West Virginia's Brownfields Legislation:

•  Which cleanup standards should be employed?
•  How should environmental risks be assessed?
•  To what extent should the public be involved?

    This article describes the history of the West Virginia Brownfields law, the activities of the various parties and interest groups that became involved, how the major issues were resolved, and the voluntary redevelopment process that was established. Finally, issues that remain to be addressed during rule-making and implementation will be discussed, along with recommendations for resolving those issues.
 
 
    Brownfields Redevelopment Policies

    Land use decisions, and particularly decisions about potential industrial land use, are often contentious. Industrial development involves investment of significant financial resources and can result in a substantial boost to the local economy, but it also often brings concerns about traffic, pollution, property values, and its overall impact on the quality of life.

    In West Virginia, suitable land with adequate infrastructure for development is scarce, even though old or abandoned properties are common in many counties. Proposed industrial development in or near established residential or rural neighborhoods often meets with resistance from local residents worried about urban sprawl, pollution, traffic, noise, and the disruption of their established life styles. The brownfields concept, where old industrial sites are cleaned up and reused, is an attractive alternative because it returns industrial development efforts to areas already dedicated to industrial use, thereby allowing economic development while avoiding conflicts with established residential or rural areas.

    The three most common incentives used to encourage brownfields redevelopment are modified cleanup standards, financial assistance (often in the form of a low interest loan or grant) and a total or partial waiver of liability to the property owners. Typically, the property owners or developer must clean up the contamination at the site only to levels commensurate with the proposed commercial or industrial use. These cleanup levels are based on an analysis of sitespecific environmental risks, and are accompanied by a waiver of liability and/or official certification of completion to assure the property owner that additional environmental remediation will not be required in the future. By removing the uncertainty associated with the potential liability for future cleanups on the site, the property owner can better predict the cost of developing the site, making it more likely that it will be developed. Local economic development authorities, for example, report that many small to medium-sized companies looking to expand in their areas have indicated that they have gone elsewhere because they could not afford to wait a year or two to construct a building on a new site. By making existing buildings on brownfields sites more economically attractive, local development authorities can significantly expand the number of sites available for new companies to consider.

    Although stringent cleanup standards and concerns about liability have hindered redevelopment of brownfields sites, these policies force businesses to take pollution prevention seriously. This is because preventing the release of pollutants is usually substantially cheaper than cleaning them up. In fact, it is the fear of cleanup costs, much more than any fear of regulatory enforcement, that pushes companies to prevent releases in the first place.

    Much of the controversy about the brownfields proposals revolves around the issue of whether weakened cleanup standards and waivers of liability might encourage companies to relax their pollution prevention efforts, thereby resulting in more contaminated sites than are cleaned up by the programs. Poorly applied brownfields policies could allow existing industries to "walk away" from their polluted sites, leaving behind economically devastated communities and highly polluted environments. This scenario is the exact opposite of what brownfields policies are intended to achieve, but it is a real possibility if those policies are not carefully structured to account for the real-world behavior of environmental regulators, industrial operators, developers, and financiers.

Legislative Models for Brownfields Programs

    The federal government has only recently become involved in brownfields redevelopment. In 1995, the U. S. Environmental Protection Agency (EPA) offered financial incentives for brownfields pilot projects in ten cities. The program expanded to an additional 20 projects in 1996. These pilot projects encouraged local initiatives and communitybased redevelopment efforts, rather than rigid, federally mandated remediation programs. EPA emphasizes community involvement, environmental justice, and new technological and managerial processes by providing regulatory flexibility and modest start-up grants.

    The states are also relatively new to brownfields redevelopment. Minnesota and Illinois started the first state brownfields programs in 1988 and 1989, respectively (Powers 1995). In Minnesota, the Minnesota Pollution Control Agency oversees every stage of the cleanup. But instead of establishing rigid cleanup standards for the redevelopment of each site, it provides both the community and the property owner/developer a variety of "assurances" that the cleanup is in compliance with the law's intent. Each site's cleanup, for example, must take into account both the contamination on the site as well as potential contamination from off-site sources. Also, the property owners are issued a certificate of completion after the site has been cleaned up. Minnesota's program has been popular with both businesses and environmentalists because it is a less expensive and faster alternative to the state's more rigid SuperFund program.

    Illinois also does not mandate rigid cleanup standards for all sites. Instead, it allows the property owner to hire specialists to conduct a risk-based analysis for the site. The analysis determines, subject to the Illinois EPA's approval, how clean the site must be to meet the law's intent. Once that level is achieved, the property owner is provided a waiver of future liability (a Covenant Not To Sue).
 
    Ohio has adopted an entirely private sector approach. Private certified professionals are authorized to conduct cleanups and certify their completion. In addition, Ohio's program explicitly limits agency or public oversight by making information associated with cleanups legally privileged, nondiscoverable, and inadmissible in court proceedings. Public involvement is not required except for limited public comment when a variance to standards or a consolidated permit is requested by the property owner.

