4
Selection Criteria

The health and safety of the public and the protection of the environment were clearly factors in the citizen concerns expressed at the public forum, discussed in Chapter 1, in other citizen communications, and in numerous publications. These concerns, as well as the committee's review of the accident and routine operation health risks of stockpile storage, handling, and disposal, guided the conduct of this study.

In view of the overriding public and committee concern for health and safety, the committee selected as its primary criterion, in formulating its recommendations, the minimization of the cumulative adverse consequences from all relevant risks over the full duration of the disposal program . Consequences of concern extend to effects on the environment, as well as on the health and safety of workers and citizens, resulting from the release of agents and other pollutants during accidents or in routine operation. They include both immediate and delayed effects. For brevity in the following sections, and with recognition that the consequences of sustained risk accumulate in time, this is referred to as the "cumulative total risk minimization criterion."

RISK

Risk can be defined as the expected value of a consequence. For accidents and normal operations, it is quantified by such measures as fatalities per year or latent cancer fatalities per year. As time increases, the cumulative effect of a continuing risk is an increase in the potential consequences. For example, if the risk from driving is one fatality per year for every 10,000 drivers, and that risk remains constant, then the expected consequences would be two fatalities after two years, ten after ten years, and so on. The expected consequences will continue to rise as long as the risk persists.

For the stockpile disposal program, risks will persist and consequences will accumulate until the entire stockpile is destroyed and all of the facilities are decommissioned. Because chemical agent and munitions inventory, storage conditions, and surrounding population and community conditions vary



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions 4 Selection Criteria The health and safety of the public and the protection of the environment were clearly factors in the citizen concerns expressed at the public forum, discussed in Chapter 1, in other citizen communications, and in numerous publications. These concerns, as well as the committee's review of the accident and routine operation health risks of stockpile storage, handling, and disposal, guided the conduct of this study. In view of the overriding public and committee concern for health and safety, the committee selected as its primary criterion, in formulating its recommendations, the minimization of the cumulative adverse consequences from all relevant risks over the full duration of the disposal program . Consequences of concern extend to effects on the environment, as well as on the health and safety of workers and citizens, resulting from the release of agents and other pollutants during accidents or in routine operation. They include both immediate and delayed effects. For brevity in the following sections, and with recognition that the consequences of sustained risk accumulate in time, this is referred to as the "cumulative total risk minimization criterion." RISK Risk can be defined as the expected value of a consequence. For accidents and normal operations, it is quantified by such measures as fatalities per year or latent cancer fatalities per year. As time increases, the cumulative effect of a continuing risk is an increase in the potential consequences. For example, if the risk from driving is one fatality per year for every 10,000 drivers, and that risk remains constant, then the expected consequences would be two fatalities after two years, ten after ten years, and so on. The expected consequences will continue to rise as long as the risk persists. For the stockpile disposal program, risks will persist and consequences will accumulate until the entire stockpile is destroyed and all of the facilities are decommissioned. Because chemical agent and munitions inventory, storage conditions, and surrounding population and community conditions vary

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions considerably among the sites, the cumulative total risk minimization strategy should ultimately be determined on a site-specific basis. In keeping with an overall risk minimization criterion, and with the recognition that any selected technology must satisfy technical disposal requirements, the following criteria were used to choose among potential alternatives to the baseline system: Can the technology contribute to a program of disposal and associated storage that is safer than that of the baseline program? Can the technology treat agent, energetics, metal parts, and dunnage? Can the technology destroy all agents? Do waste products meet environmental disposal requirements? Can the technology achieve treaty requirements for irreversible agent destruction? The primary criterion—safety—involves a number of issues, including the basic chemistry of the agents, operating temperature and pressure, the composition and quantity of waste streams, the disposal system complexity, and timely implementation. These issues are considered in the evaluation of potential alternative technologies in Chapter 6. Combinations of technologies may be necessary to satisfy some criteria. No technology can single-handedly satisfy both treaty and hazardous waste disposal requirements. For example, incineration must be followed by an effective pollution abatement system. While it is the purpose of this report to recommend technologies on the basis of these criteria, technology decisions may ultimately be dictated by other than technical criteria. If, for example, those who oppose incineration demand use of an alternative technology, timely exploration and development of potential alternatives will be necessary even though "substantially safer" options may not be available. RISK AS A MEASURE OF HEALTH AND SAFETY (WORKERS, THE PUBLIC, THE ENVIRONMENT) With cumulative total risk minimization as the overriding criterion for the consideration of stockpile disposal technologies, it is important to clearly define what is meant by risk, particularly health risk. The U.S. chemical stockpile presents two distinctly different and important types of health risk to workers at the sites, persons living near the sites, or plant and animal life in the regions surrounding the sites:

