_,.-;-\. - d n Y Y. 2: i LA. A = iu - J. /I ._._ -___ .__._ -.---- .._._ -----.-.._-.-. _ ___ .__..._ - _ _ _. .- .._ I.. .- -.----. II --.. -- Resources, Community, and Economic Development Division B-239166 June 8,lQQO The Honorable Mike Synar Chairman, Environment, Energy, and Natural Resources Subcommittee Committee on Government Operations House of Representatives Dear Mr. Chairman: At your request, we have assessed implementation of the Safe Drinking Water Act program by the Environmental Protection Agency (EPA) and the states. Specifically, this report discusses (1) the extent to which community water systems (systems serving permanent residents) have complied with the act’s requirements for monitoring water supplies and meeting drinking water standards, (2) the effectiveness of EPAand state enforcement programs in ensuring compliance with these requirements, and (3) the impacts of new drinking water requirements on the program. As arranged with your office, unless you publicly announce its contents earlier, we will make no further distribution of this report until 30 days after the date of this letter. At that time, we will send copies to other appropriate congressional committees; the Administrator, EPA; and the Director, Office of Management and Budget. We will also make copies available to other interested parties. This work was performed under the direction of Richard L. Hembra, Director, Environmental Protection Issues, who may be reached at (202) 275-6111. Other major contributors to this report are listed in appendix IV. Sincerely yours, / Y J. Dexter Peach Assistant Comptroller General Executive Summary While improved treatment by public water supply systems has virtually Purpose eliminated the threat of typhoid fever, cholera, and other diseases that once plagued the nation’s drinking water supplies, waterborne disease outbreaks continue to occur. In recent years, public health and environ- mental officials have also become increasingly concerned about a proliferation of man-made chemical contaminants found in drinking water supplies. Many of these contaminants have been linked to cancer, birth defects, and other serious health problems. Concerned about the effectiveness of the Environmental Protection Agency’s (EPA) efforts to safeguard water supplies from contamination, the Chairman, Environment, Energy, and Natural Resources Subcommit- tee, House Committee on Government Operations, asked GAOto assess key elements of the agency’s safe drinking water program. Among the issues GAOexamined were (1) the extent to which community water sys- tems (systems serving permanent residents) have complied with requirements for monitoring water supplies and meeting drinking water standards, (2) the effectiveness of state enforcement programs to ensure compliance with these requirements, and (3) the impacts of new drink- ing water requirements, mandated by the 1986 amendments to the Safe Drinking Water Act. The Safe Drinking Water Act, enacted in 1974, required EPAto establish Background drinking water standards, covering certain drinking water contami- nants, to be met by the nation’s 58,000 community water systems. The act also required water systems to monitor the water delivered to con- sumers to detect whether it exceeds the standards. EPAgenerally dele- gates primary responsibility to the states for enforcing the monitoring requirements and drinking water standards and for overseeing commu- 6 nity water systems. In 1986 Congress amended the act to significantly increase the number of contaminants to be regulated, strengthen EPA enforcement authority, and establish various other requirements. To comply with the program, a water system must collect samples of its drinking water and have them tested in an approved laboratory for a variety of contaminants. The test results are then reported to the state, which analyzes the data to determine whether (1) the system has met its monitoring requirements and (2) the water quality has violated a drink- ing water standard. The state, in turn, reports violations of monitoring requirements and drinking water standards to EPA,which maintains a national data base on system compliance. Page 2 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Jbcutlve Summary Systems with serious and/or chronic problems are deemed to be in “sig- nificant noncompliance” and receive priority attention for enforcement. EPApolicy requires states to take “timely and appropriate” enforcement action against these significant noncompliers. The policy sets specific time frames within which the state must take enforcement action and establishes criteria for what actions qualify as appropriate. If a state does not take timely and appropriate action in a particular case, that water system becomes a target for EPAenforcement. Despite EPAreports that water systems are largely meeting monitoring Results in Brief requirements and drinking water standards, GAOfound substantial evi- dence that (1) violations are probably going undetected and unreported by water systems and (2) identified violations are going unreported by states to EPA.Although states have a number of quality assurance mea- sures at their disposal that would alleviate these compliance problems, financial constrairits are leading many to curtail these measures. Based on its detailed review of enforcement cases in six states, GAOalso found that enforcement is often neither timely nor appropriate against significant noncompliers. More important, state enforcement actions are often ineffective in returning these violators to compliance. Of particu- lar concern is that many of the significant violations GAOreviewed, some posing serious health risks, have persisted for years. In some of these cases, states took no enforcement action; in others, enforcement action did not bring about compliance or did so only after lengthy delays. Addi- tional hindrances to a return to compliance include (1) the difficulties small systems have in paying for costly corrective actions and (2) tech- nical barriers such as a lack of alternative water sources. The addition of new regulatory requirements to the drinking water pro- & gram will make an already complex problem more difficult for EPA,the states, and water systems. EPAestimates, for example, that these new requirements, which will affect nearly all water systems, will add about $2.6 billion in annual compliance costs. With compliance becoming increasingly difficult $01 water systems, EPAand the states are bracing for substantially increased regulatory costs of their own in assisting sys- tems, monitoring systems’ compliance, and taking enforcement action against violators. Page 3 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Executive Summary Principal Findings Considerable EPAdata show that the large majority of community water systems com- Noncompliance ply with all drinking water requirements and that only a small percent- age are in significant noncompliance. However, interviews with state and federal program managers and EPAstudies reveal that the number of violations is considerably understated. The reasons reflect problems at the water system, state, and federal levels: . At the water system level, some violations are not detected due to sam- pling error by water system operators. EPAand state officials cite as causes (1) the increasingly technical nature of water sample collection and (2) inadequately trained or inexperienced operators, particularly at small systems. GAOfieldwork in three EPAregions and six states also dis- closed cases of intentional falsification of data, although the full extent of this problem is unclear. l At the state level, EPAstudies show that (1) some identified violations are not reported to EPAand (2) some states have adopted policies sus- pending or restricting certain EPAmonitoring requirements. As a result of the policies, water systems are not performing all required water tests, Both findings were substantiated by GAO'Swork. . At the federal level, EPAlacks key data needed to determine water sys- tem compliance and must rely instead on state tracking systems which, in some cases, are known to be inadequate. Among the most effective tools states use to help ensure compliance are periodic visits to water systems called sanitary surveys. During these surveys, state officials may test water quality, observe operator proce- dures, and/or check the condition of equipment. However, although EPA regulations require the surveys, financial constraints are leading many a states to cut back on these and other quality assurance activities. Enforcement Not Timely, In evaluating state enforcement efforts, GAOreviewed 95 cases of signifi- Appropriate, or Effective cant noncompliance in six states and found that states took timely and appropriate enforcement action, as defined by EPApolicy, in only 24 cases. Of greater concern is that state efforts to return significant noncompliers to compliance were often ineffective, or succeeded only after years of continuous violation. GAOfound, for example, that 46 of the 95 cases persisted for more than 4 years. While extenuating circum- stances (such as funding constraints among small water systems) help to explain the problem in some instances, in others, states allowed lengthy Page 4 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program F;xecutive Summary delays for voluntary compliance before initiating appropriate enforce- ment action, Where state actions were delayed or ineffective, EPArarely stepped in and exercised its own enforcement authority, Acknowledging the need to improve enforcement, the EPAAdministrator announced in April 1989 that the agency will encourage states to increase the number of enforcement actions to be taken against noncom- plying water systems. As discussed in chapter 3, GAOsupports this pol- icy but emphasizes the need to focus on the quality of such actions-on whether they will achieve compliance. Impacts of New As problematic as compliance and enforcement already are, they may Regulations become more so in coming years as EPAestablishes new standards and other requirements for water systems. EPAestimates that annual water systems’ costs may increase by $2.5 billion in the coming years. Already faced with funding shortages, states will see their own regulatory costs increase by hundreds of millions of dollars annually. With water sys- tems and states both experiencing increased difficulties implementing the program, EPAalso expects a correspondingly greater burden on its own resources. Ultimately, the program’s success will depend on the ability of systems, states, and EPAto meet these resource needs. In chapter 4, GAO discusses recent EPAinitiatives to deal with the financial problems and cites infor- mation from GAO'Sown review supporting the need for such efforts. Among GAO'Srecommendations in chapter 2 is that the Administrator Recommendations improve water systems’ compliance by (1) encouraging more consistent use of state-sponsored operator certification and training programs in order to reduce operator error, (2) improving internal controls to detect and deter intentional falsification of sampling data, and (3) encouraging more consistent implementation by states of sanitary survey programs. GAOalso makes a number of recommendations in chapter 3 to improve compliance through better EPAand state enforcement. GAOdiscussed the facts in this report with EPAofficials, who generally Agency Comments agreed with their accuracy. GAOhas included their comments where Y appropriate. However, as requested, GAOdid not obtain official com- ments on a draft of this report. Page 5 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Contents Executive Summary 2 Chapter 1 8 Introduction How EPA’s Drinking Water Program Works Objectives, Scope, and Methodology 9 ’ 10 Chapter 2 13 Many Water Systems EPA’s Statistics Indicate Substantial Compliance Definition of “Significant Noncompliance” Understates 13 14 Are Not Complying Seriousness of Many Compliance Problems With Monitoring and Violations Are Going Undetected or Unreported by Water 17 Contaminant Level Systems States Are Underreporting Violations to EPA 26 Requirements EPA Data Management Problems 28 Conclusions 29 Recommendations 31 Chapter 3 33 Enforcement Safe Drinking Water Act Requirements