PFAS concerns mount

Perfluoroalkyl and polyfluoroalkyl substances (PFAS) have gained significant attention in the news lately due to their persistent nature and potential health impacts. PFAS refers to a class of chemicals discovered in 1947 renowned for resisting heat, water, and stains—substances that are found in everything from firefighting foam to nonstick pans to raincoats. Their strong carbon-fluorine bonds resist degradation, earning them the nickname “forever chemicals.” Unfortunately, the strong and stable qualities that have been central to the many use cases of PFAS allow them to persist in the environment and migrate to drinking water and human bloodstreams. The health implications associated with PFAS exposure include various cancers, liver damage, endocrine disruption, and immune system effects, among others. Despite ongoing scientific debates, a substantial body of research has linked PFAS to these adverse health outcomes, resulting in extensive litigation.

Since the first major lawsuit concerning PFAS exposure was filed in 1999, the number of PFAS-related lawsuits has increased to more than 10,000 complaints involving almost 500 companies across many industries. Known settlements to date are over $16 billion, a figure that is likely to grow substantially as more complaints continue to surface. The litigation landscape is complex and involves multiple types of insurance coverage, such as general liability, pollution legal liability, and directors and officers (D&O) coverage. Ongoing coverage disputes hinge on policy language and limited jurisdictional precedent, further complicating the ability of insurers to quantify their potential liabilities.

States have been addressing concerns over PFAS drinking water contamination for a few years by establishing regulatory standards or guidance for certain PFAS chemicals, totaling over 20 states at the time of this publication. In April 2024, the U.S. Environmental Protection Agency (EPA) took a significant step by establishing a countrywide regulatory standard, setting maximum contaminant levels (MCLs) for various PFAS compounds in drinking water, with the lowest MCL set at 4 parts per trillion (ppt) for perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS).

Litigation over PFAS water contamination was well under way prior to the EPA drinking water regulations, with plaintiff claims supported by state regulations and scientific studies linking PFAS to health concerns. However, the EPA rule is significant as it now puts the impetus on water districts nationwide to test for PFAS instead of relying on the EPA for testing. If tested PFAS levels exceed the MCL, water districts are then required to remediate their drinking water, which can be a costly endeavor. Funding for these projects may come from a variety of sources, including the federal infrastructure bill, state grants, and increased payer rates. While these sources of funding will certainly help, the proposed amounts will likely be nowhere near the total amount it will cost to undertake the necessary remediation nationwide.

Many suits have already been brought forward by water districts against chemical producers and other manufacturers that added PFAS to their products. A significant portion of these suits were consolidated into a multi-district litigation (MDL) centered on contamination from aqueous film-forming foam (AFFF), a firefighting foam used to fight liquid-based fires, as the use of AFFF creates runoff that can seep into water sources. To date, the water districts have reached settlements with four companies, as detailed in the table in Figure 1.

The water districts in the MDL brought suits before the EPA regulation was finalized. It is likely that the amount of affected water districts and subsequent litigation will only grow from here as new districts discover PFAS and districts that already settled with the chemical giants find other parties that are also partially responsible for the contamination.

Hesitancy to quantify PFAS liabilities

Many insurers have been hesitant to quantify potential PFAS liabilities and set aside reserves for PFAS claims, often suggesting that PFAS liabilities are inestimable due to the difficulty in producing reliable estimates using industry-accepted methodology. Traditional actuarial methods, both triangle-based and those tailored for asbestos, pollution, and health hazard (APH), rely on stable historical claims data to predict future liabilities. The evolving nature of PFAS claims means there is insufficient historical data to apply these traditional methods effectively. PFAS-related injuries take time to manifest and possible exposure dates span decades, further complicating the use of conventional actuarial approaches.

The hesitancy to quantify and reserve for PFAS liabilities also stems from uncertainties in the legal and scientific domains. Most notably, the development of case law regarding PFAS coverage is ongoing, with outcomes varying based on jurisdiction and the specific policy language. The Insurance Services Office (ISO) rolled out forms that broadly exclude PFAS from various coverages, including commercial general liability, umbrella, and business owners in 2023 and some insurers had already developed their own exclusions a few years earlier. While many policies today include PFAS-specific exclusions, older occurrence-based policies may only contain general pollution exclusions, or none at all, leading to disputes over coverage applicability. Beyond exclusions, policy triggers and the duty to defend are frequently contested in courts. These ongoing coverage disputes make it difficult for insurers to estimate potential liabilities accurately. Additionally, the scientific understanding of PFAS exposure and its health effects continues to evolve, adding another layer of uncertainty.

A full discussion of actuarial approaches for PFAS can be found in the Milliman white paper “The ‘Forever Chemical’ Risk Is Now: An Actuarial Reserving Study Can Help Insurers Prepare for PFAS Claims and Litigation,” published in November 2023.