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions the acute and latent health and environmental effects of direct exposure to agent and disposal products associated with accidents and incidents during the storage, handling, and processing of chemical munitions; and the health and environmental effects of long-term or chronic exposure to agent and disposal products during and after routine operations of storage, handling, and chemical agent and munitions processing facilities. These two types of risk are referred to as the risks from accidents and the risks from normal operations, as illustrated in Figure 4-1. To date, the development of risk assessments for the Chemical Stockpile Disposal Program has been motivated largely by consideration of the risks from accidents. With respect to the risk from accidents, considerable progress has been made in recent years in designing safety into engineered facilities (e.g., nuclear power plants). This progress is a direct result of considering safety as a fundamental performance indicator similar to throughput rates, process efficiency, product quality, and cost. In this case there is a need to measure safety in some quantitative form and to integrate the results into the basic design and performance assessment process, and then into process safety management. One means of measuring safety is through a risk assessment that calculates the frequency of events, or series of events, leading to consequences of different levels of severity. In particular, this type of risk assessment answers three basic questions (Kaplan and Garrick, 1981): What can go wrong? How likely is it? What are the consequences? The first question is answered by structuring various scenarios (sequences of events) of the different ways that the facility can get into trouble. This step requires a clear understanding of how the facility operates and its sources of risk in order to identify those events that can initiate an accident scenario. The key to developing a complete set of initiating events is to recognize that the various contributors to risk can be many, including equipment failures, external phenomena (earthquakes, severe storms, lightning strikes, aircraft impact, terrorist attacks, etc.), human errors, and institutional failures (Figure 4-2). The likelihood of such rare events is difficult to estimate accurately. The degree of uncertainty in the estimate needs to be considered when risk decisions are made. The question of likelihood and magnitude of an incident often requires a detailed analysis since, in many instances, there are too few data to assign statistically based frequencies. There are almost always gaps in quantitative data as well. Expert opinion must in some cases be substituted for the missing

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions Figure 4-1 Types of Risk. data, thereby adding one more source of uncertainty to the analysis. The most common approach is to develop frequencies and probabilities based on all available evidence; to integrate them into detailed logic models that decompose the system into components for which there are data and increased knowledge; and to do so in such a way as to recognize the uncertainties involved. The Army's 1988 programmatic risk assessment of the Chemical Stockpile Disposal Program (discussed later in this chapter under risk assessment activities) notes that uncertainty in risk assessment arises from many causes, including the inadequacy of data, inaccuracies of modeling, and the incomplete identification and understanding of accident phenomena. Error factors used to characterize the uncertainty inherent in each accident probability "point estimate" were obtained from detailed analyses of accident scenarios (U.S. Army, 1988). The Programmatic Environmental Impact Statement (PEIS) took into account a number of uncertainties and assessed the variabilities of the estimated means used to present the results of the risk analysis. Finally, the consequences axe defined in terms of injuries, fatalities, facility damage, environmental damage, etc., or combinations thereof in accordance with the purpose of the assessment. Thus, a risk assessment may be viewed as a structured set of scenarios, and their likelihoods and consequences. The results not only provide the likelihood of different levels of severity but also convey the analysts' confidence or uncertainty in the results. Equally important, the risk assessment for a specific facility (i.e., a site-specific risk assessment) exposes and ranks the importance of each contributor to risk. The results provide the ability to manage the risk at the most basic and effective level

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions Figure 4-2 Sources of risk. Source: PLG, Inc.