and EPA Enforcement Policy 33 Inadequate in States Rarely Met EPA Enforcement Criteria 35 Deterring and Many Significant Noncompliers Have Remained in 39 Correcting Noncompliance for Years Long-term Significant Noncompliers Often Involve 40 Noncompliance Difficult Compliance Issues Conclusions 48 Recommendations 50 a Chapter 4 51 New Drinking Water Major Drinking Water Requirements Mandated by the 1986 Amendments 51 Requirements Will EPA Has Proposed or Promulgated Regulations for Most 52 PoseAdditional New Drinking Water Requirements Compliance and Resource Constraints Will Increase Compliance and 52 Enforcement Problems Enforcement EPA Efforts to Help Water Systems and States Obtain 58 Challenges Additional Resources Conclusions 60 Page 6 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program - Contents Appendixes Appendix I: Adverse Health Effects Associated With 62 Ingesting Contaminated Drinking Water Appendix II: Scope and Methodology for Review of State 66 Enforcement Cases Appendix III: Summary of EPA’s Efforts to Issue New 70 Drinking Water Regulations Required by the 1986 Amendments to the Safe Drinking Water Act Appendix IV: Major Contributors to This Report 72 Tables Table 3.1: State Performance Against EPA Timeliness 36 Criteria Table 3.2: Number of Months GAO Review Cases Met SNC 40 Criteria Table 4.1: Estimated Annual Costs to Water Systems for 53 Implementing 1986 Amendments to the Safe Drinking Water Act Table 4.2: Estimated Initial and Annual Costs to States for 56 Implementing 1986 Amendments to the Safe Drinking Water Act Table 1.1: Adverse Health Effects Associated With 63 Drinking Water Contaminants Table II. 1: Review Cases by Reporting Category and by 68 State Abbreviations CDC Centers for Disease Control EPA Environmental Protection Agency GAO General Accounting Office MCL maximum contaminant level SNC significant noncomplier Page 7 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 1 I Introduction Most Americans take the availability of safe drinking water supplies for granted. In the United States over the past several decades, improved treatment practices and drinking water regulations have virtually elimi- nated such diseases as typhoid and cholera and have reduced the inci- dence of other debilitating diseases. However, outbreaks of some diseases, such as giardiasis, continue to occur. According to the Centers for Disease Control (cDc),~ 485 disease outbreaks caused by the ingestion of contaminated water,” involving over 110,000 individuals, were reported to state and local health authorities between 1971 and 1985. Moreover, CDCand Environmental Protection Agency (EPA) officials esti- mated that the actual number of such outbreaks is 20 to 80 percent more than the number reported because contaminated drinking water is often not suspected as the cause of illness. Perhaps of greater significance are the long-term adverse health effects caused by ingesting contaminated drinking water. EPA and CDC research- ers have found that over time, some drinking water contaminants can damage the liver, kidneys, heart, and other body organs. While uncer- tainty exists about the potential long-term effects of certain contami- nants, health and environmental officials are concerned that prolonged consumption of some contaminants, even at low levels, can cause cancer, leukemia, and other serious health problems. To protect the public from these risks, the Safe Drinking Water Act, enacted in 1974, required EPAto establish (1) standards or treatment techniques for contaminants that could adversely affect human health and (2) requirements for monitoring the quality of drinking water sup- plies and for ensuring the proper operation and maintenance of water systems. To oversee the program, states assuming “primacy” (responsi- bility) would, with EPA’S oversight, assess compliance with these stan- dards and requirements and take enforcement action when warranted. a If necessary, EPA would step in with its own enforcement, as detailed under the act’s federal enforcement procedures. By the mid-1980s however, EPA had not regulated many contaminants, and had not revised most interim drinking water regulations, which had ‘An agency within the U.S. Department of Health and Human Services, the CDC is responsible for, among other things, identifying and defining preventable health problems, surveilling diseases, and developing and applying disease prevention and control. ‘As defined by CDC and EPA officials, an outbreak occurs when two or more persons experience a similar illness after ingesting drinking water and evidence implicates the water as the source of sick- ness A single case of chemical poisoning constitutes an outbreak if evidence shows that the water has been contaminated by the chemical. Page 8 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program , Chapter 1 Introduction been in effect since 1977. Compliance with drinking water requirements was uneven, with the most problems occurring at small community water systems serving nearly 25 million people, or approximately 11 percent of the nation’s population. Congress amended the act in 1986 to (1) establish deadlines to accelerate EPA'Sefforts to set standards, (2) establish a monitoring program for certain unregulated contaminants, (3) require EPAto issue criteria for determining which surface water sys- tems must filter their supplies, and (4) require disinfection of all public water systems. The act also gave EPAnew authority to ensure timely and effective enforcement of all drinking water regulations. Under the drinking water program, EPAand the states rely heavily on How EPA’s Drinking water systems:’ to demonstrate compliance with monitoring require- Water Program Works ments and with water quality standards called “maximum contaminant levels” (MCI,). To meet monitoring requirements, the water system opera- tor periodically must collect water samples at the locations and frequen- cies specified by WA and have the samples tested in an approved 1aboratory:l The frequency of required monitoring varies depending on the contaminant, the water source (surface water or groundwater), and, in some instances, the size of the population served. Some states test the system’s water directly, at least for a portion of the contaminants being monitored. The test results are then reported to the state, which (1) determines whether the system has met its monitoring and reporting requirements and (2) analyzes the test data to determine whether the system has vio- lated any drinking water standard. If violations have occurred, the stat,e is responsible for taking enforcement action, giving priority to systems deemed to be in “significant noncompliance.” As detailed in chapter 2, such a designation is based on the frequency and/or magnitude of viola- 6 tions. EPApolicy requires states to take timely and appropriate enforcc- ment action against significant noncompliers (SNC), and to that end establishes time frames for such action and criteria for determining what actions qualify as appropriate. EPAis responsible for taking enforcement action in cases where the state does not take such action. “As defined in the act, a public water system provides piped water for human consumption and musl have at least 1R scrvicc connections or regularly serve at least 26 individuals. ‘To be approved, testing laboratories must demonstrate to EPA or the states that they arc capable of performing the analytical measurements required in the drinking water program and obtaining accu- rate results. Page 9 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter1 Introduction States report to EPAviolations of monitoring and/or water quality requirements by systems, and whether they took timely and appropriate enforcement action against SNCS.As part of its oversight responsibility, EPAmaintains a national data base on system compliance and publishes national compliance statistics. Citing the importance of safe drinking water to public health, the Chair- Objectives, Scope,and man, Environment, Energy, and Natural Resources Subcommittee, House Methodology Committee on Government Operations, asked GAOto assess EPAand state implementation of the Safe Drinking Water Act program. Based on the Chairman’s initial request and our subsequent discussions with his office, GAOagreed to 9 determine the extent of community water systems’ compliance with monitoring requirements and drinking water standards under the act; . evaluate EPAand state enforcement actions to bring violators into com- pliance; and l determine the status of EPAefforts to set standards for contaminants specified in the 1986 Safe Drinking Water Act amendments, and the effect those standards and other new requirements will have on the program. We performed the bulk of our work at the Office of Drinking Water, at EPAheadquarters; three EPAregional offices, and two state drinking water program offices in each of the three EPAregions. Choosing EPA regions for geographical diversity, we selected Regions I, VI, and X, headquartered respectively in Boston, Massachusetts; Dallas, Texas; and Seattle, Washington. Within the regions we also conducted fieldwork in Massachusetts and Vermont, in Region I, Oklahoma and Texas, in Region VI, and Oregon and Washington, in Region X. Our review focused on b community water systems, which are the primary source of drinking water for most Americans, The review did not address noncommunity water systems.” To obtain information for this review, we collected (1) documents on specific compliance problems, the procedures used to identify and report violations, and enforcement cases; (2) EPAregulatory and economic “Community water systems primarily serve year-round residents, while noncommunity water sys- tems, operating at sites such as campgrounds, lodges, and other public accommodations, serve tran- sients or intermittent users at least 60 days out of the year. For an assessment of EPA’s noncommunitv water svstem orogram in Redon III, see the EPA Office of the Inspector General’s report Nonco&munity kater System Program (ElfiW7-03-0171431928, Sept. 26,.1988). Page 10 GAO/RCED-90.127 Noncompliance Undermines EPA Drinking Water Program ChaPtM 1 Introduction impact analyses; and (3) Federal Managers’ Financial Integrity Act reports (also prepared by EPA). The Financial Integrity Act reports pro- vided information on internal control weaknesses in processing and reporting water system compliance data. We also performed a number of activities to respond to each of the three objectives. To respond to the first objective, we interviewed state and federal program managers and reviewed documents on state quality assurance and internal control programs intended to ensure that water systems comply with drinking water requirements and accurately report compliance data to the state. To assess the reliability of national compli- ance statistics, we (1) analyzed EPAregional studies on the accuracy of the compliance data states report to EPAand (2) studied how EPA'Sdata management system identifies significant noncompliers. In addition, we asked state and EPAofficials about the extent to which they believe water system operators file test results that are erroneous or falsified and about the measures they are taking to detect and deter these practices. For the second objective, we examined state and EPAenforcement poli- cies and procedures, including EPAcriteria on how states should address systems in significant noncompliance with the act’s regulations. To test the states’ enforcement performance, we reviewed 95 cases of signifi- cant noncompliance at 75 community water systems located in six states. We designed our review to include all possible types of cases, including those in which the water system returned to compliance, cases involving enforcement action, and pending cases. Individual review cases within each category were randomly selected. We reviewed the states’ files on the selected cases in detail and interviewed state program managers about enforcement actions and other efforts to return the water systems to full compliance. Where necessary, we followed up on selected cases with EPAregional officials. We chose 1987 as the base year for this review because we wanted to include cases that were initiated after EPA'Senforcement policy