Insurance Recovery for Manufactured Gas Plant Liabilities

Fortnightly Magazine - April 15 1997

Valuation, optimization and settlement strategies

oth gas and electric utilities face a variety of environmental issues arising from more than 1,500 former manufactured gas plant (MGP) sites, which supplied a major source of energy in the United States from the early 1800s to the mid-1900s. Using the standard operating procedures of the day, MGPs created and often disposed of byproducts such as coal and oil tars, tar/water emulsions, sludges, spent oxides (including cyanides), lampblack, ash and clinker.

Today, simply identifying these MGP sites and their history of operation and ownership can be a significant chore. It calls for review of company records and historical association proceedings, industry directories, journals and other documents that describe individual plant operations and site layouts. MGP site environmental liability issues include the type and extent of contamination, the risks posed and current and future use of the site. State laws and regulations also must be considered. Site remedies and expected costs pose another issue, including how those costs are best allocated. Other potential liabilities include property damage and toxic tort claims. %n1%n

Once a company identifies these liabilities, questions soon follow concerning recoverability of costs from insurers, such as: How to place a value on a potential insurance recovery or damage award? How to maximize recovery through litigation or settlement? And, how can coverage disputes be mediated to avoid litigation?

Valuing Potential Recovery

Utilities contemplating litigation against insurers to recover costs associated with MGP cleanups face difficult business decisions. By analyzing the expected recovery from litigation, a company can determine if such an action is worth the effort by comparing its anticipated litigation costs with expected recovery. The analysis of expected recovery can also help determine appropriate settlement values and document prudence.

In multi-site/multi-insurer litigation, many complex factors and uncertainties complicate evaluation of the expected value of litigation. These factors include policy limits and language, site costs and the state's case law interpretations. The many uncertainties and complexities make it seem difficult, if not impossible, to predict the value of an insurance suit. This uncertainty could discourage utilities from taking legal action or lead them to accept settlements that are far less than fair. However, this need not be the case.

By combining the tools of decision analysis %n2%n with the experienced judgment of attorneys and engineers, utilities can evaluate the merits of alternative litigation strategies. A probabilistic approach assesses the likelihood and cost of each possible litigation outcome to estimate both the expected value and the risks of litigation. Insured parties have used this information to determine whether to sue, to compare the value of global versus separate insurance actions, to select litigation forums, to evaluate settlement offers and to develop testimony for trial. This approach consists of five steps:

Step 1:

Organizing Policy Data

The first step in the process is to locate all relevant insurance policies and develop information about both their individual characteristics and the structure of the insurance coverage hierarchy in each year. Finding all of the pertinent policies may be difficult for MGP sites that could have been owned and operated by various companies extending back as far as the 1800s. An insurance archaeologist can assist in the reconstruction of policy histories.

Once all relevant policies are located, they must be examined carefully to identify the features that will bear upon potential coverage. This includes identifying the insurer, the period covered, the coverage layer (e.g., primary, excess, umbrella) and the responsibility for coverage within a level (e.g., quota share, joint and several). In addition, companies should determine the following: the deductible, per-occurrence limit, aggregate limits, occurrence definition, pollution exclusion language, treatment-of-defense costs versus damages and self-insured retention level.

Step 2:

Evaluating Site Facts Relevant to Coverage

A number of site-specific facts will affect coverage. Dates of site events combined with policy trigger arguments will determine which policies are pertinent for a particular site. Other site features will determine whether policy criteria for coverage have been met. Consequently, site records and knowledgeable individuals must provide site information that relates to the following issues:

• Potential trigger dates. When did the site operate? When were shipments made or contaminants released? When was a problem discovered? When was a claim made?

• The owned property exclusion. Is there damage or the threat of damage to public property or to the property of others?

• "Unintended or unexpected occurrence" language. Were damages foreseeable?

• "Sudden and accidental"

language. What caused the contamination? Were there abrupt contamination events at the site? %n3%n

• Notice requirements. When was each carrier notified?

• Existence of claim or suit. Has a suit been filed that would trigger defense obligations? What incurred expenses are considered defense costs?

Step 3:

Estimating Site Costs

The expected recovery for any site clearly depends on its potential environmental costs which may include studies, mobilization, construction, operation and maintenance, project oversight, natural resource damages, toxic tort and third-party property damage awards, government oversight and legal costs. Past costs are documented through existing records and the knowledge of company employees, technical consultants and attorneys familiar with the site.