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions Risk assessment for releases of agent and pollutants that occur during normal operations and during transients or upsets must also be evaluated. In contrast to the risk assessment of accidents, these risk assessments address two basic questions: What is the magnitude of the release? What are the consequences? An operational release is the release of a small concentration of agent below the limits of detection technology. It also refers to agent destruction by-products (i.e., stack emissions or other process discharges). It does not imply that agent or emissions are actually released in a deliberate sense. For example, Table 2-4 indicates that some releases of agent from storage do occur on a continuing basis. Most are detected during routine inspections inside igloos, and few if any are detectable at the boundaries of the storage area. The magnitude of exposure is defined by an assessment that involves source definition (magnitude and chemical composition); environmental transport and transformation; and human, animal, and plant uptake. The consequences of this exposure, as in the case of accidents, are defined in terms of health effects, environmental damage, etc. Generally, for operational releases, the health risk of major concern is the incremental cancer risk to the workers and surrounding population as a result of releases of agent and other pollutants. The standards for such releases are established by the U.S. Environmental Protection Agency (EPA), the U.S. Army, and regional agencies. Compliance with these standards requires specific actions such as monitored trial burns prior to operation and continuous monitoring during full-scale operations. Risk assessments of accidents, transients or upsets, and normal operations can be developed at different levels of detail depending on the available knowledge and the intended uses of the results. Screening risk assessments can be used to identify major safety, health, and environmental concerns, even in the absence of detailed technical specifications. In addition, detailed, site-specific risk assessments can be used to evaluate and manage the risks associated with either accidents or normal operations. Choice of the cumulative total risk minimization as the basis for technology selection has far-reaching effects on both the characteristics of acceptable technology and the timing of the disposal program. This influence on schedule is independent of treaty obligations or political agenda, which the committee considers of lesser importance. Rather, the significance of timing with regard to cumulative total risk arises from evidence that acute risk is dominated by risks from storage rather than from disposal operations. It suggests that prompt disposal will minimize total risk, as discussed in detail

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions later under risk assessment activities. In fact, even if the storage risk is not currently large compared to the disposal risk, the cumulative consequence of storage risk will eventually outweigh that of any disposal operation. THE RISK OF CHEMICAL STOCKPILE STORAGE Communities near continental U.S. sites are already exposed to the risks associated with the storage of the unitary chemical agent and munitions stockpile as discussed in Chapter 2. Although these risks currently appear to be low and the condition of the aging stockpile is monitored continually, the agents and munitions will be a source of concern until they are destroyed. Of particular concern are the M55 rockets with propellant stabilizers that are deteriorating. Also, some aging mustard agent is deteriorating, producing a gel. In some cases, it produces a gas that increases pressure inside the storage containers and munitions. GB has also been found to produce a gel. Corrosion is another concern that both degrades the condition of the munitions in storage and increases the difficulty and associated risk of their ultimate disposal. Finally, there are accident risks that are still present due to external factors such as earthquakes, lightning strikes, and aircraft crashes (MITRE, 1993d; U.S. Army, 1988). As discussed in Chapter 2, the Stockpile Committee is not as confident as MITRE of the stockpile's stability over realistic estimates of the duration of the disposal program. In fact, the committee is concerned that continuing deterioration will increase the risks of disposal operations. THE RISK OF CHEMICAL STOCKPILE DISPOSAL OPERATIONS Planned or contemplated disposal operations pose their own independent risk. For the baseline system, there is evidence that the main source of risk to the public from accidental agent release is posed during transport of munitions and containers from the storage area to the disposal facility. This transport risk would be present for any other disposal technology. Once inside the disposal building, the risk of agent release to the environment is greatly reduced because of extensive physical containment safeguards inside the disposal facility. In addition to the risks of accidental agent release during storage and handling, there is the risk from releases during normal disposal operations. All candidate disposal options must ultimately eliminate the same stockpile materials, and in the long term, all will eventually be released to the environment, though the exact form, timing, and location of the discharge may

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions vary with the selection of agent disposal technology. Both the risk of storage and the risks of disposal operations are discussed later in terms of their relative contributions. HEALTH EFFECTS The Health Effects Of Accidental Releases Significant exposure to GB or VX results in neurotoxic response, and exposure to H (mustard) causes epidermal blistering. It is possible that short-term exposure to high levels of mustard agent may result in cancer induction (discussed later). Since there have been no significant agent releases during operations of the Johnston Atoll Chemical Agent Disposal System (JACADS), Rocky Mountain Arsenal, or the Chemical Agent Munitions Disposal System (CAMDS) at Tooele Army Depot over the past 30 years, it has not been possible to conduct traditional dose-response exposure studies on individuals in or around the destruction facilities. In order to obtain reliable estimates of health effects from any specific event, it is first necessary to demonstrate actual physical release to the environment and to detect agent or destruction products in or on biological species. The Army has established sensitive no-observed-effect-level (NOEL) values for each of the chemical agents. There are extensive air monitoring programs at Tooele and JACADS, as there were at Rocky Mountain Arsenal. However, there are currently no dose-response monitoring programs at JACADS or Tooele, since accurate health effects studies cannot be performed unless it is possible to measure exposure dose. The Army performs routine occupational health monitoring, including general health assays at all chemical agent and munitions storage sites. It also assays the neurotoxic-specific AChE (acetylcholinesterase) inhibition at facilities where nerve agent exposure might occur. In addition to this current monitoring, it may be possible to monitor other parameters, such as any decrease in maximal dilation of the pupils of the eyes or any changes in electroencephalograph (EEG) patterns. However, these are not being performed currently at any of the sites. Studies at the Rocky Mountain Arsenal in the 1970s demonstrated identifiable changes in EEG patterns after symptomatic minor worker exposures to GB, but the effects were very slight and required computer evaluation to be reproducibly identified (S. Leffingwell, personal communication, 1993). No long-term effects of any of these acute, symptomatic exposures have been documented. Similar EEG recordings were made of workers at Newport in the late 1970s; however, those tests were part of a medical monitoring program, not a research effort, and no analysis was undertaken (S. Leffingwell, personal communication, 1993).