was implemented in October 1986. In addition, we wanted SNCviolations that were as recent as possible without being so new that one would not expect the states to have responded. We decided that selecting SNCviola- tions identified during 1987 would allow sufficient time for state and EPAenforcement actions to occur. We reviewed these actions through May 1989. We called each state to update our information as to whether systems that were still in significant noncompliance as of May 1989 remained so as of February 1990. Our results represent only the cases Page 11 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program -- -- - ( Chapter 1 Introduction reviewed and not all significant noncompliance cases nationwide. (See app. II for a detailed description of our methodology in selecting the enforcement cases.) For the third objective, we interviewed officials in EPA'SOffice of Drink- ing Water to determine the status and implications of EPAefforts to set additional drinking water standards and implement other requirements contained in the 1986 Safe Drinking Water Act amendments. To obtain additional information on the impact of these activities on the program, we also interviewed EPAregional water program officials, drinking water program officials in the six states, and representatives from the Associ- ation of State Drinking Water Administrators, the American Water Works Association, and the League of Women Voters. We also reviewed EPAregulatory and economic impact analyses for promulgated and pro- posed drinking water regulations and examined impact studies con- ducted by the Association of State Drinking Water Administrators and the League. Our audit work was conducted between July 1988 and March 1990 in accordance with generally accepted government auditing standards. During our review, we discussed our audit findings with EPAofficials responsible for implementing and enforcing the Safe Drinking Water Act program, and have incorporated their comments where appropriate. However, in accordance with the wishes of the requester’s office, we did not solicit formal comments from EPAon a draft of this report. Page 12 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water SystemsAre Not Complying With Monitoring and Contaminant Level Requirements Despite EPAreports showing that community water systems are largely complying with drinking water requirements, the extent of compliance has been considerably overstated. Part of this discrepancy can be explained by how EPAdistinguishes between “significant” violators and the much larger universe of other violators. We found that under EPA'S criteria, violations not classified as significant can include cases of chronic noncompliance with monitoring requirements and instances where MCLSare exceeded by substantial margins. In addition to this definitional issue, we found deficiencies in the detec- tion and reporting of violations at each major point in the regulatory process, from the time a system samples its water supply to the time EI'A records the system’s compliance status in its national data base. Specifi- cally, we found deficiencies at the water system, state, and federal levels. l At the water system level: Some violations are probably going unde- tected because of errors in the way system operators take and test water samples. We also found instances where test results were intentionally falsified, although the extent of this problem is unclear. While states can take certain internal control and quality assurance measures to increase water system compliance, such measures are sometimes not imple- mented effectively, or not implemented at all. . At the state level: States, for a variety of reasons, are not reporting some water system violations to EPA.The existence of state and regional policies that revise or suspend certain EPAmonitoring requirements con- tributes fundamentally to the problem of underreporting. . At the federal level: EPAlacks key information needed to determine water system compliance and must rely instead on state tracking sys- tems In the absence of such information, the agency’s data management system sometimes produces incomplete or overstated compliance rates. 1, EPA'Sannual statistics disclose the number of water systems that comply EPA’s Statistics fully with drinking water requirements, systems classified as SNCS,and Indicate Substantial “other noncompliers” that have at least one violation and whose Compliance problems are not serious enough to cause the systems to be classified as SNCS.Using fiscal year 1988 statistics, the most recent year for which complete statistics are available, EPAestimated that (1) 72 percent of all community water systems had no reported violations, (2) only 2 percent of community water systems were classified as SNCS,and (3) about one quarter of the water systems were identified as “other noncompliers”- Page 13 GAO/RCED90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systema Are Not Complying With Monitoring and Ckmtaminant Level Requirements a relatively small percentage considering that a single violation of only one requirement would warrant such a classification, Such statistics would appear to indicate that most water systems are monitoring their water and meeting quality standards as required and that the large majority of violations that do occur are not serious. How- ever, as discussed in the remainder of this chapter, (1) the criteria EPA uses to distinguish between SNCSand other violators minimizes some potentially serious problems and (2) the number of water systems reported to be in full compliance may be overstated by a significant margin, reflecting problems at the water system, state, and federal levels. According to EPAprogram managers, the criteria for significant noncom- Definition of pliance were established to focus limited enforcement resources on the “Significant systems with the worst problems. However, the narrowness of this defi- Nokompliance” nition has excluded substantially more systems with serious and/or chronic compliance problems. Understates Seriousnessof Many Compliance Problems SomeViolators Not Community water systems may be classified as SNCSdepending on either Classified as “Significant” (1) the frequency with which they violate program requirements or (2) the severity of their violations (e.g., the extent to which they exceed a Nonetheless Have Serious contaminant level). For one contaminant group, for example, water sys- Problems tems are required to sample for microbiological and turbidity contami- 4 nants and meet MCL requirements on a monthly basis.’ A system that violates MCL requirements for 4 or more months during any 12-month period would be classified as in significant noncompliance, as would a system that fails to conduct any required monitoring for 12 consecutive months. EPAalso employs an”aggregate” criterion for significant non- compliance, which encompasses water systems with any combination of monitoring or MCLviolations for 12 consecutive months. ‘Microbiological contaminants, such as bacteria, may cause acute health effects, including gas- troenteritis, diarrhea, and hepatitis. Some of the smallest water systems may be allowed to sample their water for microbiological contaminants quarterly instead of monthly, and others sample monthly, but determine compliance with the MCL requirement quarterly. High levels of turbidity, which is a “cloudiness” in water caused by minute suspended particles, may reduce the efficiency of disinfection treatment and mask the presence of microbiological contaminants. Turbidity require- ments apply only to water systems that obtain their water from surface sources. Page 14 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program . * chapter 2 Many Water Systems Are Not Complying With Monitoring and Contaminant Level Requirements However, under such a restrictive definition, water systems with serious and/or chronic compliance problems may be classified as “other,” or “nonsignificant,” noncompliers. For the same contaminant group described above, for example, a system may have three violations of MCL requirements within 12 months-or fail to take any samples for as many as 11 months out of 12-and still not be considered a significant noncomplier. Similarly, a system may have a combination of three MCL violations plus eight monitoring violations within a 12-month period and still not be classified as a significant noncomplier. The SNCcriteria for chemical and radiological contaminants also leave room for serious violations that EPAwould not classify as significant. In general, a water system is considered to be in violation if its test results exceed the MCL for a contaminant. However, for each chemical and radiological contaminant, EPAhas established “SNC levels” that deter- mine when a violation warrants classification as an SNC.Although for some contaminants, the MCL and the SNClevels are the same, for others, the SNClevel may far exceed the MCL. For example, a nitrate violation is considered to be significant noncompliance if the test result exceeds 200 percent of the MCL, while the SNClevel for a selenium violation is 500 percent of the MCL. To find out more about the violators classified as other noncompliers, we asked EPA'SOffice of Drinking Water to break down its compliance data for microbiological and turbidity contaminants. These statistics confirm that many water systems categorized as other noncompliers appear to have serious compliance problems. For example, we found that although only 134 water systems were identified as SNCSbecause they had four or more microbiological MCLviolations within a 12-month period, 602 systems had three within the same period. a Thus, a considerable number of water systems have violations that may be serious but would not cause them to be categorized as SNCSunder EPA'Scriteria. The distinction between SNCSand other noncompliers is important because, as discussed in the next section, the SNCcriteria were established based on available enforcement resources and have tradi- tionally accounted for the majority of EPAenforcement targets. Page 16 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program .-Ii .., ,. ” , ‘;;yY ’ Chapter 2 Many Water Systems Are Not Complying With Monitoring and Contaminant Level Requirements SNC Criteria Largely Although EPAencourages states to take action against other violators, SNCS have been the primary focus of EPA'Senforcement policy and track- Determined by ing system and state enforcement goals. According to EPAdrinking water Enforcement Resources officials, when the agency developed its SNC criteria in the mid-1980s, its Rather Than Public Health major consideration was controlling the enforcement work load. Public Considerations health was also considered, but only to the extent that the criteria allowed states to focus their limited enforcement resources on the worst problems. The SNC criteria replaced an earlier classification of high pri- ority violators called “persistent violators.” The shift to SNCS did cut the number of high priority enforcement targets considerably. For example, the officials estimated that under the previous criteria, 6,000 to 7,000 “persistent” violators of microbiological and turbidity requirements were identified in fiscal year 1984. However, when the new SNC criteria were applied to fiscal year 1985 violations in the comparable contaminant category, the number of SNCS dropped to approximately 2,000. While we agree that it makes sense to set priorities for enforcement when resources are scarce, EPA'Scurrent SNC criteria minimize the num- ber of potentially serious problems that may be targeted for enforce- ment action. Officials with EPA'SOffice of Drinking Water acknowledge that the SNC criteria need to be expanded, explaining the need in part by the fact that in some regions, few systems meet the existing criteria. One manager added that the lack of SNCS in a particular region does not indi- cate a lack of water systems with serious problems. In his view, expanding the SNC criteria will help EPAavoid the appearance that the program is not addressing all serious problems. In April 1989 EPAestablished a national work group to consider possible changes to its SNC criteria to better reflect public health considerations. a According to an