It’s Time to Take a Longer View of Risk for Critical Infrastructure
XL Catlin responds to EU STREST Goals
Fukushima was a wake-up call to Europe. The EU realized that stress-tests for critical infrastructure (CI) across Europe did not account for the potential of chain-reaction disasters. The EU has engaged seismologist Dr. Domenico Giardini and his 8-country team to design a new, stress-test framework for critical infrastructure.
XL Catlin's Alan Milroy discusses how Europe can increase infrastructure resilience by taking a longer view of risk.
Mr. Milroy, are insurers concerned about cascading disasters in Europe?
Cascading disasters are not a new topic for the insurance industry, but the EU STREST project certainly promises to offer new insights into this area. The past two decades have shown that global warming fuels more volatile weather patterns. Violent, deadly storms which once struck once in 100 years are becoming significantly more frequent. That means that 300-year, 500-year events are occurring every 100 years or more often.
The Insurance industry has been analyzing risk for over a century, and has developed a large body of detailed knowledge of both risk and losses.
Dr. Giardini is right about redefining risk. Even LP-HI needs to be redefined. Storms which were once LP could become highly probable. As Dr. Giardini also points out, governments and businesses need to include vital economic sectors or regions in the category CI. We need to reevaluate resilience preparation in these sectors, as well as in the traditional CI.
How does insurance help clients assess the potential impact of LP-HI events and cascading disasters?
First, we examine the client’s global portfolio, and identify vulnerable or especially critical properties in the context of both the client’s and the insurer’s overall portfolio risk. We deploy risk engineers to explore those properties and their surroundings, in order to pinpoint and quantify risks.
A single building can be complex. Companies often add on to factories as needed, so the age, structure, materials, and construction style can vary from one area to the next. Operations may vary, so the impact on one building to the business will depend on specifics within that building. This is not always a direct function of the relative size of a building on a site. Small buildings can also be complex. Building vulnerability can therefore vary for each kind of LP-HI event.
Outside of the building itself, isn’t it also difficult to calculate the potential impact of a cascading disaster in the surrounding area?
It is certainly complicated. After the building assessment, we investigate the surroundings.
Is the factory near a chemical plant or other high-risk facility? What are the conditions at that facility, especially if it belongs to a different company? Is the area exposed to natural disasters? Is it near a dam? What is the condition of the dam? We have a long list of location risk criteria, which we add to and refine continuously.
We then incorporate scientific meteorological, geological, and hydrological modeling into the location risk analysis. This helps us to understand the possible losses in rare events. We must also understand our models’ weaknesses and strengths. Does the model include the Tsunami loss? Did we factor in the different, associated risk features identified during our site investigations?
Earthquakes frequently cause fires, for example, due to ruptured gas lines, or gas and other material flowing from the damaged site to another industrial facility. A site may have the benefit of a sprinkler system. However, without bracing, this system could become inoperable, just when you need it to stop a fire from a ruptured gas main. Our engineers will recommend bracing, as well as gas shut-off valves.
Historically, when emergency services were not equipped to respond swiftly and fully—very often the case for an LP-HI event—fires have consumed industrial complexes and reached into surrounding towns, resulting in a tragic loss of human life and heavy blows to business and CI.
Risk engineers ask: How fast and to what degree are emergency services prepared to respond in an LP-HI event? Are employees trained in the best ways to seek safety in a disaster? How would they react if the fire station were damaged, the water main system is ruptured, and the fire water is not available?
Once the humanitarian rescue is complete, how quickly will the site be safe enough for claims adjustors to access, so that insurance can pay out? If a state of emergency requires the area to be cordoned off, as with Hurricane Katrina and the Fukushima tragedy, access to the damage zone will be impeded, and losses will escalate.
In the aftermath of a disaster, the answer to these questions can make the difference between CI resilience, long-term setbacks, and utter failure to recover.
What about regional concentrations of certain industries, like Dr. Giardini’s example of the Jura watchmakers?