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions There is little information about the neurotoxic effects, due to the rare instances of GB and VX exposure during production, storage, or warfare usage. It has been possible to chronicle the effects of mustard agents from historical exposures. The United States and England have ''mustard pensioners'' who were exposed to HD during World War I, and who subsequently suffered from chronic bronchitis and increased rates of cancer when compared to World War I amputee casualties not exposed to the agent. During World War II, the Japanese had a chemically contaminated facility on Okuna-Jima Island where workers experienced severe blistering exposure, primarily to H mustard, but the badly contaminated sites also contained phosgene and Lewisite. Apparently, increased cancer rates have been documented from exposure during work-related activities at mustard production facilities. In addition to numerous reports on cancer incidence in the Okuna-Jima cohort, there have also been reports of increased cancer incidence among British workers from a World War II-era plant that manufactured mustard (S. Leffingwell, personal communication, 1993). It has been difficult to provide a quantitative evaluation of the carcinogenicity of mustards since the mutagenicity is relatively low and the animal studies were limited; however, an Oak Ridge Study in 1980 suggested that H was three times as carcinogenic as benzo[a]pyrene. In 1993, the Institute of Medicine published a report, Veterans at Risk: The Health Effects of Mustard Gas and Lewisite, that considerably expands the list of health consequences of mustard (IOM, 1993). The Army currently enforces stringent, short-term, total concentration limits for working-level exposure and long-term general population limits that are far more protective in terms of public safety than are standards for other hazardous chemicals. This provides a comfortable safety margin to normal occupational exposure under current standards. The Health Effects From Normal Operations Most of the risk assessments that have been performed relative to chemical stockpile disposal have dealt with the accidental release of agent. However, there is increasing public concern over potential long-term, chronic health effects due to cumulative, low-level exposure to either agent or other pollutants (e.g., NOx and dioxin) from the normal operations of chemical storage or disposal facilities. These long-term public health effects are much more difficult to evaluate than the effects of accidental releases, and they suffer from incomplete, variably interpretable data for effects that may have a latency of 20 to 30 years. Such studies usually lack pre-exposure controls and rely on nonrandomly collected data such as self-reported effects. In addition, site-similar epidemiology at chemical agent and munitions disposal

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions sites or related hazardous waste disposal sites is currently quite limited in the scientific literature. These issues cannot be thoroughly addressed in this report; however, a separate National Research Council committee will address the nature and probability of health effects associated with incineration. This proposed evaluation has been titled "Health Effects of Waste Incineration" and will examine the potential health effects from incinerators in general (rather than from incinerators that handle specific substances such as the agents in the chemical stockpile), under both steady and upset operating conditions. It will be sponsored by the Agency for Toxic Substances and Disease Registry (ATSDR), the U.S. Environmental Protection Agency, and the U.S. Department of Energy. In the limited studies that have evaluated the chemical nature of stack emissions, it has been observed that the levels of destruction and the nature of organic compound emissions vary significantly, depending on the parent compound. Thus, the principal organic hazardous constituents (POHCs) and the products of incomplete combustion (PICs) can be quite complex. For example, the combustion of natural gas may include more than 100 elementary chemical reactions, which may result in over 100 different products of incomplete combustion. Although it is possible to determine whether any individual product is released at significant toxicity levels, the overall biological effects of the mixtures are unknown. Only a few incinerator emission studies have been performed in detail, primarily because of the complexity of the chemical analysis and the lack of correlation of presumed effects with dosage. In such studies, it is critical to be able to separate presumed insult (neuropathy, allergy, cancer induction, etc.) from coincidental circumstances. These studies have generally concluded that the normal "nonupset" operation of incinerators results in stack emissions that are equivalent to those from comparable combustion systems such as fossil fuel power plants. Off-normal operation is not monitored by the EPA, but disposal facility permits will require cessation of agent feed if furnace conditions are outside normal operating limits. All such facilities must satisfy EPA emission standards. As a point of reference, the EPA Health Effects Research Laboratory in North Carolina, and others, have provided data on the mutagenicity of stack emissions suggesting that incineration facilities, if operating properly, produce no more toxic pollutants than many residential furnaces, cars, and wood stoves (Watts et al., 1989; 1992; Driver et al., 1990). Recent studies of industrial incinerators have provided some evidence of the low risk of proper routine operation (Dempsey and Oppelt, 1993). An incinerator siting study oriented toward evaluating long-term, low-level toxicity effects is currently under way in North Carolina (ATSDR Division of Health Sciences). This study relies on self-reported symptoms from complex exposures that included direct worker exposure and fugitive emissions, not just stack emissions (S. Leffingwell, personal communication,