October 1989 proposal issued by the work group, a number of the changes under consideration will have the effect of increasing the number of high priority enforcement targets. The work group also proposed a three-tiered system for categorizing violations and prioritizing enforcement targets, with the first tier including SNC violations and representing the top enforcement priority. The second tier would include violations that are serious but that have not yet reached the SNC level, and the third tier would contain all other viola- tions. WA plans to complete the redefinition of SNC criteria by the end of fiscal year 1990. Page 16 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 lbny Water Syetems Are Not Complying With Monitoring and Ckmtaminant Level Requirements Beyond the issue of how EPAcategorizes known violations, we found evi- Violations Are Going dence suggesting that violations are either not being detected at all by Undetected or water systems or are being detected but are not being reported. As noted Unreported by Water in chapter 1, states vary as to how much they rely on water system operators to collect and analyze water samples and then report the Systems results to state drinking water authorities. Where water systems, and not states, are responsible for sample collection, the potential for obtaining inaccurate information is greater-particularly in the case of smaller water systems, which are less likely to have trained full-time operators. Based on our discussions with program managers in the six state pro- grams, it appears that violations are, in fact, going undetected because of unintentional errors in the way system operators take and test water samples. The program managers also identified instances of intentional falsification of test results, although most states do not actively monitor the extent of this problem. Unintentional Sampling To obtain test results that accurately reflect water quality, sample col- Error by Water System lectors must follow exacting procedures established by EPAand state regulatory agencies. Consequently, sampling is best done by trained Operators individuals who understand how to take the samples and interpret the results. Untrained collectors are more likely to produce invalid test results. EPAand state program managers told us about a number of circum- stances that may lead to errors by water system operators. In the case of turbidity, for example, inaccurate readings may result if on-site test equipment is not regularly calibrated or maintained. In the case of some organic compounds that are volatile and may dissipate in the air, inaccu- b rate readings may result unless the sample collector seals the container properly, making sure that it contains no air bubbles and is not exposed to sunlight or high temperatures. These officials expressed concern about operator sampling technique and the accuracy of the test results. For the most part, they attributed potential problems to inadequate operator training and the lack of full- time operators or the high turnover among operators at small water sys- tems. They also indicated that errors will increase as additional MCI, and Page 17 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program :,. ,:, :, ‘, Chapter 2 Many Water Systems Are Not Complying With Monitoring and Contaminant Level Requirements other requirements are implemented as a result of the 1986 Safe Drink- ing Water Act amendments. They explained that under these require- ments, more contaminants will be regulated and that many will require increasingly sophisticated sample collection procedures. EPA region X officials, for example, told us that while the quality of labo- ratory analysis in the region is generally good, they are not confident that operators are using proper sample collection techniques and attrib- uted the problem to insufficient training. Despite these concerns, how- ever, the officials note there is little evidence that this problem affects public health. Program managers in Oregon and Washington confirmed that some system operators lack adequate training in key functions such as proper sample collection and maintenance. They indicated that opera- tors of small systems are particularly likely to lack training, and Wash- ington officials said that high turnover among these operators is an additional complicating factor. Operator sampling error is also a concern in New England because approximately 75 percent of the water systems there serve 500 people or less. According to an EPA region I manager, small systems have the most difficulty attracting trained operators and all too often, the person who takes the samples and performs other tasks is “whoever happens to be around.” In contrast, EPA region VI officials were fairly confident about their com- pliance data because the states collect many samples themselves. These officials told us that most region VI states collect organic and inorganic chemical samples themselves. Until recently, Louisiana even collected the microbiological samples for its water systems. However, the officials said that resource constraints resulting from implementation of the 1986 4 amendments to the Safe Drinking Water Act may require some states to turn over these activities to system operators, perhaps increasing problems with compliance data. According to a Louisiana program man- ager, about 2 years ago resource constraints forced the state to establish a fee system to help pay for state sample collection. Then, because the fees were inadequate to cover costs (a situation exacerbated by the 1986 amendments), the state revised its regulations to require system opera- tors to collect their own samples as of July 1, 1990.” Texas officials also “However, water system operators were very concerned about the impact this would have on their own operating costs; the Louisiana Municipal Association filed suit on their behalf and obtained an injunction preventing the state from implementing the new regulations. As of March 1990, state offi- cials were debating a major budget increase for the drinking water program, which would provide sufficient funds for continued state sample collection. Page 18 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program chapter 2 Many Water Systems Are Not Complying With Monitdng and Contaminant Level Requirement8 told us that as the provisions of the 1986 amendments begin to take effect, it will be difficult for the state to continue collecting samples without additional resources. More ConsistentUse of Operator Operator certification and training programs can be an effective means Certification and Training of preventing operator sampling error. These programs are intended to ProgramsCould Help Reduce ensure that water systems are operated and maintained by qualified Operator Error individuals, sampling techniques are properly employed, and the pro- gram generally complies with drinking water regulations. Several years ago, EPARegion VI conducted an informal study, comparing small sys- tems with and without certified operators, to assess the impact of certi- fied operators on small water system compliance. The study found more compliance problems at systems lacking certified operators than at those having them. A study conducted in Utah, analyzing legislation to convert the state’s voluntary operator certification program to a manda- tory one, obtained similar results. However, according to an Office of Drinking Water official, there is no national operator certification program or any regulation requiring states to have such programs themselves. Although information from the Association of Boards of Certification” shows that 45 states have mandatory operator certification programs and 2 others have voluntary programs, the same organization collected data indicating that at least 11 states exempt systems serving 500 people or fewer from having certi- fied operators. An EPAofficial told us that other states use different cri- teria, such as the number of service connections, to exempt small water systems from operator certification. These exemptions are significant because over 60 percent of all community water systems nationwide serve 500 people or fewer. Within the three EPAregions we visited, we found that all but one of the 4 states have some type of operator certification program, although the requirements vary considerably.4 Consistent with EPA’S findings, our findings show that some states exempt small water systems from opera- tor certification requirements because (1) the requirements are consid- ered unnecessarily burdensome and (2) the smallest systems are often operated by part-time employees or volunteers and cannot attract or “The Association of Boards of Certification seeks to improve certification laws and promote certifica- tion as a means of ensuring effective operations by personnel of water utilities and pollution control systems. The organization assists certifying authorities to develop strong administrative programs and effective uniform certification criteria and standards. ‘Within EPA Region I, Rhode Island does not require its water systems to have certified operators. Page 19 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systema Are Not Complying With Monltmlng and Contaminant Level Requirements pay for an operator who meets all qualifications. Although the percent- age of the population served by these systems is low, small water sys- tems can represent a significant proportion of the total number of water systems. For example, systems serving 500 people or fewer account for 76 percent of Louisiana’s systems and 90 percent of New Mexico’s. System operators’ compliance with operator certification requirements also varies considerably from state to state. Although both Oklahoma and Washington officials estimate compliance levels to be over 90 per- cent, the Vermont program manager estimates that fewer than 5 percent of the state’s community water systems have certified operators. He told us, however, that with recent legislation strengthening the state’s authority to enforce the requirement, the state expects to have all sys- tem operators certified by 1993. Massachusetts’ compliance rate is about 50 percent, with small systems largely accounting for the problem. Mas- sachusetts officials explained that the certification program is managed outside of the state environmental agency by a state Board of Certifica- tion, which, in their view, has neither the staff nor the expertise to properly certify all water system operators. Massachusetts program officials introduced legislation to bring the certification program under their control, thus allowing them to administer and enforce the require- ments, but it was not enacted. Under recently issued EPAregulations, all surface water systems are required to have operators that are “approved” to the satisfaction of the state. However, according to an Office of Drinking Water official, EPA has not established any minimum qualifications for water system opera- tors; it will be up to the states to determine what they consider acceptable. In addition to providing the training associated with operator certifica- tion programs, some states have developed special training initiatives that focus on small water systems because these systems tend to have the most compliance problems. Organizations such as the American Water Works Association and the National Rural Water Association also sponsor some training activities. However, no requirements exist for states to conduct training programs, and resource constraints often limit the frequency and content of the training that is provided. Moreover, with the addition of increasingly technical drinking water regulations pursuant to the 1986 amendments, training will need to become more widespread and comprehensive as the program itself grows in complexity. Page 20 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systems Are Not Complying With Monitoring and Contaminant Level Requirements Falsification of Data or Water system compliance is likely to be overstated not only because vio- Manipulation of Samples lations caused by operators’ sampling errors go undetected, but also because some operators may be deliberately falsifying test results, How often this practice happens is unclear, however, because