    Other states have adopted legislation specifying special corrective action criteria for brownfields development using risk-based procedures and standards. They differ significantly in how the risks are evaluated, the kinds of sites eligible to participate, the degree of statutory guidance offered, the level of regulatory agency oversight, the kinds of liability protection offered, and the level of public involvement afforded to citizens.

The Brownfields Debate in West Virginia

    Economic development has been a major issue in West Virginia for a very long time. Attracting high paying, full-time jobs with benefits has been one of the highest priorities for West Virginia lawmakers for decades. Since the industrial sector provides these kinds of jobs, attracting industrial development that is "environmentally friendly" is a prime political objective in West Virginia. Targeted industries have frequently cited severe infrastructure limitations as a major consideration in their industrial siting decisions. These limitations were highlighted in a 1994 report from the West Virginia Economic Development Authority which identified only two sites in West Virginia larger than 50 acres with adequate infrastructure to support a major industrial development (Dilger and Witt 1994).

    One result of this often-cited report was the passage of the 1994 Infrastructure and Jobs Bond Amendment which provided $300 million for new roads, water and sewer construction. These infrastructure enhancements are targeted toward communities not meeting basic services, but a portion is specifically earmarked for infrastructure needs to attract economic development. The Infrastructure Amendment not only expanded the state's fiscal capacity to address its infrastructure needs, but it also allowed the state legislature to focus increased attention on measures needed to spur redevelopment of existing industrial sites.

    Concurrent with the 1994 Infrastructure and Jobs Bond Amendment efforts, state environmental agencies began a discussion with representatives of state business and environmental groups regarding the use of risk assessmenVrisk management procedures in decision-making regarding cleanup levels at a variety of contaminated sites. Eli McCoy, then Deputy Director of the West Virginia Division of Environmental Protection (DEP), recognized "that this concept is viewed with some degree of skepticism by certain citizen segments of the population, and that there must be a fairly broad base of support for the use of any such methodology. If either side is not comfortable with the motives of the agency, or if they feel the agency process does not take into account their particular system of values, agency decisions could end up in the court system. Litigation of cleanup levels does nothing to advance our goal, which is adequate cleanup at sites as quickly as we can get it done" (McCoy 1994).

    A series of meetings held during June and July of 1994 clarified the participants' concerns. DEP personnel were concerned that existing statutes and rules did not give them adequate guidance for determining cleanup standards less stringent than pristine background levels. They pointed out that some cleanup efforts continued indefinitely, even though there was no perceived risk to human health or the environment and, in some cases, even though it was known that the technology available was inadequate to ever achieve a pristine zero background. Moreover, they lacked statutory authority to provide the private sector with any assurance that a site did not need to be cleaned up.

    DEP cited the use of air strippers to remove gasoline from contaminated ground water as an example. Although gasoline concentrations in the ground water were still detectable after the use of air strippers, the treatment technology was incapable of lowering concentrations significantly. And even though the particular aquifer was not being used for drinking water, DEP did not feel they had the authority to authorize the termination of the pump and treat program. This resulted in cleanup resources being expended at a site where there was no significant risk, and little likelihood of further environmental benefits.

    The West Virginia Manufacturers Association (WVMA) extended this position by arguing that overly stringent standards resulted in a misappropriation of societal resources and an overall reduction in the degree of environmental protection that could be achieved, given a more reasonable standard. They identified the following three obstacles to industrial land reuse:
 

• New property owners are automatically held liable for cleaning up past pollution on a site, even if they were not responsible for the contamination.
   
•  Cleanup standards automatically assume that the site must be cleaned up to a pristine condition, even if it is to be used for a heavy manufacturing plant and not a daycare center.
   
•  State tax, liability, and cleanup policies provide greenfields sites with lopsided economic advantages over brownfields sites.

    The WVMA argued that the state needed to adopt a risk management standard similar to those under consideration in nearby states that would make brownfields sites more competitive with greenfields sites.

    Representatives from the West Virginia Environmental Council (WVEC) expressed their willingness to seek a consensus on resolving DEP's concerns, especially regarding some of the illogical cleanup situations associated with past policies. However, they raised a number of concerns regarding both the risk assessment process and any perceived weakening of cleanup standards for land remediation and reuse proposals. These included:
 

• A general lack of trust in risk assessment practitioners'objectivity and scientific capacity.
   
• A fundamental, ethical disagreement with risk assessment assumptions that voluntary and involuntary risks are equivalent, that risks and benefits are distributed equally in a community, and that some risk is inevitable, thereby precluding a preventative, zero-emissions approach.
   
• A concern that public input into decision-making would be lessened, preventing the public from assessing the validity of scientific conclusions or incorporating societal values into government decision-making.
   
• A concern that non-quantifiable, qualitative values would be ignored or not given proper attention, even when required by other state and federal environmental statutes.
   
• A concern that weakened cleanup standards could create industrial "sacrifice zones" where developable land resources are devoted to the same out-dated polluting industries that created the contamination in the first place, leaving more modern, clean industries to compete for a shrinking pool of developable sites.
   
• A fear that any weakening of cleanup standards will merely put off the cleanup to future generations and represents a societal subsidy to industries that might not otherwise be competitive if they were required to pay their operation's full costs.
   