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions 1993). Another community health evaluation has been performed with the Vertac Incinerator in Jacksonville, Arkansas, in which the result of dioxin contamination was examined before and after remediation. Only immediate acute effects were detected, and those were correlated with immediate symptomatic exposure. The study suffers from a lack of precontamination or fugitive exposure information. Overall, it is difficult to extract consistent conclusions from the generic occupational studies that have been performed over the past several years. The difficulty involved in performing an epidemiological study before and after agent destruction could be very great. Such a study would be extremely difficult, if not impossible, because no correlation with dosage could be determined—since no measurable agent release would be expected and by-product releases could be very small. In addition, the maximal level of effect from chronic low-level exposures would be expected to be less than 5 percent differences, based on the correlations for AChE changes determined following symptomatic exposures described previously. Thus, the studies would be expected to produce data at the margin of detection accuracy, and any observed results might be strictly coincidental. Should a sublethal release occur, it would be possible to mark and develop a prospective cohort study. It might be possible to evaluate the impact of incineration technologies at various hazardous waste sites around the world in a comprehensive health effects study, such as the National Research Council evaluation previously discussed. Such a study is beyond the intent of the stockpile disposal program. Such general population evaluations might be reinforced by insights gained from laboratory animal studies; however, these require extrapolation from short-term, high-dose treatments to long-term, low-dose exposures more likely to be encountered in the environment. In addition, the laboratory animal response must then be converted to potential human responses in which the target organs might have differing sensitivities or vary significantly for particularly sensitive subgroups. Such extrapolations have been shown to be extremely difficult to make and to be inconsistent in results. It is not possible to assess the long-term impact of alternative technologies at this time because their engineering development is preliminary, and the extent of exposure to chemical agents or other destruction pollutants during processing is largely unknown. Resolution of the issues of the long-term public health and environmental effects of normal, low-level exposure to either agent or other processing pollutants is beyond the scope of this report. For the purposes of this report, the number of useful studies is quite limited and the data base is incomplete. There is certainly no clear indication of increased incidence of cancer, neurological disruptions, or other negative health effects that can be

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions Figure 4-3 Risk of stockpile storage and disposal. Source: U.S. Army, Office of the Program Manager for Chemical Demilitarization.

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions stockpile deterioration, as shown in Figure 4-4. In this case, the added cumulative risk is the rectangular shaded area under the risk curve between "Begin 1" and "Begin 2." As can be seen, this added area is much larger than any possible reduction of area under the disposal curve. Neglecting the adverse effects of deterioration is reasonable in the near term, but given the uncertainties in stockpile deterioration, this assumption would not be acceptable for extended program delays. In any case, at least for relative storage and disposal risks as indicated here, any decrease in disposal risk afforded by a new technology (lower disposal risk curve) would be quickly offset if a delay in starting date permits greater accumulation of storage consequences. This programmatic risk assessment forced consideration of the risk of accidents for specific activities and operations in the context of the overall disposal program. Thus, insights were developed on the relative risks of such activities as storage, handling operations, transportation, munitions disassembly, agent destruction, pollution abatement, and final waste disposal. Figure 4-3 brings into sharp focus the importance of considering both storage and disposal risks in minimizing cumulative total risk. If storage risks are dominant, then prompt disposal is the remedy. If processing risks are dominant, improved alternative disposal technologies are called for. It is important to note that Figure 4-3 conveys only the acute risks of storage and disposal operations. It does not consider latent health and environmental effects from both accidental and low-level releases of agent or other pollutants from either storage or disposal operations. Latent and acute risks are separate issues that are not easily compared. Low-level releases from storage and transport operations external to the disposal facility are the same, regardless of the selected disposal technology, except that the total number of low-level releases from storage will of course increase if disposal is delayed. Low-level releases from the disposal operation itself may vary with the technology selected. As discussed in previous sections, this latter risk is not well documented for any alternative technology, but it is believed to be very low. Furthermore, only a fraction of the low-level release from disposal operations will be influenced by the choice of an alternative technology. Fortunately, separate and nonconflicting remedies can be recommended to reduce both acute and chronic exposure risks should either prove unacceptable (Chapter 7). The Stockpile Committee is concerned about risk and is aware that it is a significant public concern as well. The conduct of the programmatic risk assessment was a necessary component in the construction of existing facilities. In light of experience with the baseline system since that assessment,