most states do not actively seek it out. While most EPAand state officials we inter- viewed stated they do not believe data falsification is extensive, they did tell us that (1) falsifying test results is relatively easy, (2) incentives to do so will increase, and (3) they had already identified suspected instances of falsification. Data falsification occurs when water system operators intentionally compromise their compliance data to make it appear that they have done the required sampling and that water quality is within acceptable limits. Program officials described several methods of falsification: . Although most water samples are analyzed in approved laboratories, system operators often have control over sample collection and thus have an opportunity to influence test results. One way for water system operators to ensure that test results are within acceptable limits is to take samples from sources known to be free of contamination. . Although system operators are supposed to take daily water samples for turbidity, test them using on-site equipment, and report the results to the state at the end of each month, the operators can simply report plau- sible test results without ever actually testing their water. l To falsify microbiological tests, operators can take measures to elimi- nate any contamination before the sample is tested. For example, boiling or microwaving the sample will kill bacteria, as will rinsing the container with chlorine prior to collection of the sample. EPAand state officials acknowledge that water system operators have several incentives to falsify compliance data: for instance, to avoid hav- b ing to employ costly corrective treatment or having to notify the public that its drinking water is contaminated. In an illustrative case, detected by EPAregion I officials, one operator falsified data so that his system would not have to treat its water. In 1988 EPAdetermined that the oper- ator of a Vermont surface water system had reported exactly the same test results for turbidity every day for months, despite storms, seasonal changes, and other factors that normally affect turbidity levels. There- fore, regional officials observed the system operator for several days and verified that the operator was neither taking nor analyzing samples. In addition, EPAtook and analyzed samples and found that several exceeded the MCL. Page 21 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systems Are Not Complying With Monitoring and Contaminant Level Requirements When confronted with this information, the operator admitted he took samples only 2 or 3 days per month but reported daily tests. He explained to EPArepresentatives that despite several meetings at which state officials expressed concern about the system’s lack of filtration, the local water department board had strongly resisted employing such treatment. According to the operator, he realized that low turbidity results would help the department avoid having to install a filtration system. In December 1988 EPAissued a notice of violation that cited the system’s failure to comply with turbidity monitoring requirements since January 1979. Additional legal action is underway. While most of the EPAand state officials we interviewed said that such problems are not widespread, they all cited cases in which data falsifica- tion had been detected or was strongly suspected. For example, program managers in all six states had identified cases in which reported turbid- ity results were too consistent to be credible. In Oklahoma and Texas, program managers estimated that from 3 to 5 percent of the surface water systems may have falsified data. Oklahoma and Oregon managers also cited instances in which system operators had dosed microbiological samples with chlorine in an attempt to eliminate bacteria. State officials provided other examples of questionable practices that would lower the number of reported violations, including the following: . In Oregon, state officials cited cases in which system operators repeat- edly took turbidity samples until they found one complying with stan- dards; the systems then reported only the satisfactory result to the state. According to these officials, they have been informed by testing laboratory representatives that this practice also occurs with other contaminants. 6 l When Oklahoma officials investigated one system with suspiciously con- sistent turbidity results, the operator said that his predecessor told him to take a water sample, “hold it up to the light, and if it looks pretty clear, give it a .3.” He was also told not to report a result over 1 (the MCL) under any circumstances. While EPAand state officials asserted that such problems are not wide- spread, we found relatively few efforts among them to actively seek out the problems. An EPAregion X official acknowledged that state efforts to detect data falsification occur haphazardly. He noted that state officials identify such cases if they notice incongruous or overly consistent test results and if they have sufficient resources to follow up on them. Wash- ington officials agreed that part of the reason they have not identified a Page 22 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systems Are Not Complying With Monitoring and Contaminant Level Requirementzi larger number of definitive cases of data falsification is that the state has not had the resources to investigate suspicious data. Similarly, pro- gram managers in several other states told us that they investigate only if they notice suspicious data or receive consumer complaints about poor quality water. Still, we did identify a few limited efforts among the states we visited to identify data falsification systematically. An EPAregion I enforcement official obtained, from the Vermont program office, special computer printouts that flagged systems whose turbidity results showed little or no variation. He then determined whether the test results were logical in light of the systems’ condition and other factors. In addition to identify- I ing the case described earlier in this section, the official identified 13 others in the region warranting further investigation. Oklahoma and Texas systematically review monthly turbidity reports to identify ques- tionable data. In addition, Oklahoma’s program manager told us that testing laboratories in the state periodically analyze microbiological samples to detect the presence of excess chlorine, since adding chlorine can disguise microbiological contamination problems. We found nothing in our review to suggest that the majority of water system operators do not make a good faith effort to comply with pro- gram requirements. However, based on the information obtained during our state visits, data falsification may be occurring more frequently than either EPAor state officials suspect. Moreover, as discussed in chap- ter 4, the incentives to do so will increase as water systems are required to comply with the broader and more stringent requirements in the 1986 amendments to the Safe Drinking Water Act. Accordingly, we believe that EPA needs to encourage these types of efforts to detect and deter data falsification to ensure the credibility of self-monitoring under the program. CL Sanitary Surveys Are According to EPAand state officials, comprehensive inspections of water Often Not Implemented systems, or sanitary surveys, are among the most important tools states can use to help ensure water system compliance with drinking water Effectively requirements. In addition to being an overall review of the facility and its operations, sanitary surveys provide states an opportunity to con- duct specific activities that may reduce the potential for both operator sampling error and falsified test results. Such activities may include sampling the water, observing the system operator’s sampling and test- ing procedures, reviewing collection procedures to ensure the operator Page 23 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systems Are Not Complying With Monitoring and Contaminant Level Requirements understands them, and checking the system’s turbidity test equipment to ensure its proper calibration and functioning. EPAregulations also emphasize the importance of sanitary surveys, requiring them as a condition for states to obtain primacy. However, the regulations do not specify what states must do during the surveys or how often states must conduct them. As discussed in this section, we found that sanitary survey programs can vary widely in both frequency and content and that resource constraints are substantially affecting many of these programs. State Emphasis on Sanitary Both Texas and Oklahoma place strong emphasis on sanitary surveys, Surveys Varies conducting them the most frequently of the six states we visited. In both states, survey results are directly linked to enforcement, thus reinforc- ing the importance of ensuring that water systems are properly designed, operated, and maintained. Texas officials told us that they conduct comprehensive surveys of surface water systems annually and groundwater systems biannually. They believe that sanitary surveys are the most important tool in ensuring safe drinking water and that peri- odic sampling only reveals “the tip of the iceberg” about water quality. In Oklahoma, sanitary surveys are conducted more frequently than in Texas, but are somewhat less comprehensive. According to EPAregion VI officials, although all six states within the region have emphasized sanitary surveys, the need to emphasize compli- ance monitoring and enforcement activities will move states away from sanitary surveys and other quality assurance efforts. In Louisiana, financial problems have forced the state to make severe cutbacks in its sanitary survey activities; the number of comprehensive surveys has been reduced by more than two-thirds and the briefer inspection-type surveys have been eliminated almost entirely. According to region X officials, region X states’ sanitary survey pro- grams have also been affected by limited resources and the increased emphasis on enforcement. The Washington program manager told us that his state has not conducted routine sanitary surveys since the late 1970s. Although the state does maintain a field presence through spe- cial-purpose investigations, these reviews generally focus on a specific problem and do not entail a comprehensive system review. The Wash- ington manager told us that the lack of sanitary surveys is the most glaring weakness in the state’s drinking water program. Page 24 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systems Are Not Complying With Monit.