• A concern that local land use planning will be preempted.
   
• A concern that providing liability waivers would reduce the incentive for self-monitoring and pollution prevention by industry, thereby exacerbating the need for future governmental enforcement efforts.

    Little further progress was made at the agency level in 1994, but legislation drafted on behalf of the WVMA was introduced during the 1995 legislative session. Based on legislation from Ohio, the bill would have directed DEP to establish numerical cleanup standards and provide for riskbased variances to those standards. It would have also allowed a certified professional to issue a "No Further Action Letter" to property owners who met applicable standards, and mandated that DEP issue a "Covenant Not To Sue" to the property owners once a "No Further Action Letter" was submitted to the agency. The "Covenant Not to Sue" would remain in effect as long as the property complied with applicable standards and would also apply to all of the property's future owners or operators. The WVMA argued that this kind of self-implementing program would encourage redevelopment and voluntary remediation of brownfield sites.

    The WVEC objected to several of the bill's provisions, especially those that allowed minimal cleanups of sites with existing operators. They worried that some operators would abandon their sites instead of cleaning them up for re-use. They also objected to provisions that limited public access to information, reduced cleanup standards, required the issuance of a Covenant Not To Sue whenever DEP received a No Further Action Letter, and limited site monitoring.

Development of a Consensus Brownfields Bill

    The 1995 bill was not approved by the legislature, but it did attract a great deal of interest from legislative leaders who (1) established an interim legislative committee to develop a brownfields bill and (2) requested that environmental and industry representatives meet to reach a consensus on a bill.

    Early discussions among the interest groups yielded minimal progress, but the sides eventually agreed to abandon models based on bills from other states and draft a new approach crafted to address West Virginia's needs. It was also agreed to set aside some of the most contentious technical issues for later rule-making and focus the bill on general legislative policy issues. Then-Governor Gaston Caperton added significant impetus to the negotiations when he added Brownfields legislation to his agenda during his 1996 State of the State address.

    A general understanding was reached early that the bill would encourage voluntary cleanups by providing flexible standards and liability protections to achieve a balance that encourages economic development while assuring environmental protection. Nevertheless, consensus on specific wording to achieve that balance did not come easily. Eventually, meetings between interest groups, agency personnel, legislative staff, and legislative leaders resulted in agreement on language that offers substantial incentives for voluntary cleanups, but with provisions that narrowly target the eligibility for those incentives to assure that they are not misused. Unlike the controversy in the 1995 bill, the 1996 brownfields bill (HB 4605) passed both houses unanimously, and was signed into law by then-Governor Caperton on March 26, 1996.

HB 4605

    While many elements of HB 4605 are adapted from other states, it is one of the first brownfields bill in the United Statesdrafted as a negotiated consensus between environmentalists and industry representatives. As such, it is a model against which similar legislation in other states may be evaluated.

    No single piece of legislation can solve every problem, and, as a compromise, no one got everything they wanted with HB 4605; thus, there will inevitably be people who are dissatisfied with the law. Nevertheless, an overwhelming consensus on the major policy issues in the law developed in both the legislature and in the governor's office.

Policies and Definitions

    HB 4605 focuses on voluntary remediation activities. However, because hazardous substances and the public's health are involved, DEP will propose specific rules to guide cleanup efforts. But the process is inherently different from most other environmental regulatory actions. DEP will not initiate or mandate the cleanup. Instead, the cleanup will be initiated by the land owners (usually a company or public development agency). The act, however, does not preempt any of DEP's statutory duties to protect the public's health.

    The law applies the term "Brownfields" to properties not being actively used by the owner on the article's effective date (July 1,1996). This definition specifically excludes properties subject to unilateral enforcement orders or listed under the federal SuperFund law. This assures that only voluntary actions are covered because sites under an enforcement order are, by definition, not being voluntarily remediated. It also encourages existing businesses to continue their pollution prevention efforts because they are not eligible for brownfields incentives if their property becomes contaminated. Of course, existing businesses can still conduct "voluntary actions" to remediate contamination that occurs afterthe law's effective date. However, voluntary action sites are not eligible for public financial assistance available to brownfields under the law.

    The act also defines the "engineering" and "institutional" controls that may be used to meet cleanup standards. Engineering controls include both the treatment of hazardous substances and measures that merely contain them (i.e. caps, liners, leachate collection systems, etc. ). Institutional controls involve legal or contractual restrictions on the property's use. They are intended to limit the public's exposure to hazardous substances. These include deed restrictions or "land-use covenants" which limit property uses to activities appropriate to the remediation standards that have been met. Although specific use standards are set during rule-making, brownfields laws typically allow certain industrial uses on properties that would not meet the environmental standards for residential or agricultural uses. These lower standards must still protect industrial site users, but are typically far less expensive to attain than cleaning up to a residential use standard. In exchange for the weakened cleanup standards, the property owner accepts land use covenants that prevent the property from being used for purposes inappropriate to the level of contamination remaining on the site. The weakening of cleanup standards is considered justified because the alternative, in many cases, is for the contaminated property to remain abandoned, thereby creating a perpetual environmental threat. The in" dustrial use standards and accompanying land use covenants are meant to allow the property to be economically productive while eliminating its environmental or health risks to the community.