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions Figure 4-4 Potential cumulative consequences of delayed disposal, for constant storage and disposal risk.

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions and taking advanced assessment methods into consideration, the committee recommended an update of this assessment in the form of individual, site-specific risk assessments for each storage site (NRC, 1993c). The committee believes that sufficient investigations were performed in the cited PEIS, including site-specific variations, to reach the conclusion that, with the possible exception of the Blue Grass site where estimated storage and disposal risks were about equal, storage risks exceed risks from disposal operations at all sites. In some cases, the storage risk was calculated to be more than 100 times greater than that for disposal with existing disposal technology. These analyses showed that the highest storage risk was at the Aberdeen site due to the presence of ton mustard containers stored outdoors in an area close to airport operations. The committee believes that the new site-specific risk analyses, for which time has been incorporated in the Army's most recent construction schedule (Table 1-3), will confirm the relative risk estimates of the previous assessment. Site-Specific Risk Assessment The Army has committed to performing site- and facility-specific risk assessments for all of the U.S. disposal sites.1 This action follows the recommendation of the Stockpile Committee, discussed above (NRC, 1993c). The Army also performed a preliminary and limited-scope site-specific risk assessment for the Johnston Atoll Chemical Agent Disposal System (GA Technologies, 1987). Site-specific risk assessments using quantitative risk assessment techniques provide a way of assessing the safety performance of the chemical agent and munitions disposal and storage alternatives. Such analyses have been used successfully in other industries as the basis for managing risk, and these have given assurance regarding the safety of the facilities and the handling and processing of hazardous materials. Site-specific risk assessments identify the intersystem and intrasystem dependencies and the human roles in controlling the risks. They provide a quantification of risk from all reasonable causes, including both internal events (plant and plant-people failures) and external events. 1   A site-specific risk assessment accounts for unique features related to the distribution of population (and its transportation patterns) around the site, nearby topography, and land uses (e.g., truck farming versus grazing versus residential). A facility-specific risk assessment accounts for differences in the design of facility components (e.g., the site-specific layout of components; the particular mix and quantities of the on-site stockpile).

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions Each analysis evaluates site-specific hardware, software, procedures, training programs, maintenance practices, and operator response (including site-specific storage facilities and munitions handling activities). The contributions to risk in terms of random failures, common cause failures, multiple failures, and human errors are assessed. The total scope of a site-specific accident risk assessment includes a number of factors. A list of these factors ranging from human reliability and pollution abatement systems to final plant decommissioning is given in Table 4-1. RELATIVE RISKS ASSOCIATED WITH INDIVIDUAL STEPS OF DISPOSAL OPERATIONS The risk analysis (U.S. Army, 1987) supporting the PEIS considered the risks of accidents for individual operations in the baseline system as applied at JACADS. Individual operations involve such activities as punch-and-drain, disassembly (including the operation of rocket shearing machines), in-process agent storage, and furnace/incinerator performance. The results are characterized only in terms of accidental agent release, as opposed to other process pollutants, and whether the release conditions are initiated by internal or external events. The only consequence considered was acute effects from accidental releases. Among the important general results are the following highlights: Overall Accidents resulting from disposal operations have not been calculated to be major contributors to the risk from the overall Chemical Stockpile Disposal Program. The accident risk is estimated to be greater for those activities of the program having to do with storage, transportation, and munitions handling. The risk from disposal operations is less. The accident risk from storage and transport to the disposal facility, although much higher than for disposal operations, is still very much lower than the routine risk to the public of accidental death from such non-stockpile-related causes as automobile accidents, falls, drownings, or industrial accidents. These latter risks, for each individual, are in the range of a 10-4 to 10-5 chance of fatality per year. The storage risk shown in Figure 4-3, about 1.8 × 10-2 expected fatalities per year for a total at-risk population of 4 million, works out to an individual risk of 0.45 × 10-8 fatalities per year.