&ng and Contaminant Level Requirements Similarly, Oregon’s program managers told us that the state has been conducting sanitary surveys on a 3-year cycle but cannot meet this time frame any longer. According to the fiscal year 1989 state/EPA agree- ment, Oregon now plans to survey its water systems once every 5 years. Here, too, state officials pointed to the lack of resources as the underly- ing cause of the cutback. A program manager in EPARegion I told us that all of the New England states are having financial problems and that a reduction in the fre- quency of sanitary surveys is among the effects. In Massachusetts, pro- gram officials told us that they survey every water system at least once every 3 years. Their goal is to conduct a “short-form” review at each system annually and a comprehensive review every 3 years. Currently, the state conducts approximately 50 comprehensive reviews and 100 shorter ones each year. In Vermont, according to state officials, the state reviews groundwater systems every 3 years and surface water systems annually. Thus, while some states’ sanitary survey programs appear to be compre- hensive, other states’ surveys are either less comprehensive or have been discontinued altogether. Although most program managers agree on the value of sanitary surveys as a quality assurance tool, recent resource cutbacks have made it increasingly necessary to reduce or elim- inate this program element. According to a 1988 review of the costs of implementing the 1986 Safe Drinking Water Act amendments, the single largest expenditure in state drinking water programs is for water sys- tem inspections and sanitary surveys.” Although 43 percent of the respondents cited inspections/sanitary surveys as the most important program activity for ensuring the safety of water supplies and protect- ing public health, 75 percent of the respondents indicated that they were likely to reduce this activity unless they receive additional 1, resources to implement new regulatory requirements. EPA’Sdilemma is that while sanitary surveys are required as a condition of retaining primacy, the agency is reluctant to “force the issue” in the wake of state funding problems. When asked whether states that have discontinued their sanitary surveys, such as Washington, are violating EPAregulations, an official with EPA’SOffice of Drinking Water stated that because EPAhas not established any requirements or criteria for “EPA and the Association of State Drinking Water Administrators conducted a joint survey in 1988 and issued a report in June 1989. A total of 36 states and one territory responded to the association’s questionnaire, Page 26 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systems Are Not Complying With Monitoring and Contaminant Level Requirements how frequently these reviews must be conducted, the states might not be in technical violation as long as they have conducted sanitary surveys at some point. This approach, however, seems to reflect a policy of “looking the other way” at state noncompliance with the requirement for sanitary surveys. We believe that this policy not only contravenes the spirit of EPA’S own requirement, but is ill-advised in light of the qual- ity assurance these programs provide. As discussed in chapter 4, to the extent that EPAcan assist states in dealing with the resource issue, sani- tary survey programs may be salvaged and a valuable quality assurance tool preserved. However, the first step is for EPA to clarify its regulatory requirement to make plain when sanitary surveys are required and how they should be conducted. Once the state receives water system test results, it determines whether States Are the results indicate a violation of either monitoring or MCL requirements Underreporting and reports identified violations to the EPA regional office. The regional Violations to EPA office periodically audits the state-reported compliance data through “data verification studies,” which are intended to detect systemic reporting problems. We found that the EPA verification studies contained substantial evidence that states are underreporting violations to EPA. Information on state reporting practices obtained during our site visits confirmed the findings reported by EPA. To conduct their “data verification” studies, EPA regional officials select a random sample of water systems and then compare the state’s raw compliance data, such as monitoring records and laboratory reports, with the information states reported to the agency’s data management system. In total, we reviewed data verification studies conducted by all 10 EPA regional offices covering 38 states plus Puerto Rico and the Vir- gin Islands.” b Our examination of EPA'Sstudies disclosed that although the percentage of total errors identified from state to state varied widely, the percent- age of errors found to involve the underreporting of violations was con- sistently high. For example, EPA regional offices reviewed the accuracy of microbiological compliance data in all 40 states in which data verifi- cation studies were performed. EPA found reporting errors in over 25 percent of the sample cases in 15 states; in 6 of the 15 states, the error “We obtained the most recent study available in each case. The frequency of EPA’s reviews depends on the resources and priorities of individual EPA regional offices. The studies included in our analysis used data from fiscal years 1984 through 1989; 50 percent of the studies were based on data from fiscal year 1987 or later. Page 20 GAO/RCED-99-127 Noncompliance Undermines EPA Drinking Water Program Chapter2 Many Water Systema Are Not Complying With Monitoring and Ckmtaminant Level Requirement43 rate exceeded 50 percent. In 25 states, over 75 percent of all reporting errors involved the underreporting of violations. EPA reviewed the accu- racy of compliance data for turbidity, chemical contaminants, and radio- logical contaminants in fewer locations, but the results were generally consistent with what the agency found in its examination of data for microbiological contaminants, Reasonsfor The studies cite a variety of reasons why violations were underreported. Underreporting of Four EPAregions indicated that reporting violations to EPAwas some,- times given a lower priority than other activities. At least five regions Violations reported that state program staff were not always sufficiently knowl- edgeable about federal regulations to recognize all violations and report them properly. In Region II, for example, one state was incorrectly clas- sifying water systems with monitoring violations as “non-reporters.” State and regional policies that revise or suspend certain monitoring requirements are a major factor contributing to the problem of underre- porting violations. Specifically, according to EPA’Sdata verification stud- ies in at least six regions, monitoring for chemical and/or radiological contaminants was suspended under state policies. For example, the fis- cal year 1988 study for Michigan disclosed that the state had not enforced federal monitoring requirements for inorganic chemicals since 1982 because EPAand the state agreed that limited resources should be used to test for volatile organic compounds instead. Other states estab- lished similar policies for a variety of reasons. In addition to revealing the suspension of monitoring requirements, the data verification studies disclosed that some states did not have systems to track water system compliance with chemical and/or radiological monitoring requirements and thus could not determine whether monitoring violations had occurred. During our site visits, we also identified regional and state policies that led to the underreporting of violations. For example, under an EPA region X policy, microbiological monitoring violations are not reported to the national data base as long as a water system has taken over 50 percent of the required samples during a month. In the case of turbidity, the regulations require daily monitoring, but in Region X, monitoring viola- tions are not reported as long as the water system has tested its water at least 20 days during the month. As a result of such policies, monitoring is partially completed or not required at all, and monitoring violations are not being reported to EPA. Page 27 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program . Chapter 2 ‘1 Many Water Systems Are Not Complying With Monitoring and Contaminant Level Requirements The problem presents a dilemma for EPA:On one hand, states or EPA regions may present a compelling case why such policies may be war- ranted. On the other hand, EPAis tolerating state and regional policies that directly conflict with existing requirements. Whether or not the policies are justified, our primary concern is that the present situation undermines a program that relies primarily on adherence to published regulatory requirements. In addition to encouraging noncompliance, these policies also lead to statistics that mislead EPAmanagers and the public into believing that required monitoring is being conducted and that compliance is being achieved. Once states determine whether water systems have violated monitoring EPA Data Management or MCLrequirements, they report identified violations to EPA,where the Problems data are compiled and summarized to develop national compliance rates. However, because EPAis missing key information, it cannot determine accurate compliance rates for many contaminants. This complex problem arises, in part, from the fact that EPA'Sdata man- agement system is an “exception” system; states only report violations. Thus, a lack of reported violations is taken to mean a water system is in full compliance. However, in the case of chemical and radiological moni- toring requirements, the fact that no violations have been reported for a particular system could mean that the water system is in full compli- ance-but it could also mean that (1) required monitoring has not been conducted, but the compliance period has not ended yet or (2) a viola- tion has been detected, but has not yet been reported. These ambiguities are complicated by inconsistencies in how states track these violations and report them to EPA.The required monitoring fre- quency for chemical and radiological contaminants is every 1,3, or 4 b years, depending on the contaminant and type of water source, and EPA requires no set point within these periods when tests must be conducted. As already noted, EPA'Sdata verification studies disclosed that some states do not have systems to track compliance with chemical and radio- logical monitoring requirements. In addition, even when states report violations to EPA,the agency does not know when the compliance period begins and ends for a particular contaminant and water system or whether the state is reporting violations when they occur or at the end of the compliance period. Although states are required to report such information, some do not or report data subject to multiple interpreta- tions. According to an official from EPA'SOffice of Drinking Water, the Page 28 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systema Are Not Complying With Monitoring and Contaminant Level Requirements agency is unwilling to assume that all systems for which no monitoring violations have been reported are in full compliance. Because of these concerns, EPAcannot use its data management system to generate an SNClist for chemical and radiological monitoring viola- tions and must instead rely on the states to compile such lists. In addi- tion, EPAcan only report limited information on overall compliance. For example, in fiscal year 1987 EPAreported that 2.3 percent of all water systems were SNCSbecause they had never tested for chemical and radiological contaminants and, for the remaining 97.7 percent, could report only that these water systems had monitored at least once since the requirements took effect in June 1977.7 However, these figures do not reveal much about compliance, since some water systems have been required to monitor some contaminants annually since 1977. There could well be a significant gap between “monitoring at least once since the requirements took effect” and full compliance. Unless states improve their tracking systems, EPAwill continue to have problems in determining the extent of water system compliance with chemical and radiological requirements. Although published EPAdata show that (1) most water systems are com- Conclusions plying with monitoring and contaminant level requirements and (2) the relatively few violating systems have generally committed minor infrac- tions, we found considerable noncompliance with these requirements. Part of this discrepancy is a matter of definition: Under EPA’S criteria, “significant noncompliance” may exclude both cases of chronic noncom- pliance with monitoring requirements and cases where MCLSare exceeded by substantial margins. Beyond these definitional concerns over how EPA categorizes known vio- , lations, many violations are either not being detected at all, or are being detected but are not accounted for in EPA compliance data. This problem reflects weaknesses in the manner in which (1) water systems sample water supplies, analyze the results, and report them to the state; (2) the state reports on systems’ compliance with monitoring and MCLrequire- ments; and (3) EPA determines compliance. 