    The law requires a "Licensed Remediation Specialist," certified under DEP rules, to supervise the remediation of contaminated sites. This is intended to assure technically competent site assessments, remediation, and monitoring of contaminated sites so that cleanups will, in fact, be protective and worthy of the liability waivers for which they are eligible.

    The law also defined a "Risk" as the probability that a contaminant, when released into the environment, will cause an adverse effect in exposed humans or other living organisms. Establishment of risk-based corrective action standards and procedures for conducting risk assessments will be discussed later. But the principle of allowing risks from hazardous substances contradicts nondegradation policies established under other environmental statutes, and its application will likely be one of the most controversial aspects of the brownfields law.

Brownfields and Voluntary Action Processes

    West Virginia's brownfields program is integrated into the state's voluntary remediation program. In essence, brownfields remediation is a special case of the voluntary action program. In most respects, brownfields sites follow the same application, voluntary agreement, and remediation processes as other voluntary actions. They are, however, eligible for several incentives not available to existing businesses conducting a voluntary remediation, including site assessment loans to help pay for the costs of determining the extent of any contamination, Tax Increment Financing, and remediation loans from the West Virginia Economic Development Authority.

    The first step in a brownfields remediation is completing an application to DEP. The application includes a site assessment made by a licensed remediation specialist. Any person, corporation, local development authority, or other entity may apply, and. if accepted, can enter into a remediation agreement.

    The remediation agreement requires that a licensed remediation specialist supervise all work, and specifies the remediation's tasks, deliverables, and schedules. DEP must assure that these specialists are technically competent to design and oversee remediation work. A strict code of ethics requires the specialist to protect the public's safety, health, and welfare. Moreover, the specialist is legally responsible for any release of contaminants that occurs during remediation activities, and may lose his or her license, or be subject to civil or criminal penalties for violations. A final report with all information needed to confirm that the required work was completed, and a listing of the technical standards to be applied relative to the property's proposed future use, is also required. The agreement also must provide for full recovery from the applicant of any costs DEP may incur in overseeing the remediation plan. Finally, the agreement specifies that DEP will not initiate any enforcement actionagainst a person or business who is in compliance with the agreement, unless there is an imminent threat to public health.

    DEP retains the authority to make inspections, to collect appropriate samples, and to have access to any records relating to the remediation. Reports filed with DEP may not disclose confidential business information, but are otherwise available to the public as required under the Freedom of Information Act.Once a voluntary remediation action has been completed and the site meets the applicable standards, DEP can issue a certificate of completion to the land owner. It remains effective as long as the property meets the standards and relieves the property from any further remediation requirements, unless certain reopener provisions apply. These reopeners allow DEP to require further remediation if fraud was committed, previously unknown contamination is found, the remediation method fails, the level of risk associated with the site is increased significantly beyond the established level of protection due to new information about the contaminant or new land uses which alter exposure assumptions, or if the release occurred on a new industrial site where the remediation relied on institutional or engineering controls and the removal or destruction of the contaminant becomes economically practicable. This latter reopener is intended to apply to voluntary remediations on new industrial sites as an added incentive to prevent continued environmental degradation.

    The certificate of completion must include a land use covenant if institutional or engineering controls are applied and shall include whether residential or nonresidential standards were used. A property owner who knowingly violates a land use covenant by converting from nonresidential to residential use is guilty of a felony.

    Waivers of liability are limited to liability for further environmental remediation. A party responsible for injury or property loss due to the release of a contaminant is still required to provide compensation for any third party losses, but is not required to conduct additional remediation unless the aforementioned reopeners apply. Responsible parties are not relieved of their responsibility for contaminants cleaned up by a new, innocent owner under a voluntary action. In essence, the law retains the principle that the person or business causing the release is responsible for all damages associated with it, but that a person or business who voluntarily cleans up a release can limit their liability for cleanup costs by coming forward voluntarily, rather than waiting for an enforcement order. Liability protection is also available for contractors, financiers, utility workers, etc., so that they are not held responsible for losses simply by virtue of being engaged to clean up another party's release.

    These liability limitations are intended to be quite narrow. Releases or damages caused by gross negligence or willful misconduct are not covered by the liability protection, so due diligence on the part of all involved remains the rule.

    Many of the technical details, such as fees for recovering agency costs, risk assessment procedures, and numerical cleanup standards, will be worked out during rule-making. Many of these will obviously involve critical public policy issues. The consensus-building process used in negotiating the bill is being continued and expanded in the rule-making process. It is expected that the required rules will take effect, and the program will become operational later this year.

Risk Assessment Procedures and Risk-Based
Corrective Action-How Clean is Clean?

    The new law authorizes voluntary remediations of sites to levels commensurate with the actual environmental risk at the site. These risk-based standards and risk assessment procedures represent an evolving, but controversial, approach to setting regulatory standards.