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions TABLE 4-1 Scope of Risk Assessment Assessment Factors ▪ Human Reliability ▪ Equipment Performance ▪ Integrated Process ▪ Mitigation Systems ▪ Pollution Abatement Systems ▪ Waste Handling ▪ Munition Handling Operations ▪ Release Condition ▪ Atmospheric Dispersion ▪ Health Effects Evaluation ▪ Operating, Maintenance and Management Procedures ▪ Emergency Response ▪ Decommissioning   SOURCE: PLG, Inc. Internal Events The historic frequency of releases associated with accidents initiated by equipment failure and human error has been low, so as to make its contribution to risk insignificant. In terms of magnitude, the largest amount of agent vapor release would occur for a metal parts furnace explosion simultaneously with ventilation failure. A similar event in terms of frequency and consequence involves a munition detonation in the explosive containment room vestibule with subsequent fire spreading to unpacked munitions outside the vestibule in the unpack area. The highest-frequency event calculated covering disposal operations would be initiated by an inadvertent feed of an unpunched munition containing a burster to the dunnage incinerator (10-2 per year for mines; ˜ 10-3 per year for other munitions). The feature of this event that keeps it from

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions being an important contributor to risk is the small amount of agent that might be involved, perhaps up to 15 pounds. Thus, the consequences were estimated to be small. External Events Aircraft crashes dominate the external event risk among all the risks that were considered. Fire and lightning are real but unevaluated risks. These apply to both storage and disposal operations. For example, the frequency of accident scenarios induced by the crash of a large aircraft is estimated to be higher at the Anniston Army Depot than it is for the Tooele Army Depot. The frequencies of earthquake-induced accident scenarios also vary (e.g., they are generally higher for Tooele than for Anniston). The earthquake scenario having the highest relative frequency involves a munition fall and subsequent fire, leading to the detonation of munitions in the unpack area. The low frequency, of the order of 10-6 per year, keeps this scenario from being a major contributor to risk. Other external events such as tornadoes and meteorites did not initiate accidents of sufficient frequency to be a significant contributor to risk. RELATIVE RISKS OF ALTERNATIVE DISPOSAL TECHNOLOGIES The risks of accidents associated with agent and munitions disposal using the baseline system are now reasonably well identified. Detailed site-specific quantitative risk assessments are possible for both accidents and normal operations. In contrast, for many of the potential alternative technologies, full systems have not been designed, risks are not fully identified, and indeed, technical feasibility is yet to be proven. Obviously, comparable quantitative risk assessments cannot be completed at this time. It is the opinion of the Stockpile Committee, based upon its members' cumulative scientific, industrial, and programmatic experiences in similar technologies (see Appendix G) that, given sufficient time and funding, all technically feasible disposal technologies could be engineered to roughly the same level of safety. The immediate consequence of this opinion is that no potential alternative technology has been eliminated because it was believed to be inherently unsafe. However, if ''sufficient time and funding'' turns out to involve long delays and substantial increases in cost, engineering judgments can help identify alternatives that should be dropped from further consideration. At the same time, as more is learned of the alternatives and as new site-specific risk

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions analyses become available, the potential for any increased safety when using alternative technologies should be reexamined. The risks considered in this program are indeed low in comparison to other more familiar risks. However, they are important to program planning and management in at least three ways: Risk is the major source of public concern. Although there are widely varying opinions on the significance of the different sources of risk, risk is virtually the only common ground for communication among interested parties. Anxiety and uncertainty tend to be the response to low-probability, high-consequence, involuntary risk. The risk of stockpile storage, while not high, is nevertheless gradually increasing due to such phenomena as destabilizing propellants and deteriorating containment. Uncertainties in both the stockpile condition and the disposal schedule leave open the possibility of a significantly increased hazard if disposal is sufficiently delayed. The result is the need for a better understanding of not only the levels of risk involved but also the contributing factors to that risk. Experience has indicated that improved understanding best comes from a systematic and deliberate analysis of the risks and their sources. ISSUES OTHER THAN RISK In addition to the principal criterion of risk minimization, a number of other factors are of concern to the communities and to the committee. Socioeconomic Impacts The construction, use, and decommissioning of the disposal facilities will cause community impacts related to jobs, transportation, and property values. As noted, there are serious concerns about the safety and environmental performance of the operations that need to be addressed. One concern involves a general distrust of incineration and fears about proper operation of such facilities. Use of an alternative technology to allay fears is a possibility, but the actual trade-offs between an alternative technology and any component of the baseline system needs to be considered explicitly in any decision about which technology to use. Another of the concerns is that the facility might continue to be used for hazardous waste disposal after all agents and munitions have been disposed.

