Dominik ScheibenRisk Consulting Manager, Environmental risks, AXA XL
February 7, 2022
Environmental liability insurers rely on environmental assessments to underwrite their coverages and support clients’ efforts to better manage and mitigate their environmental liability exposures. When the policy covers only a handful of sites, the assessment process usually isn’t overly burdensome. However, what about clients that have tens, hundreds or even thousands of locations?
For multinational companies with large numbers of sites, AXA XL uses a three-stage risk assessment process. First is a portfolio-level analysis to develop an “environmental sensitivities score” for each location; that’s an aggregate measure showing the relative environmental liability risks for every facility in the portfolio. Next, additional desktop research is conducted on highly scored sites to better understand the exposures and potential liabilities. Finally, on-site surveys are carried out at facilities with the most significant environmental sensitivities; the objective here is to develop loss prevention and mitigation strategies to lessen the risks.
The environmental sensitivities scores that are the foundation of the assessment are developed by collecting and analyzing data on four dimensions: the source of the environmental pollution, the pathway by which a pollutant could travel, the receptor it could impact, and the baseline conditions at the site. The source could be a tank or pipeline leak, a toxic gas emission or a chemical spill. The pathways are soil or groundwater, surface water or the air. The receptors include nearby urban or agricultural uses (i.e., third-party liability exposures), water bodies or protected natural areas. The baseline conditions reflect pre-existing pollution/contamination levels of the air, surface water and groundwater surrounding the site.
In other words, the inputs to the environmental sensitivities score are a variety of dynamic variables drawn from a wide range of sources.
For example, we recently conducted an environmental sensitivities assessment for a major cosmetics manufacturer’s European operations. It identified three locations where the composite environmental sensitivity scores were considerably higher. One is next to a major river and is surrounded by natural areas with high biodiversity, and another is located in a productive aquifer with high surface water permeability. The third is within a densely developed district where the third-party liability risks are considerable, e.g., from a fire, explosion or large spill. While the factors contributing to the relatively high scores vary considerably, the assessment showed that these three sites warrant more in-depth analyses, and the company should consider further remediation measures to minimize the risks.
Given all the variables that must be considered, it is not surprising that the portfolio analysis often is time-consuming and labor-intensive. Insurers—or an environmental consultant working on behalf of the client—typically consult area maps and manually collect data from diverse sources to assess a site’s environmental sensitivities and potential exposures. This manual approach is cumbersome, and the ratings assigned to individual sites could be applied inconsistently when different data sources are used to evaluate many operations in geographically dispersed locations.
In response to these inefficiencies and shortcomings, AXA XL has created an environmental sensitivity tool that enables us to assess the risks and prioritize the needs faster and more consistently. Like the Risk Scanning solution we launched last year, this new tool uses machine learning applications and artificial intelligence (AI) capabilities to collect, synthesize and analyze massive amounts of data. The datasets include structured information from public agencies, streamed data from sensors or meters, climatological data and customer-specific information like location and site activities.
The tool currently includes comprehensive datasets for the geographic region of Europe, including the UK, and the example I cited above was one of the first test cases.
The benefits of this new tool for clients are twofold. First, it augments and standardizes a client’s existing understanding of its environmental liability exposures at individual locations. That is, companies where environmental liabilities are a potential concern typically have at least a basic knowledge of a site’s most significant environmental issues. However, when the portfolio comprises many facilities in diverse locations, the client’s data on various sites often will vary in age and have been generated by different external consultants using dissimilar sources and methodologies. Thus, it is difficult, if not impossible, to accurately prioritize the sites where the environmental exposures are most significant.
Second, the end product of the assessment is a detailed and consistent evaluation of every location within a portfolio based on the latest data from the most sophisticated sources. That means the assessments produced by our new tool will help ensure that clients’ environmental liability policies fairly and accurately reflect the exposures. With the economic headwinds facing most companies nowadays and a hardening insurance market, this more precise assessment of potential environmental liabilities could prove highly beneficial.
In sum, the Risk Scanning solution and environmental sensitivities tool are, in my view, good examples of how insurers can harness “big data” to help clients more readily evaluate and prioritize their exposures and, in turn, better manage the risks and create more resilient businesses.