     Historically, many environmental regulatory practices had their genesis in the concern over cancer. While "No Observable Effect Levels" had been determined for other toxic health effects, the U. S. Food and Drug Administration (FDA) concluded in 1950 that a No Effect threshold could not be proven for cancer and that any exposure to carcinogens, no matter how small, increased the risk of contracting the disease. In response, Congress adopted the Delaney Amendment to the Food, Drug and Cosmetic Act in 1958. It prohibited the addition of carcinogenic compounds into the food supply. Pesticides, preservatives, other additives, or their derivatives which were demonstrated to induce cancer are not allowed in any detectable amounts. Although the Delaney Amendment was appropriate for 1 950s technology, modern chemical analytical methods have improved so much that extremely minute traces can now be detected at concentrations that arguably pose negligible cancer risks.

    Because specific types of cancer are relatively unusual occurrences with numerous natural causes, proof of carcinogenicity involves detecting an increase in cancer rates above normal background rates. These rates are usually expressed as the probability of developing a particular type of cancer, and this probabilistic approach was based on dose-response research where animals were exposed to various, usually high, doses of the carcinogen. Cancer rates at very low doses were then extrapolated from these high dose data.

    Risk assessment, at its simplest level, calculates an expected dose (or exposure) for a situation and multiplies this dose by the cancer rate per unit dose to determine the probability of an adverse outcome. Advances on this model were applied to other types of toxic agents and to other health effects. Importantly, unlike the Delaney Amendment's nothreshold approach, regulation of other environmental risks balances risks against thresholds determined by potential benefits.

    In 1970, the FDA defined a regulatory standard for quantitative risk assessment for nonfood exposures, setting a maximum acceptable lifetime risk as one excess cancer in 100 million people exposed. Later, both the FDA and the U. S. Environmental Protection Agency (EPA) relaxed these guidelines to between one excess cancer in one million and one in 100,000.

    Generic risk assessments usually contain a variety of conservative, "worst-case scenario" assumptions to assure that the maximally exposed individual will be protected. Site-specific risk assessments, however' like those used to determinerisks at brownfields sites, use site-specific factors instead. These factors typically demonstrate a risk level that requires less stringent cleanup levels than would be required under a generic risk assessment. This, in turn, creates a debate over which factors should be used to ensure the public's safety, the more stringent assumptions under the generic approach or the less stringent assumptions under the sitespecific approach.

    Criticisms of risk-based standards fall into the following three broad categories:
 

• Technical disagreements over scientific issues.
   
• A concern that risk management issues and politics often interfere with risk assessment processes.
   
• Differences over values, where ethical dimensions of risk assessment are challenged (i.e. what is the relative importance of voluntary versus involuntary risks, what are appropriate "acceptable risk" levels, and does it matter if those who profit from risk-based decisions are not the same people who will be exposed to those risks).

    In an effort to standardize the risk assessment process, the National Research Council (1983) developed a model process which has been adopted by several federal agencies and incorporated into their guidance documents on risk assessment. The NRC model involves four steps:
 

•  Hazard identification.
•  Dose-response assessment.
•  Exposure assessment.
•  Risk characterization.

    Hazard identification evaluates an agent's potential to cause harmful effects. All possible hazards associated with the site are evaluated. Dose-response estimates are obtained for each hazard and their potential effect on human health is identified. The target population's estimated exposure to each hazardous agent from the source is determined. This information is integrated in the risk characterization step to determine the nature and magnitude of the risks and the uncertainties associated with various outcomes.

    Since completely accurate assessment of risks is impossible, the starting assumptions and other policy choices can have a strong influence on risk predictions. The NRC distinguishes risk management policy from risk assessment policy and specifically warned that risk management policies (policies used to determine the selection of regulatory options), should not be allowed to control risk assessment policies. In setting this principle, the NRC recognized that risk management issues, such as the regulatory control option's cost or the acceptable level of risk, involve value/political judgments that should not interfere with the scientific process of risk assessment.

    The brownfields law anticipates that site-specific risk assessments will be used to define cleanup standards. The goal of risk-based cleanups is to minimize the risk from an environmental contaminant, rather than to focus solely on reducing the amount of the contaminant. Remediation that isolates a contaminant may reduce exposure, and therefore risk, at lower cost than permanent removal. But permanent isolation of contaminants in media such as soil or groundwater is difficult, so risk reduction techniques need to be carefully designed and monitored to prevent remedy failure and to assure continued prevention of unacceptable exposures.

Pros and Cons of Risk Assessment

    Risk assessment is attractive because it provides a mechanism for integrating data on numerous environmental processes in a comprehensive, consistent, quantitative, objective fashion. Regulatory agencies and the courts have consistently preferred objective, quantifiable criteria over subjective criteria as a basis for informed decision-making. In a practical sense, risk assessment also provides a scientific basis to prevent resources from being wasted on arbitrary cleanup efforts that provide little or no enhancement of environmental safety.

    The promise of risk assessment is a comprehensive, systematic process to incorporate objective information into decision-making, and to do so within the broader context of environmental protection and public health. It offers a mechanism to consider the interdependence of multimedia problems using the best available scientific information and to identify the gaps in knowledge that create uncertainty in environmental decision-making. While this goal is laudable, inherent structural weaknesses in implementation often prevent this goal from being realized.