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions The facility would then become a continuing hazard in itself, even after the intended elimination of the stockpile hazard. The committee has considered this concern explicitly in its recommendations and agrees that continued use for nonstockpile destruction purposes imposes additional, but unquantifiable, risk. Open Decision-Making Process A formal and well-publicized process for public communication, as well as for receiving and considering public inputs to decisions on technology choices and monitoring operations, needs to be established on a continuing basis. Successful public participation processes for some hazardous waste facilities have included features such as having members of the board of directors nominated by the community, having funds to support expenses incurred by community members of a plant co-management committee, requiting the support of local government for any expansion of treatment capacity or change in technology, and guaranteeing nearby property values and crop prices (Castle, 1993; OFFE, 1993). A growing literature summarizes case studies and approaches that have been used to ensure public safety and to enhance the public's confidence in the safe operation of potentially risky hazardous materials facilities (Hance et al., 1988). Schedule Congress has set a date for completion of demilitarization by the end of 2004. Independent of that, schedule affects total risk since delays prolong exposure to cumulative storage risk. Although the committee is committed to expeditious disposal, it believes that schedule is secondary, per se, to overall safety. It is operating on the assumption that this date might well be extended if a valid reason for doing so exists. Cost Although cost is not a consideration in the committee's criterion regarding the reduction of overall risk, it is nevertheless a consideration after some adequate level of safety is reached. Storage costs, total capital investment, operating costs, costs of disposal of residual solid materials, etc., are all included in the costs of the disposal program. The committee could not develop accurate costs for as yet undeveloped alternatives. However, because the incremental costs for the process equipment in the baseline system are

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions low compared with the total costs of the facilities and the overall program, cost has not been considered an important selection criterion in judging any technology. Nevertheless, cost is a factor that will ultimately be considered by Congress in committing federal revenues to the Army Chemical Stockpile Disposal Program. The disposal program also presents the unusual possibility that maximum safety may correspond to minimum cost. This follows from the foregoing indications that risk may be dominated by storage rather than disposal operations. Investment in the research for and development of improved disposal technology could then be accompanied by delays that increase the total cumulative risk regardless of the characteristics of the new technology. Citizen Involvement In setting one primary criterion, cumulative total risk minimization, the committee recognizes that, while acting in what it believes to be the best interests of the public and of the environment, its views and recommendations cannot possibly satisfy all involved participants. The final selection of a technology or process should be made by considering the recommendations of the local citizenry as well as those of this committee. Effective mechanisms must be established to develop and communicate citizens' concerns throughout the selection process, and thereafter in the operation of the facilities. This must be accomplished in a manner that ensures safe and effective operation and monitoring of the disposal facility. SUMMARY Based on previous analyses, the risk from accidents associated with continued storage of chemical agents and munitions, with one possible exception (where the risks may be equal), exceeds the risk of disposal operations using the baseline system. Until disposal is completed, the risk of releases from storage will persist, and if the storage period is extended sufficiently, that risk will increase significantly as the munitions and agent deteriorate. The Army's programmatic risk assessment was an adequate initial evaluation of the overall acute risks associated with storage and disposal using the baseline system. Both acute and chronic risks associated with storage operations and transport external to the disposal facility will be unaffected by the selection of disposal technology. Chronic risks from disposal operations

OCR for page 61
Recommendations for the Disposal of Chemical Agents and Munitions are not well documented for any technology, but are believed to be very low for the baseline system and recommended alternatives. In the absence of schedule and cost constraints, a number of agent disposal technologies can be engineered to satisfactory levels of safety, with regard to both acute and chronic health and environmental effects. To varying degrees, however, alternative technologies will incur both cost and delays (which add cumulative storage risk penalties) to complete necessary research and development work, and to demonstrate adequate performance. Finally, the committee is keenly aware of citizen concerns other than risk associated with the Chemical Stockpile Disposal Program. These concerns include socioeconomic impacts, public participation in the decision-making process, program schedules, and program costs. It is important that all be considered in the program to eliminate the stockpile. The committee's emphasis on cumulative total risk minimization is an attempt to be sensitive to the overriding concerns of the public and of the committee itself.