Catastrophe risk assessment for individual structures can prove a useful tool for profitable decision-making

When it comes to extreme natural and man-made events - earthquakes, hurricanes, severe thunderstorms and acts of terrorism - professionals concerned with the risk of prominent, high value or strategically significant facilities have a lot riding on the design and engineering of the structure.

While the chance of a catastrophe impacting a specific building is small, property damage, casualties and business interruption losses can be devastating should the worst occur. To proactively manage the risk, building owners and insurers need to be fully aware of both the potential for catastrophes at their locations and the vulnerability of their specific properties.

Catastrophes can impact every region of the US, leaving virtually all properties at risk from one peril or another. While risk managers are aware of the obvious risk of hurricanes along the Gulf Coast and Eastern seaboard, earthquakes in California and the Pacific Northwest, severe thunderstorms in the Midwest and terrorism in major urban centres, they must also take into account that within these regions there is wide variability in the actual loss potential.

Insurers and reinsurers rely on catastrophe modeling to estimate the potential impact of various perils on individual policies and their portfolio as a whole. In contrast, property owners and risk managers frequently turn to engineers to obtain probable maximum loss (PML) assessments for their structures. Alone, neither approach is sufficient to significantly reduce the uncertainty regarding an individual structure's likely behaviour under extreme conditions, or determine the return on investment for various options required to mitigate the risk.

It is the synthesis of these two approaches, often referred to as a detailed catastrophe risk assessment, that can provide risk managers, brokers, loss control experts and facultative reinsurers with the information necessary to make a wide variety of mission-critical risk management decisions for individual structures, including insurance purchasing and underwriting decisions for property, life and workers' compensation lines.

Science questions

What is the likelihood of a catastrophe affecting a property? Science still cannot predict the time and location of catastrophic events. Science, however, can determine the probability that an event of a given severity will occur at a specific location. With computer models, extreme events are mathematically simulated to reflect the frequencies and intensities that we can expect to see in reality. For example, the brevity of the historical earthquake record and the resulting scarcity of data is overcome by simulating thousands of possible future scenarios. With this approach, an event that may occur only once every thousand years can be analysed.

Detailed catastrophe risk assessments are similar to standard catastrophe modeling in that they consider the full range of possible events that could impact a property. PML analysis, the approach most commonly used for individual structures, only considers a single event. The use of only one event to quantify loss potential offers little context to base significant risk management decisions. The limited scope of a PML analysis is exacerbated by the variation in PML definitions adopted by different engineering firms.

Why is it important to simulate the full range of possible events? Because each event is unique and will impact a property differently. For example, each event in the simulation mathematically replicates the location (or path) of the event, the effects of the local geography and, most importantly, the intensity level and duration of forces that directly impact the specific structure. Furthermore, the most devastating losses don't necessarily stem from the closest or most intense event. Consequently, the only way that strategic risk management decisions can be made with confidence is with a full understanding of the likelihood of losses resulting from a full distribution of potential events.

To accurately quantify the likelihood and extent of the resulting damage - and subsequently the likelihood and extent of loss - the probability and variability of different scenarios must be carefully depicted. Doing so requires simulating thousands and sometimes hundreds of thousands of events to uncover subtle but significant variations in event characteristics.

Engineering questions

To fully understand the likely performance of a structure under extreme conditions, a detailed engineering analysis must be carried out. In this respect, detailed catastrophe risk assessments are significantly different from standard modeling analyses and PML studies.

Standard modeling approaches estimate damage based on the performance of similarly engineered structures of some typical design and average vulnerability. They rarely include a site visit. PMLs usually require a site visit by an engineer to make qualitative assessments about the building condition. In contrast, detailed catastrophe risk assessments include site visits to study the structure and local hazard conditions, such as soil type for earthquake risk. The visit is supplemented by an examination of engineering plans to digitally create 2-D and 3-D computer models of the structure. These computer models are used to simulate the impact of various catastrophic events on the structure.

