the seismic hazard; and (4) calculating probability distributions of seismic-performance metrics rather than just mean values for a better characterization of the risk. This methodology will be applied to study the risk reduction and transfer effects on communities. Seismic risk-reduction will be achieved through implementation of seismic control devices to structures and risk-transfer is achieved through financing mechanisms, such as catastrophe bonds and other parametric models, used to carry over risk to risk-takers.

The following objectives were defined:
• The development and the assessment of site-specific seismic-hazard and general intensity-measure models that can incorporate prior physics-based knowledge and characteristics of recorded data;
• The development models that can represent risky assets accurately by reducing the number of limiting assumptions, and study the effects of risk-reducing devices on the spatial distribution of losses in risky communities;
• The assessment of traditional seismic-intensity measures used in Catastrophe Modelling, and propose novel seismic-intensity measures that can describe better the seismic hazard in case of catastrophic events; and
• The development of risk-transfer seismic parameters that can be used to design sufficient and efficient payout triggers for risk-transfer financial products.
The work performed in the current project was structured in different tasks split in four main work packages (WPs). The scheme attached, also known as a Project Evaluation Review Technique (PERT) diagram describes schematically the WPs in the PARTNER project and the interdependencies between them through the red arrows, the direction of the arrow showing which activity is the input to which. The main activities in each WP are represented in the dark-blue boxes, in which also references to the proposed deliverables are made.

Work Package 1: Seismic Hazard
The objective WP1 was to develop site-specific models for the generation of synthetic ground motion time histories, and it was successful completed. The hazard model developed is general and can be used for producing catalogues of synthetic ground motion records statistically consistent with recorded data, even in areas with poor historical data. A comprehensive characterisation of earthquakes is done in the model proposed, as function moment magnitudes and epicentral distances, as well as source and site conditions, all essential parameters for the appropriate description of the frequency content of earthquakes.

Work Package 2: Exposure
WP2 was aiming to provide a methodology to account more accurately for the exposed to catastrophe risk in order to assess more accurately and meaningfully the special distribution of losses. The work in this work package is summarized by the following two activities: (1) to propose advanced methodologies to analyse more accurately complex structures, such as multi-story buildings and infrastructures, e.g. bridges, subjected to seismic loads; and (2) to assess the spatial distribution of losses with respect to retrofitting methods applied to assets exposed to earthquake risk, and more importantly.

Work package 3: Vulnerability
The aim of WP 3 was to formulate a framework that produces vulnerability functions for structures. These functions show how vulnerable different types of structures, such as multi-story buildings made of different materials, infrastructure, are to different types of earthquake characterised by their moment magnitudes and epicentral distances. Based on these functions losses can be estimated for different types of buildings based on their occupancy type, material built, foundation, in case they are struck by an earthquake event.

Work package 4: Risk Reduction & Transfer
WP 4 is mainly the result of the catastrophe-modelling framework and also represents ways of using this framework for different purposes, summarized best by the two main activities of this WP: (1) risk transfer and (2) risk reduction.
The risk-transfer activity developed herein treated alternative risk-transfer products, unlike the traditional insurance. These novel products are based on earthquake-related indexes, which determine in the case of a catastrophic seismic event, whether the risk-transfer product triggers a payout or not, irrespective of the level or whether the damage actually occurred following that event. Thus the losses produced by earthquakes and these indexes must have a good statistical dependence for the product to be fair. The risk-reduction activity studied several seismic-vibration-reduction devices such as (a) the viscous damper, (b) the tuned-mass damper and (c) the tuned-inerter damper (TID). A higher focus was emphasized on the last device, with applications for multi-story buildings and cable-stayed bridges.
The project was highly innovative with several aspects of the catastrophe-modelling framework used to price the impacts of natural disasters on risky assets, in the (re)insurance industry. The results are expected to have a high impact in industry and are summarized below through the following main components:
- Hazard: the ground motion simulation model is site-specific, consistent with recorded data and produces synthetic ground motion with natural non-Gaussian, non-stationary characteristics;
- Exposure: non-intrusive, accurate methods to calculate response of structures to earthquakes efficiently without using Monte Carlo simulations were developed;
- Vulnerability: vulnerability functions were calculated as functions of moment magnitude and epicentral distance, which correlate better the response of structures to earthquake characteristics;
- Risk-trasnsfer: extreme-value theory was used to determine payout trigger mechanisms in case of earthquakes to ensure a fast payout for the affected parties;
- Risk-reduction: innovative vibration-reduction mechanisms were tested/designed in a stochastic environment to protect buildings from damage, when subjected to earthquakes.