    Risk assessments have been criticized by scientists, industry, regulators, environmentalists, and ethicists (Ginsburg 1993; Montague 1992; Silbergeld 1994).

    Scientists recognize, as a fundamental principle of science, that trends in data only apply within the limits of the experimental conditions. But risk assessment often takes data generated under fairly narrow experimental conditions and extrapolates them to whole ecosystems and across wide ranges of environmental conditions. Proponents and opponents differ as to whether this represents estimation based on the best available predictive knowledge, or oversimplified guessing and wishful thinking. Some in industry object to risk assessment partly because of its use to ban certain products, but also because of its instability, where two equally valid experiments can produce data generating different estimates of risk with no objective way of choosing between them. Other criticisms of risk assessments that have been voiced include that they:
 

• often neglect preventable risks, merely because they are small. This is fundamentally incompatible with toxics reduction and pollution prevention.
   
• often overemphasize cancer endpoints, primarily because few data on other endpoints exist for most agents.
   
• sometimes oversimplify all possible environmental threats to a few risk drivers, thereby ignoring multiple exposure pathways, multiple endpoints, or multiple chemical interactions; producing an underestimate of cumulative risks.
   
• often implicitly assume that a risk agent is innocent until proven guilty, causing it to underestimate the risks associated with newly discovered risk agents- are sometimes based on an inherent presumption of a right to pollute instead of a right not to be polluted.are often based on models of emissions averages and health effects estimates of average populations, and often ignore the impact of unforeseen accidents or atypical susceptible populations.
   
• often attempt to predict incremental increases in risk without considering cumulative exposures that may exceed biological or ecological thresholds.
   
• can complicate the regulatory process, leading to paralysis by analysis.
   
• are fundamentally undemocratic, substituting pseudoscientific values and calculations for democratic participation in decision-making.

    When risk assessments are manipulated, or used to justify that a particular policy choice is "safe enough," the aforementioned criticisms are legitimate. When used as part of a comprehensive assessment of alternatives, with peerreview and democratic involvement procedures, risk assessment can be a useful tool to assist in making informed decisions. Some policy guidance and protocols for conducting risk assessments were included in HB 4605 to minimize these potential problems. Rule-making will need to identify procedures that address these criticisms so that acceptable risk assessments will, in fact, provide an accurate and complete analysis of risks from a site.

Ecological Risk Assessment

    The EPA has expanded the risk assessment process from its emphasis on human health effects to include broader ecological risks (EPA 1992). In their framework, ecological risk assessment is composed of three phases. The first phase is problem formulation, where systematic planning determines the assessment's feasibility, scope, scale, and endpoints. It includes a preliminary characterization of exposure and effects, data needs, policy and regulatory issues, and site-specific factors. Since ecological impacts may occur to diverse species, habitats, trophic levels, or ecosystem-wide functions, choosing the appropriate endpoints to assess and the appropriate questions to answer is no easy task.

    The second phase of ecological risk assessment analyzes the risk agent's exposure and ecological effects. The thrid phase, risk characterization, uses the results of the exposure and ecological effects' analyses to evaluate the likelihood of adverse ecological effects, as well as to summarize the assumptions used, the uncertainties and magnitudes of the effects, their spatial and temporal patterns, and the likelihood of recovery. The ecological risk assessment summarizes the ecological significance of the consequences for the risk manager.

    Under the EPA framework, ecological risk assessment is an iterative, nonlinear process where information needed to reach conclusions in early phases does not become available until later phases. EPA recognized that interactions between the risk assessor and the risk manager were needed to assure that the assessment information was timely and relevant to decision-making. They also recognized the need for verification and monitoring to validate specific predictions made by the risk assessment and to provide necessary feedback for future improvements in decisionmaking.

    The issue of whether ecological risk assessment is needed has been reviewed extensively. Although many have argued that human health-based standards contain conservative assumptions and safety factors that lead to negligible risks, Hegner (1994) showed that human health-based Maximum Contaminant Levels (MCLs) were higher than ambient water standards for aquatic organisms and thus not protective of ecological health for half of the 18 chemicals compared. In fact, the ecologically protective standard was often lower than minimum analytical detection limits specified for RCRA or CERCLA monitoring programs, indicating that chemical monitoring could not predict whether or not ecological effects would occur at those sites.

    Suter (1993) summarized several reasons why human health risk assessment was insufficient for environmental protection:
 

• Some routes of nonhuman exposure, such as root uptake or drinking from waste water ponds, are not credible for humans.
   
• Some nonhuman species are more sensitive than humans, often because they experience physiological responses that humans do not have (e.g. egg-shell thinning from DOT).
   
• Some ecosystem level effects, such as lake eutrophication, have no human analog.
   
• Nonhuman species may be exposed more intensely than humans, even when the exposure routes are the same (i.e. humans rarely eat every fish meal from the same stretch of a contaminated river, but a heron or otter will, and they will eat the whole fish, not just fillets of relatively uncontaminated muscle).
   
• Most birds and animals have higher metabolic rates than humans, so they receive a larger dose per unit of body mass because of higher consumption rates.
   