Milliman’s approach to quantifying PFAS liabilities

To address these challenges, Milliman built a dedicated team that has spent years developing a proprietary model for estimating PFAS liabilities, following actuarially accepted approaches. This stochastic methodology accounts for the numerous uncertainties inherent in PFAS exposure and litigation by calibrating key assumptions with the latest data and incorporating knowledge from actuarial, legal, claims, and environmental experts. By simulating tens of thousands of potential scenarios, a distribution of probabilistic outcomes can be generated.

Milliman’s PFAS model provides actuaries with a robust and flexible framework that can be incorporated into various actuarial workflows, including reserving, pricing, underwriting, and enterprise risk management (ERM). The model is well equipped to produce actuarial central estimates and tail scenarios, including incurred but not reported (IBNR) liabilities for claims that are yet to emerge, as well as incurred but not enough reported (IBNER) losses for individual insureds who have already brought PFAS claims, offering a clearer picture of potential liabilities.

The flexibility of this approach allows for continuous refinement of assumptions as new information becomes available. This adaptability ensures that the estimates remain current and reflective of the latest scientific findings, remediation techniques, settlements, regulation, and legal precedents.

Quantifying potential PFAS liabilities today allows companies to develop strategies to mitigate the risk:

• Understanding PFAS risk and the potential threat it poses to company value through ERM modeling
• Avoiding new PFAS exposure through effective pricing and underwriting on prospective policies and retrospective reinsurance coverage
• Reducing the likelihood of significant adverse reserve development in the future by considering PFAS liabilities in reserve estimates.

Each company will have a different level of exposure to PFAS and, therefore, will be at different stages in the process of understanding their risks. In all cases, quantifying PFAS today equips companies with the tools to protect their bottom line before PFAS liabilities can grow.

Milliman water district remediation estimate

As mentioned earlier, water districts in the United States have been at the forefront of the PFAS litigation over the last few years, as districts seek funds to remediate drinking water that has been contaminated with PFAS. We have applied our stochastic modeling approach to develop estimates of the drinking water remediation cost for each individual water district. Milliman’s model produced a range of possible estimates and determined that a reasonable total expected cost to remediate drinking water systems for U.S. water districts is $120 billion to $175 billion.

The current estimate relies on the contamination thresholds utilized in the AFFF MDL and considers the presence of around 30 different PFAS compounds to determine whether a water district has PFAS contamination that rises to a level where it could pursue litigation.

Model simulation results varied considerably at this stage due to the uncertainty surrounding key stochastic assumptions, such as frequency of PFAS contamination and the cost of remediation systems. Some examples of situations that could materially impact the model assumptions include, but are not limited to, efficiencies gained in remediation technology, demand surge for remediation technology, water districts pursuing alternative strategies to meet remediation needs (e.g., utilizing different water sources), and improved sampling methods that are more sensitive to detecting PFAS, among others.

Each water district has unique risks, and our model separately calculates the probability of requiring future remediation along with the expected cost to remediate the drinking water. Many water districts still need to be tested to determine whether they have exposure to PFAS, a process that is estimated to take years, and the number of PFAS compounds previously tested by district varies widely. To address this, our model was developed through extensive testing on over 30 explanatory variables to produce the estimated PFAS levels at each water district. Modelers considered geospatial, regulatory, and industry data from numerous sources to determine an expected remediation cost estimate for over 140,000 U.S. water districts. Some water districts have an expected remediation cost as low as a few thousand dollars, while others are expected to be in the hundreds of millions. Figure 2 provides an example of how modeled probabilities vary due to the unique aspects of each water district and the surrounding area.

The water district drinking water remediation estimate discussed above represents one of several analyses that Milliman has been conducting to estimate potential types of liabilities related to PFAS (e.g., environmental liability, bodily injury). This estimate should not be viewed as an insurance industry ultimate liability for PFAS. Milliman is in the process of quantifying a liability estimate specific to the insurance industry. Some insurance policies are likely to be impacted by multiple types of claims that are not fully captured in this water district estimate. Therefore, the insurance-specific estimate will be released later as it is dependent on the other liability estimates mentioned above.

In summary, PFAS-related risks present significant challenges for the insurance industry and corporations alike. Our innovative methodology provides a reliable means to manage these uncertainties today. By leveraging advanced modeling techniques and continuously updating key actuarial assumptions, the estimation of PFAS liabilities can be improved in real time as this emerging risk evolves.

Special thanks to the following individuals for their significant contributions:

Carl Ashenbrenner, FCAS, MAAA
Easton Becker
Halle Dorsey
Laura Eaton
Alyssa Grove, ACAS, MAAA
Brekk Hayward, ASA, MAAA
Joe Michels, ACAS, MAAA
Jaxon Parmley
Karen Rinehart
Mason Spitz, ACAS, MAAA
Kyle Stowe

The information contained within is for general information or educational purposes only and shall not be considered as specific advice. As such, no action or decision should be taken solely on the basis of the information set out herein without obtaining specific advice from a qualified advisor. The authors are members of the Casualty Actuarial Society and consulting actuaries with Milliman. The estimates, thoughts and opinions contained herein are those of the authors.