7As of March 1990, EPA’s fiscal year 1988 annual compliance report was still in draft. The draft report identified the number of water systems that were SNCs as a result of chemical and radiological monitoring violations, but was silent as to the compliance status of the remaining water systems. Page 29 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 ManyWaterSystemsAre Not Complying With Monitoring end Contaminant Level Requirements At the water system level, we found some violations are being commit- ted unintentionally by system operators who take and test water sam- ples incorrectly. This, we found, is particularly true at smaller water systems, which often do not have the resources to hire and retain highly trained operators. The addition of increasingly technical drinking water regulations in coming years will only add to the problem. While some states sponsor training programs for operators of small water systems, such programs will need to become more widespread and comprehensive as the drinking water regulations themselves grow in complexity. More consistent use of operator certification programs can also help avert such problems. While EPAis planning to require some water systems to have state-approved operators, the agency does not plan to establish any minimum criteria for how this requirement should be met. A second problem at the water system level is the potential for deliber- ate falsification of compliance data or manipulation of the test itself to produce the desired result. While the extent of this problem is unknown, we found that (1) falsifying data and manipulating test results are rela- tively easy to accomplish, (2) ample evidence exists that the practices are occurring, and (3) incentives to engage in these activities will increase because violations of new drinking water regulations may require costly treatment measures or other facility improvements. While some states have undertaken modest efforts to detect such problems, few have a systematic program to identify and investigate potential data falsification. We believe that EPAshould encourage states to more actively seek out data falsification and should provide guidance on ways to do this. The few efforts identified in our review are a good starting point for building an active and effective deterrence program. Our review also suggests that better compliance by water systems could L be achieved through more consistent implementation of sanitary surveys. Although most state program managers agree that sanitary surveys are a valuable quality assurance tool, some states’ survey pro- grams omit important functions or have been discontinued altogether because of funding difficulties. Compounding this problem is EPA’S ambiguous policy toward the surveys: Although EPAregulations require the surveys, states without them may not be considered in violation because the agency has not established how frequently they must be conducted. We acknowledge the difficulties for EPA and the states in supporting such programs as program responsibilities expand during an era of lim- ited resources. However, we believe that an unenforced requirement Page SO GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 2 Many Water Systems Are Not Complying With Monltorlng and Contaminant Level Requirements reduces the program’s credibility and that EPAshould either properly implement or eliminate it. Given the value of the surveys as a needed quality assurance tool, we believe the preferable alternative is to assist states in finding alternative ways to fund such programs. Such assis- tance could also be used to encourage states to retain or establish other quality assurance programs, such as operator certification and training programs. EPA'Scurrent efforts to help states find alternative funding sources to support these and other activities are discussed in chapter 4. In addition to the problems affecting water system compliance, another is that the states are not reporting some violations to EPAbecause of state policies that suspend or restrict federal monitoring requirements. While in certain cases, states may present a compelling case why a par- ticular requirement is counterproductive, we believe that an open disre- gard for existing requirements provides a chance for abuse and undermines the program’s credibility. In addition to encouraging non- compliance, these policies also lead to statistics that mislead EPAmanag- ers and the public into believing that required monitoring is being conducted and that compliance is being achieved, We believe that EPA needs to evaluate such policies and-within the constraints of the Safe Drinking Water Act-decide whether changes should be made to existing regulations. Once those decisions are made, however, the agency needs to ensure that the regulations are observed and enforced. At the federal level, we found that EPAis missing key information it needs to track water systems’ compliance and thereby perform its over- sight responsibilities. Without such information, EPA'Sdata management system provides incomplete information about compliance with certain requirements. Finally, an additional tool needed to encourage compliance is a credible b enforcement program. Such a program helps deter deliberate noncompli- ance problems such as those discussed in this chapter and provides sys- tems with the incentive to meet their responsibilities under the drinking water program. As discussed in the following chapter, however, improvements need to be made to ensure that EPA'Senforcement pro- gram stands as a credible deterrent to noncompliance. We recommend that the Administrator, EPA,follow through on the Recommendations agency’s reexamination of its SNCcriteria in order to more comprehen- sively identify chronic noncompliance with monitoring requirements Page 31 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program I Chapter 2 Many Water Systems Are Not Complying With Monltorlng and Contaminant Level Requirements and substantial violations of MCL requirements& n addition, we recom- mend that the Administrator improve water systems’ compliance with program requirements by directing the Office of Drinking Water to do the following: l Promote more consistent and effective use of state-sponsored training programs and operator certification programs, which would reduce operator error. These efforts should include guidance on (1) how train- ing programs should help water systems deal with the increasingly tech- nical drinking water regulations now being promulgated by EPAand (2) what the minimum criteria should be for state operator certification requirements. 9 Evaluate the extent to which intentional falsification of test data or manipulation of the water sampling process may be occurring. The Office of Drinking Water should also provide guidance to the states on how to best discourage these practices by water systems and to detect them when they do occur, so that appropriate enforcement action may be taken. l Encourage states to implement sanitary survey programs more consis- tently. Specifically, the Office should clarify to the states its ambiguous policy on whether sanitary survey programs are required. In addition, the Office should encourage all quality assurance efforts-including operator certification and training as well as sanitary surveys-by assisting states in finding alternative ways to fund such programs (as discussed in ch. 4 of this report). l Evaluate state policies that suspend or restrict federal monitoring requirements and determine (within the constraints of the Safe Drinking Water Act) whether modifications should be made to existing regula- tions. Once these policies are evaluated, the Office should ensure that the states observe the final decisions and enforce the regulations. Page 32 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 3 EnforcementInadequatein Deterring anil Correcting Noncompliance In recent years, EPAhas put more emphasis on enforcement as a means of returning violators of safe drinking water requirements to compli- ance. EPAfirst implemented criteria for identifying water systems that deviate significantly from these requirements in fiscal year 1986, saying it would give high priority to enforcement against these systems. The agency also required states to take timely and appropriate enforcement action against these violators and, in fiscal year 1987, provided the states with specific guidance as to what constitutes such action. However, our review of SNCenforcement cases in six states indicated that state enforcement actions often do not meet the timeliness and appropriateness criteria. Moreover, such actions are often ineffective in returning SNCSto compliance, or do so only after lengthy delays. Our case reviews disclosed that some systems had been in noncompliance for many years and, in several cases, became a health threat despite long- term enforcement efforts. Many of these long-term cases have been com- plicated by situations in which system ownership is in dispute or sys- tems have had difficulty paying for needed corrective action. When states do not take timely and appropriate enforcement action, the Safe Drinking Water Act requires that EPAeither issue an administrative order or commence a civil action. However, EPAhas taken such actions infrequently. Moreover, EPArarely has stepped in on its own initiative and exercised its enforcement authority when state action was delayed or ineffective. In the 1986 amendments to the Safe Drinking Water Act, Congress Safe Drinking Water granted EPAenforcement authority to issue administrative orders Act Requirements and (orders that systems take action to comply) and impose administrative EPA Enforcement penalties up to a total of $5,000 for noncompliance with such an order. In addition, the existing civil penalty authority was increased from l Policy $5,000 per day to $25,000 per day. Congress also removed the require- ment that violations be shown to be willful as a prerequisite for obtaining penalties. Another key change in the 1986 amendments was to require enforce- ment action against all violations of the drinking water regulations. If the state does not act, the law requires EPAto issue an administrative order or commence a civil action. Previously, states were only required to report to EPAon the steps being taken (which may or may not have Page 33 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program ,. .“‘, .: ,. .b Chapter 3 Enforcement Inadequate In Deterrhg and ‘, Correcting Noncompliance included an enforcement action) to bring the violator back into compli- ance. If the state action was deemed to be unsatisfactory by EPA,the earlier statute provided only that EPA“may commence a civil action.” Rather than taking enforcement action against all violations, EPA responded to the 1986 amendments with an enforcement policy that focused primarily on the most significant violators, or significant noncompliers.’ This approach reflected constrained resources among EPA and state regulators and the knowledge that some violations are more serious than others. Furthermore, rather than requiring state action within 30 days of the state’s learning of a violation (as the act requires), EPAruled that a state can meet the agency’s timeliness criteria if it takes enforcement action within 8 or 14 months after a water system is deter- mined to be in significant noncompliance.z EPAdefined four types of enforcement actions as appropriate: . Formal administrative orders or compliance orders compelling a water system to comply with drinking water requirements. The state regula- tory agency usually issues these orders to the water system directly, although some states require prior approval by their attorney general. The orders may provide for penalties if the water system does not comply. . Referral of a civil judicial case to the state attorney general. Upon receiving a civil referral, the state attorney general, or in some instances the local district attorney, files suit in civil court seeking a court order forbidding future violations and compelling the water system to take measures to come into compliance. . Filing of a criminal judicial case in an appropriate state court. . Negotiation of an informal bilateral compliance agreement by represen- tatives of the water system and the state program office. The agreement must be signed and must contain compliance schedules indicating the l steps that will return the water system to compliance. According to EPA’Senforcement policy, states should choose an enforce- ment action based on the seriousness of the violation, its circumstances, the water system’s compliance history, and the economic benefit the ‘As noted in chapter 2, although EPA encourages states to take action against ah violators, SNCs have been the primary focus of EPA’s enforcement policy and tracking system and state