The economic impact of an LP-HI disaster on a regional industry cluster can definitely hurt a national economy, potentially long-term. Depending on the sector, it can also set back the global economy.
The tragedy in Fukushima followed by the floods in Thailand actually taught us several lessons about CI in the broader economic context. Japanese car manufacturers transferred manufacturing to Thailand plants, only to be hit by the flooding the same year. The Thailand floods also hit the hard-drive industry, causing a worldwide shortage of these important technology components.
These were vital lessons, not only for the industries concerned, but in terms of global CI. Which government, business, or individual is not dependent on a computer? Computers are part of our global CI. Do you have a tested disaster plan in place for the loss of computer power or connectivity to the internet?
How can insurers help improve the resilience of CI in an LP-HI event?
Humans have very short memories, especially on the geographical timeline Dr. Giardini mentioned. We all know the story of Pompeii, yet the fertile slopes of Mount Vesuvius proved irresistible to returning settlers. Mount Vesuvius has erupted numerous times since, although never to the same extent as in AD79. However, it could have another violent eruption, flatten industry in the area, and inflict massive financial losses.
Fortunately, we now have a lot of data, and the ability to make it available to everyone, so when human memory tricks us, we can refer to evidence. But the data will never answer all the questions, there will always be surprises, events more severe than we have seen before, Earthquake faults we did not know existed, a combination of weather patterns that conspires to create a mega event. Data improves our models and helps us to make better decisions. We need to ensure we connect these groups to a much wider audience and not analyze risk in silos.
Risk-engineering and underwriting teams in particular have accumulated a wealth of historical data, on which we base advanced loss-prevention strategies. That data includes highly confidential and relevant site data from clients, which we anonymize and run through analytical and statistical models. It also includes precise location data collected over many decades.
Insurers must continue to innovate, however. We need to keep developing advanced algorithms to collect new data more rapidly, and integrate it into our analytics. We should incorporate big data—that is, real-life reports from people at disaster sites, even if they are not our clients. We should share information! The age of secrecy is over. The sooner we embrace the advantages of information transparency and sharing, the better. Of course, I don’t mean confidential, customer-specific data, but the analyses and lessons learned from it.
We can use our data to help educate businesses and governments, and assist in developing disaster recovery and CI continuity plans. This is important, because authorities sometimes lack the information they need to make decisions. Building codes, emergency-service procedures, and even public education can only address potential disasters if authorities have the necessary data.
We need an ongoing risk dialogue between business, states, insurers, and public stakeholders. Everyone needs to understand the risks we face, and the interdependencies between them. Then we can make effective decisions about how to move forward.
Insurers have a stake in the swift resilience of our communities in an LP-HI event. If key infrastructure, like hospitals, emergency services, or bridges fail (insured or not) Insurers are impacted. If the businesses we insure do not get the help they need, physical movement of supplies and products is hindered, and recovery is delayed. Insurers bring a lot of disaster-recovery experience to help clients plan their own responses.
Would you say, generally, that we need to take a longer view of the impact of LP-HI Events on CI?
Yes, the fact is that rare disasters are more frequent than before, and they warrant even more consideration in business planning.
Macro-risk decisions today may benefit from more consistent, globally minded standards, to ensure business longevity. It often seems easy, in the short term, to move facilities to cheaper, less complicated locations. But those cost savings could be deceptive. If building standards are not aligned with the real risk in the area, it can lead, as we have seen, to death, reputational damage, and supply-chain disruption. A building with robust safety margins costs on average 5% more to build, true, but those better design parameters could make the difference in an LP-HI event.
To ensure our global CI and business longevity, short-term balance sheets and political expediency must be balanced with better-informed LP-HI risk assessment. Hazards are here to stay, but intelligent, informed legislation and recovery plans will make global CI more resilient. This could save countless lives, businesses, and even the global economy.
Want to know more? You can reach Alan on: email@example.com