• Some agents are specifically designed to kill "pests" while being nonlethal to humans.These agents often affect nontarget organisms that are ecologically or physiologically similar to the target.
   
• Nonhuman species are often tightly coupled to their habitats and cannot shift to alternate food sources or other habitat resources to substitute for secondary effects, even when the species is not directly affected by the risk agent.

    Since HB 4605 establishes risk-based numerical standards for corrective action and risk assessment procedures that are protective of both human health and ecological receptor health and viability, the risk assessment procedures will need to mesh both human and ecological risk assessment models. At the same time, regulations will need to retain the flexibility to incorporate new advances as these emerging fields struggle to define themselves (Renner 1996). Progress to date in rule-making negotiations offers substantial hope that West Virginia's program will achieve these goals.

Public Involvement, Risk Communication, and RiskManagement

    Issues pertaining to public involvement are among the most controversial in the brownfields arena. HB 4605 specifies that: (1 ) risk protocols should include public notification and involvement, (2) brownfields sites that use public financing should allow public input into basic land use decisions, and (3) other routine public hearing and comment provisions under other statutes apply where appropriate. However, it also specifies that confidential business information must remain confidential. Other than these requirements, DEP has relatively little statutory guidance to follow when it establishes its public involvement procedures.

    Public involvement processes are sometimes time-consuming, expensive, and often frustrating even for the best-intentioned agency or business. Many companies are legitimately concerned about excessive public scrutiny because they believe it can bring misunderstandings and ultimately tie up benign projects in frivolous lawsuits and unfounded public controversy. Indeed, some have argued that public input rarely contributes anything useful to decision-making and that as long as a site meets adequate safety and environmental standards, how to achieve those standards on private property is solely the private company's prerogative. At the same time, secrecy arouses suspicions, leading to rumor mongering and often unjustified fears and public opposition. Thus, it is almost always in the company's interest to involve the public early in the decision-making process. The trick is to design a process that allows for meaningful public input in a timely manner. Recent research suggests that this is most likely to occur when the public is informed early in the decision-making process, trusts the information being provided as being truthful and objective, and are provided some measure of control over the project (NRC 1989; Slovic 1993; Freudenburg and Rursch 1994; Johnson and Slovic 1994; Powers 1995).

    Risk communication is a public involvement mechanism that translates the detailed technical calculations of scientists into terms the lay public can understand. Placing risk estimates in context with risks from everyday activities and illustrating the benefits of an action builds public support for decisions made by the risk manager. Some risk communication efforts involve a one-way process in which the risk communicator attempts to convince the public to accept both the information and the subsequent decision of the risk manager (Bradbury 1994). An alternative model describes communication as a two-way iterative process of sharing information and decision-making, where the participants are mutual communicators, rather than a sender and receiver of messages. The National Research Council (1989) identified the latter approach as a more productive form of democratic dialogue. Thus, risk communication should involve third parties, community groups, and affected citizens early and often, with a high regard for their concerns, so that they provide input before critical decisions are made.

    No single method for public involvement works for every situation, or for every segment of the public. For example, public hearings, legal notices, and citizen advisory committees attract more highly informed and involved citizens while informational meetings or an ombudsperson field office may attract people more representative of the community at large (Westman 1985). Approaches that are useful for educating the general public may not be useful for resolving conflict with highly motivated interest groups. Public participation programs, therefore, need multiple input processes to encourage both the involvement and input from all relevant interests throughout the process. In this way, all affected parties have an opportunity to provide meaningful input into the decision-making process and are likely to view the final decision as legitimate. Providing citizens with ownership in the process is critical to building the community-wide cooperation needed for brownfields redevelopment.

Recommendations

• Incorporate a comprehensive scientific base of ecological information into risk assessments.

    A common failing of many risk-based decision-making has been to focus solely on human health risks. Although environmental protection is specified as a goal of the brownfields law, additional guidance is needed in rule-making to assure comprehensive environmental assessments. Ecologists should play a central role in the design and conduct of ecological risk assessments, but their central role is frequently not recognized by non-ecologists (Reagan, et al. 1994). Ecological baseline data are needed at every scale of ecological organization, from the cellular to the ecosystem level. If risk is a measure of the deviation from normal ecosystem function, structure, or behavior, then knowledge of the unperturbed system is a prerequisite to valid assessments. Avoid attempts to minimize assessment of ecological risks merely because sites are highly disturbed or industrialized.

• Assure meaningful public input into risk-based decision-making.

    Meaningful input in a democracy means that the public has a level of control over decisions resulting from the process. We should avoid the common tendency to shun publicity or restrict access to information, especially when cleanups involve less stringent standards or risks over which the surrounding community otherwise has little control.

    Public involvement methods should solicit input as well as inform the community, and should recognize that technical knowledge is sometimes less important than social factors in risk management. Over-reliance on quantitative data and the failure to appreciate the importance of differing local, social, and cultural factors that alter risk perceptions and subsequent behaviors has been at the root of many conflicts. One person's extremely minute and acceptable risk may be unacceptably large to another. This is not the lack of technical knowledge, it is a difference in basic values that must be respected in a democratic society.