Estimating future costs is often more complicated because of substantial uncertainties regarding key cost factors. These factors include the type and extent of soil remediation, the type and duration of groundwater treatment, the timing of remedial expenditures and the company's ultimate share of total site costs for unowned sites.

In cases where there is substantial uncertainty, a decision analysis approach enables development of future cost estimates. A cost decision tree can prove useful. (Exhibit 1 provides a simplified example of a remedial cost decision tree for an MGP site.) Once costs and probabilities are attached to each branch of the tree, a probability-weighted expected value can be calculated, capturing the full distribution of potential costs at the site. The tree is developed to reflect all major uncertainties such as the timing of expenditures, cleanup goals, allocation of costs among responsible parties and other cost issues.

Step 4:

Estimating Coverage Likelihoods

An insurance carrier typically raises a number of defenses as to why it should not provide coverage. For a utility to be successful, all of the applicable coverage issues must end up resolved in its favor. Therefore, the likelihood of winning or losing on each potential coverage issue must be assessed. The first step is to identify the key defenses that insurers can raise, such as:

• Was timely notice given?

• Who owns the property?

• Was the damage intended or expected?

• Was the release intended or expected?

• Are cleanup costs damages?

• Was the damage sudden and accidental?

• Is there a suit or claim?

Once these defense issues are identified, attorneys experienced in insurance law must evaluate these defenses in light of site facts to determine the likelihood that each issue will be resolved in favor of the insured. This assessment takes place for each site and for each policy that could be triggered. If the utility has a choice of where to file suit, then it should evaluate the coverage likelihoods for every potential forum in light of the past rulings in each. A hypothetical example can illustrate this process (see exhibit 2).

Step 5:

Assessing the Expected Value of Litigation

In the final step, information concerning policy characteristics, site facts, site costs, and coverage likelihoods is integrated to develop an estimate of the expected value of recovery through litigation. A key element of the litigation strategy is to determine which trigger and allocation theories will maximize expected coverage. The policies that are triggered and the ways in which costs are allocated to policies generally vary according to three factors:

• Date of Injury. Is this determined by the dates of site operation, the dates of shipment or disposal, the date of a particular contamination event, the date of discovery of the problem, or the date of a suit or claim?

• Occurrences. Is there one discrete occurrence or are there multiple or continuous occurrences?

• Cost Allocation. Can costs be allocated to a single policy year or must they be prorated across policy years?

Expected recovery typically varies greatly with the trigger and allocation theory. A date-of-discovery trigger will tend to access policies in more recent years that have high coverage limits, but unfavorable pollution exclusion language. A multiple occurrence theory can often trigger policies that predate sudden and accidental language or the absolute pollution exclusion. These policies provide a greater likelihood of

coverage but typically have lower per-occurrence and aggregate coverage limits. Furthermore, if costs must be prorated across all policies triggered by a multiple occurrence theory, then an accumulation of deductibles may diminish potential coverage.

The estimate of expected recovery from a multi-site suit under each possible trigger and allocation theory starts with an analysis at the individual site level. The expected recovery is estimated for each policy that may provide coverage. The results are then summed across all sites, taking into account the impact of aggregate policy limits, to derive the total expected insurance recovery from litigation.

Optimizing Recovery Strategy

The detailed information developed in valuing an insurance recovery case as described above offers insights into ways to optimize recovery and manage risk. %n4%n Through this analysis, utilities can identify the forum, trigger theory and allocation theory that maximize expected coverage. For example, the detailed analysis may reveal that filing of the case should take place in the state where the policy was signed, rather than the state in which the MGP sites are located. The analysis may indicate that a date-of-site-demolition trigger is advantageous, or it may point to recovery optimization with a continuous trigger where the utility can pick the policy year to trigger, rather than proration across all triggered policy years. The analysis will also reveal whether site-by-site or global litigation is preferable.