Advanced science, engineering and technology converge to determine the vulnerability of the structure. The modeled structure is subjected to the full range of conditions that extreme events would generate, such as severe ground motion, extreme winds, or the shock and pressure waves from explosives. Only by using multiple event simulations and a model of the actual building can the resistivity and/or vulnerability of structural components be fully assessed.

In many cases, the failure of a single building component leads to additional stress on, and subsequent failure of, adjacent components. The advanced modeling process employed in detailed risk assessments can capture these kinds of domino effects. This will help identify other types of structural failure, such as a particularly vulnerable floor or building wing. An awareness of these details will pinpoint employees and equipment most at risk, and the likely impact of various scenarios on site operations.

Once the type and severity of damage caused by each scenario is understood, the replacement cost can be tabulated for each damaged component. This analysis can also include factors such as the post-event inflation of materials and labour costs known as demand surge. By calculating the loss from each modeled event, a distribution of potential losses and their likelihood of being exceeded can be developed. Unlike a single number, a loss distribution enables each organisation to interpret the results based on their own tolerance for risk and provides an essential foundation for decision-making.

Better decisions

An analysis is worth little more than the paper it is written on unless you can apply the findings toward making sound business decisions. For each party involved in a risk transfer transaction, the findings of a detailed catastrophe engineering analysis provide the foundation for more confident, and hence more profitable, decision-making.

Decision analysis is a method that is used to frame problems under conditions of uncertainty according to a quantitative metric, such as return on investment or internal rate of return, in order to make the optimal selection among a number of choices. In the case of a detailed catastrophe risk assessment, the distribution of potential losses and the engineering insights derived from the event simulations can be used to facilitate a number of important decisions. These can include identifying the most cost-effective combination of insurance and engineering retrofit mitigation measures to optimally price a risk, or to select the most profitable layer for a facultative reinsurer to underwrite.

By incorporating the details derived from on-site facility inspections, review of engineering plans and the site-specific catastrophe modeling processes, the behaviour, durability and potential failure of a structure impacted by a catastrophe can be determined with a high degree of certainty.

Regardless of which side of the table you are sitting on, as a transferor of risk or an underwriter of risk, the benefits of conducting a detailed catastrophe risk assessment are significant.

Catastrophe risk assessment

The owner of a high-profile shopping facility in an earthquake-prone city centre in California commissioned a detailed catastrophe risk assessment.

The property owner had concerns about insurance adequacy and the structure's potential for collapse. The facility often attracted visitors in the tens of thousands on any given day and was valued at approximately $200m.

The objective of catastrophe risk assessment was to evaluate the structure's ability to withstand a major earthquake and quantify losses due to structural and contents damage, as well as business interruption. Furthermore, the assessment was to identify an optimal risk mitigation strategy with emphasis on property insurance needs.

First, an on-site visit and a thorough review of the building blueprints were conducted. Using this information, the structure of the building was recreated digitally, using three-dimensional computer applications.

The computer generated, three-dimensional building was then subjected to tens of thousands of simulated earthquakes as determined using a catastrophe model. The results of the analysis showed that the structure was sound for frequent, less severe events, but vulnerable to large events.

Based on the results of the catastrophe modeling analysis, the probability of exceeding (PE) various levels of monetary loss in the next ten years was estimated. The analysis showed that a total loss ($200m) had a 0.5% chance of occurring in the next ten years. A loss exceeding 50% of building replacement value had a 0.7% chance of occurring in that period. In the latter case, business interruption was expected to exceed six months.

Next, a decision analysis was conducted to determine the most appropriate mitigation strategy for the property owner. It was decided that a combination of a $3m dollar partial retrofit of the building and insurance coverage for only 50% of the full replacement value of the building, with a 10% deductible, was the best option.

Before the analysis was performed, the owner was buying insurance to cover the full replacement of the building. This new strategy resulted in an $800,000 per year reduction in insurance premium, enough to cover the cost of the retrofit in just four years. The internal rate of return of the risk mitigation investment was 12%. The property owner selected this strategy, confident that a total loss was extremely unlikely and that the partial retrofit was the best investment.