• Assure technically competent brownfields cleanups.

    Although many brownfields sites do not need the extensive remediation of listed SuperFund sites, adequate cleanups may still be technically difficult or expensive. Rigorous standards for the certification of licensed remediation specialists are needed, as are follow-up and monitoring of remediation efforts, so that future owners and developers can be assured of the property's full value, and the surrounding community can be assured of the quality of their environment. Risk-based cleanups would benefit from scientific peer-review of risk assessments and remediation designs. The surrounding community should have independent technical advice so that they can adequately incorporate their concerns and assure balanced decision-making.Avoid overblown expectations of results from the brownfields law.

    Some have touted the brownfields legislation as a kind of magic answer to both environmental remediation and economic development needs. Unfortunately, some of the state's brownfields sites are severely contaminated, and will be relatively expensive to remediate even with the new regulations encouraging their redevelopment. Adequate monitoring of sites and frequent modernization of the voluntary remediation program will be needed to assure that public health and the environment are protected and that beneficial economic development is being encouraged for the whole community.

Conclusions

    To the extent that old or inadequate statutes hindered remediation, the brownfields law should minimize unnecessary roadblocks. Although the law identified some taxpayer-supported assistance for site assessments, it did not contain any major sources of new funding. Instead, it encourages private sources of capital to invest in remediation while continuing the principle that the responsible party remains liable for cleanup costs.

    Will the brownfields law stimulate economic redevelopment while assuring environmental remediation of brownfields sites? Or, will it merely encourage superficial cover ups while postponing needed cleanup to our children's generation? Or worse, will it allow existing contamination to spread and indirectly encourage new brownfields sites by allowing existing industrial sites to be abandoned under inadequate engineering controls? Hopefully, stringent rules, meaningful public involvement, adequate monitoring, and enforcement will prevent these adverse outcomes. The consensus rule-making process currently underway offers real hope for a favorable outcome, but only time will tell.

References

Bradbury, Judith A. 1994. "Risk Communication in Environmental Restoration Programs." Risk
    Analysis 14:3: 357-363.

Dilger, Robert Jay and Tom Stuart Witt. 1994. "The $300 Million infrastructure
    ImprovementAmendment. "The West Virginia Public Affairs Reporter 11:4: 2-10.

Freudenburg, William R. and Julie A. Rursch. 1994. "The Risks of 'Putting the Numbers in Context':
    A cautionary tale." Risk Analysis 14:6: 949-958.

Ginsburg, Robert. 1993. "Quantitative Risk Assessment and the Illusion of Safety." New
    Solutions 32: 8-15.

Hegner, Robert E. 1994. "Does Protecting for Human Health Protect Ecological Health?" Risk
    Analysis 14:1: 3-4.

Johnson, Brandon B. and Paul Slovic. 1994. "'Improving' Risk Communication and Risk
    Management: Legislated Solutions or Legislated Disasters?" Risk Analysis 14:6: 905-906.

McCoy, Eli. 1994. Deputy Director, West Virginia Division of Environmental Protection. Letter
    dated June 15.

McElhinny, Brad. 1995. "Costs Driving Out Small Gas Sta-lions." Charleston Daily Mail.
    September 18.

Montague, Peter. 1992. Questions to ask about risk assessment. Annapolis, MD:
    Environmental Research Council.

National Research Council (NRC). 1983. Risk Assessmentin the Federal Government: Managing
    the Process. National Academy of Sciences National Research Council Committee on the
    Institutional Means for assessment of Risks to Public Health. Washington, D. C.: National
    Academy Press.

National Research Council (NRC). 1989. Improving Risk Communication. National Academy of
    Sciences National Research Council Committee on Risk Perception and Communication.
    Washington, D. C.: National Academy Press.

Powers, Charles W. 1995. State Brownfields Policy and Practice: A Report of an IRM
    Conference for State Officials. Boston, MA: Institute for Responsible Management.

Reagan, Douglas P., Mike Firko, and Frieda B. Taub. 1994."The Role of Ecologists in Ecological
    Risk Assessment." Bulletin of the Ecological Society of America 75:2: 96-99.

Renner, Rebecca. 1996. "Ecological Risk Assessment Struggles to Define Itself." Environmental
    Science and Technology 30:4:172-174.

Ruben, Barbara. 1995. "Fields of Dreams." Environmental Action (Winter): 12-17.

Silbergeld, Ellen K. 1994. "The risks of Comparing Risks. "New York University Environmental
    Law Journal 3: 405-432.

Slovic, Paul. 1993. "Perceived Risk, Trust, and Democracy." Risk Analysis 13:6: 675-682.

Suter. Glenn W. 1993. Ecological Risk Assessment. Chelsea, MI: Lewis Publishers.

U. S. Environmental Protection Agency (EPA). 1992. "Framework for Ecological Risk
    Assessment." EPA Risk Assessment Forum. EPA/630/R-92/002. Washington, D. C.: U. S.
    Environmental Protection Agency.

Westman, Walter E. 1985. Ecology, Impact Assessment, and Environmental Planning. New York:
    Wiley-lnterscience.