In addition, because the analysis is conducted on a policy-by-policy basis, the insured parties can determine the amounts they can expect to recover from specific carriers and specific types of coverage (e.g., primary, excess) in multi-insurer cases. It is possible that higher recovery will follow from using different trigger/allocation theories for different carriers. Thus, the analysis may indicate that during settlement negotiations with various carriers, the utility will trigger different years and argue for different allocation scenarios to maximize overall recovery. In addition to the expected value from litigation, it will often prove useful to identify the litigation arguments that maximize recovery for each carrier as preparation for settlement discussion. This will allow utilities to determine what types of settlements to accept to avoid or truncate costly litigation.

Another key feature of the valuation exercise is a better understanding of the litigation risk involved in an insurance recovery case. The expected value of overall recovery is derived from a weighted average of thousands of possible outcomes. These possible outcomes range from scenarios in which recovery is less than the litigation costs to scenarios in which nearly full coverage is obtained. This wide range of potential outcomes creates risk for both utilities and insurance companies. It is important for utility managers to understand not only the expected value of the case, but the range of potential recovery and the risks involved.

The information collected in the valuation analysis can be used to develop a distribution of possible recovery outcomes. This distribution shows the likelihood that recovery will be within a particular dollar range (see exhibits 3 and 4 for examples).

This type of risk analysis is very useful in determining litigation and settlement strategy. First, it identifies key vulnerabilities. Second, it communicates to upper management why settlement at levels different from the expected value might be acceptable. Third, the analysis provides documentation to back up the prudence of management decisions regarding settlement and litigation.

Mediating Insurance Disputes

Both utilities and insurance companies can benefit by expeditious settlement of insurance claims for MGP liabilities. However, settlement is often frustrated. Focusing solely on the "bottom line" creates impasses, in which insureds claim coverage for the maximum possible amount and insurance companies categorically deny all coverage. Inability to estimate the value of a claim due to the complexity and uncertainties involved in the valuation also stymies settlement. In both situations, the parties cannot resolve the dispute because they have not broken down the valuation problem into its component parts, and therefore cannot even begin to discuss the bases for settlement.

One way to avoid the impasse, and hopefully the "scorched earth" litigation that often follows, is to develop a valuation framework where both the utilities and the insurance companies agree on the approach, and work together to break the problem into component parts. In this approach, the parties jointly assign costs and likelihoods to each component. For those components where there is a major difference in opinion, ranges can be assigned and sensitivity analyses run to determine whether the differences will significantly affect the results. In this way, the parties can quickly determine the key points of the case, and those where they agree and disagree.

Parties often have many more areas of agreement than disagreement when insurance company "discount factors" are generally close to the insured's likelihoods of recovery. And areas of strong disagreement often do not significantly affect the recovery outcome, so do not have to be debated in detail. For example, debating the meaning of "sudden and accidental" pollution exclusion language may not prove necessary if the utility is unlikely to trigger policies that contain that language. Through careful analysis, only those areas with differences that significantly affect the recovery outcome require further research and additional negotiation, allowing the parties to focus their negotiations in an efficient manner. t

Gayle S. Koch and Kenneth T. Wise are principals with The Brattle Group, an economic, environmental and management consulting frim. Phil Hanser is a senior consultant with The Brattle Group.

1For additional information on who should pay for environmental liabilities, see "Environmental Cleanup Liabilities" by A. Lawrence Kolbe and William B. Tye, Public Utilities Fortnightly, January 1, 1992; "How Far Back Should Prudence Tests Reach: by William W. Hogan and A. Lawrence Kolbe," Public Utilities Fortnightly, January 15, 1991; and "Managing Environmental Liabilities at Manufactured Gas Sites" by Gayle S. Koch, Paul R. Ammann, and A. Lawrence Kolbe, presented at the American Power Conference, 1994.

2Decision analysis is a logical framework that combines probability theory and management science to structure and analyze the alternatives and uncertainties present in complex problems which exist in both business and litigation environments.

3The relevance of this information will vary by forum, as courts have differed in their rulings concerning whether "sudden and accidental" language has a temporal context or simply requires that a release be unintended and unexpected.

4Often, the insurance recovery analysis provides insights into management of the site's environmental liabilities. Thus, an assessment initially directed toward litigation can be used as a strategic environmental liability management tool. See "Using Decision Analysis to Manage Environmental Costs," by Gayle S. Koch, Paul R. Ammaan and Kenneth T. Wise, Journal of Environmental Regulation, Summer 1995.


Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.