enforcement goals. ‘An enforcement action is required within 8 months for microbiological and turbidity monitoring and MCL violations, and for total trihalomethane monitoring violations. The ll-month deadline applies to chemical and radiological monitoring and MCL violations. Page 34 GAO/RCED-30-127 Noncompliance Undermines EPA Drinking Water Program Chapter 8 Enforcement Inadequate in Deterring and Cwre&ng NoncomplIance system derives from noncompliance. In any case, enforcement should quickly correct the violation, deter future noncompliance, and be fair to the affected parties. To determine how successfully states were implementing EPA’Spolicy on States Rarely Met EPA timely and appropriate enforcement, we reviewed actions by six states Enforcement Criteria against 76 water systems. The systems had a total of 96 SNCviolations.” Overall, we found that states took timely and appropriate action in 24 of the 96 cases, or about 25 percent of the time. Among the specific find- ings discussed in this section are the following: 9 States took timely action less often against systems that had a record of serious and continuous violations for many years (i.e., a record of viola- tions beginning prior to the time EPA’S enforcement policy was fully implemented in October 1986). In these cases, states generally either took no action or did so only after lengthy delays. . States took inappropriate enforcement action most frequently in the more recent cases involving compliance agreements. In these cases, states often did not meet EPA’S criteria on when such actions should be taken or what they should contain. Many State Enforcement Of the 95 cases in our review, water systems returned to compliance Actions Against Long-term without enforcement action in 17 of the cases. Under EPA policy, states were required to take enforcement action in the remaining 78 cases. We SNCsWere Untimely found that states met EPA’S timeliness criteria in 43 of these 78 cases? As table 3.1 shows, cases of untimely enforcement most frequently involved long-term SNCviolations. “Appendix II contains a detailed discussion of the methodology we used to select SNC violations for review. 4To determine whether state enforcement actions were timely in accordance with EPA criteria, we counted the number of months from the end of the quarter in which an SNC violation was first identified to the date of the enforcement action. In the case of long-term violators, which would have qualified as SNCs prior to when the SNC criteria were fully implemented in fiscal year 1987, we began the count with December 1986, the end of the first quarter in fiscal year 1987. When the only enforcement action in a case occurred prior to the SNC identification date or, in the case of long-term SNCs, prior to December 1986, we gave the state credit for a timely action. Page 36 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 3 ’/ Enforcement Inadequate in Deterring and Correcting Noncompliance Table 3.1: State Performance Against EPA Timeliness Criteria SNC violations Long-term SNC identified in fiscal violations ( re- year 1987 fiscal year 1i 87) Total Criteria met -.-_---.-- System returned to compliance without enforcement action. .---.. 12 5 17 States took timely enforcement action. 19 24” 43 Criteria not met State enforcement action was late. ------~- 7 9 16 States took no enforcement action and system did not return to -_-compliance (as of..~__... May 1989). 3 16 19 Total 41 54 95 aThis figure includes 11 cases in which we gave the states credit for enforcement actions that occurred prior to the implementation of EPA’s enforcement pokey. In five instances, states took enforcement actron from 2 to 14 months before the policy took effect, and in the remaining six cases, the enforce- ment action occurred from 20 to 40 months earlier. Of the 41 recent cases (i.e., cases first identified in fiscal year 1987), 31 were resolved in accordance with EPApolicy as of May 1989. One expla- nation for the relative ease with which these cases were resolved is that a high proportion involved easily correctable problems. For example, over half of the cases involved systems identified as SNCSbecause they had never conducted required monitoring for chemical and/or radiologi- cal contaminants. To return to compliance, the water systems needed only to collect the applicable water samples and obtain acceptable test results. The results of our analysis for the 54 long-term SNCviolations were less satisfactory. Of the 54 cases, 29 were resolved in accordance with EPA policy-either the water systems returned to compliance without enforcement action or the states took timely action. As noted in table 3.1, however, 11 of the 24 “timely” actions actually occurred prior to the implementation of EPA'S policy, some of them by a matter of years. Subsequent enforcement actions were not taken despite continued non- compliance. In the remaining 25 cases, the states did not meet EPA’Stime- liness criteria: . In nine cases, enforcement actions were late, with states exceeding the deadlines by 6 months or less in six cases and by a year or more in three. Page 36 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program ;... Chapter 3 Enforcement Inadequate in Deterring and Cbrecting Noncompliance . In another 16 cases, enforcement action had never been taken. As of May 1989, the duration of these violations averaged over 6 years.” Several factors contributed to the states’ inability to meet EPA timeliness criteria when dealing with these long-term SNC violations. The lack of timely action stems in part from the states’ reliance on efforts to achieve voluntary compliance, even after years of recalcitrance by system own- ers. We noted situations in which the states postponed appropriate enforcement action until long after serious compliance problems were first identified. In one case, for example, a system had not performed required monitoring for any of the regulated contaminants since June 1980, when it took one sample for microbiological contaminants. Even so, the state’s first enforcement action did not occur until October 1987. Still, while recalcitrance explains part of the problem, some cases are difficult to resolve simply through enforcement. In some of these cases, states delayed or did not take enforcement action in favor of working with the water system to resolve difficult problems encountered in locating alternative water sources and obtaining financing for drilling new wells or connecting to neighboring systems. In other cases, systems had been abandoned and the states could not find parties willing to take responsibility for operating the systems. The facts behind these difficult cases illustrate the problems states face in achieving the ultimate goal of the enforcement program- to bring SNCS back into compliance. These issues are discussed in more detail later in this chapter. Many State Enforcement As noted earlier, water systems returned to compliance without enforce- Actions Are Not ment action in 17 of the 95 cases we reviewed, and, according to EPA policy, states were required to take action in the remaining 78 cases. We Appropriate found that states met EPAappropriateness criteria in 33 of the 78 cases. l In another 26 cases, states’ actions were not appropriate,” and states took no enforcement action in 19 cases. All 26 cases in which state “These systems were classified as in significant noncompliance when EPA fully implemented the cri- teria in fiscal year 1987. However, all of these cases involved longstanding violations (predating 1986) that would have qualified the water systems as SNCs at the time the violations were first committed. “The 69 cases in which states took enforcement action involved 66 individual actions, including 37 bilateral compliance agreements, 17 administrative orders, and 12 civil referrals. The number of enforcement actions is different from the number of cases because (1) states took multiple enforce- ment actions in 16 cases and (2) in 9 instances states used a single action to address multiple viola- tions (cases) at the same water system. Page 37 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program chapter 8 Enforcement Inadequate in Deterring and Cwre&hg Noncompliance actions did not meet EPA'Sappropriateness criteria involved bilateral compliance agreements.7 Bilateral ComplianceAgreements We reviewed the extent to which bilateral compliance agreements met Usually Do Not Meet EPA’s specific EPAcriteria for their content and use. According to EPAguidance, AppropriatenessCriteria such agreements must be written, be signed by both parties, and include a schedule with interim milestones and a final date for when compliance will be achieved. The guidance also states that actions stronger than bilateral compliance agreement& should be used when the water system has had a long history of violations or has violated the terms of a previ- ous compliance schedule. Based on our review of the case files, 31 of the 37 bilateral compliance agreements did not meet EPAappropriateness criteria. Ten were not appropriate for multiple reasons. The most common problem, found in 28 instances, was that the agreements were never signed by water sys- tem representatives. Such signatures are important because they indi- cate a commitment to take the required corrective action. In seven cases, we determined that issuing a bilateral compliance agree- ment was inappropriate in light of the water system’s poor compliance history. For example, a municipal water system had almost continuous monitoring violations since late 1982. Documents in the case file indicate that the system operator was reluctant to test the water because a dete- riorating distribution system and chronic pressure problems made it likely that the test results would exceed the MCL. Tests periodically con- ducted by the state confirmed such contamination. In July 1984 the state threatened legal action if the system did not initiate testing within 45 days. The system complied briefly, but has rarely taken required microbiological samples since August 1984. The state finally imposed a compliance schedule in March 1989, noting, “It is evident that the City a has failed to voluntarily make more than minimal efforts to insure the potability of the water. . . .” 7To determine whether state enforcement actions sctuslly qualified as “appropriate,” we reviewed the applicable case files to verify that the actions took place and that the enforcement documents had been signed by the appropriate parties and formally issued. In instances where there may have been some ambiguity about whether the state took an appropriate action, our approach was to give the state credit for the action. When states reported that an enforcement action applied to a particular violation even though the action did not specifically cite the SNC violation, we gave the state credit for the action ss long as it met EPA’s appropriateness criteria. Similarly, where states took multiple enforcement actions to address a single caee, we gave them credit for sn appropriate action in that case if any one of the actions met appropriateness criteria. Page 88 GAO/RCJZD-S&127 Noncompliance Underminea EPA Drhldng Water Program 4 Chapter 3 Enforcement Inadequate in DeterrIng aud Cwrecting Noncompliance Similar circumstances existed for the other six compliance agreements we deemed inappropriate because of the water systems’ poor compli- ance history. For example, in three instances, states issued compliance schedules after earlier formal enforcement actions had proved ineffective. We also found that four compliance agreements did not contain interim milestone dates, despite the fact that the agreements required substan- tive corrective action, Two compliance agreements did not include final completion dates. When corrective action entails a major construction project, milestone dates are essential for state regulators to monitor a water system’s progress toward achieving compliance. Other Enforcement Actions Meet Because EPAessentially leaves it up to the states to determine what Appropriateness Criteria should be included in administrative orders and civil referrals, they need only be issued to be “appropriate.” Hence, we determined that all 29 administrative orders and civil referrals reviewed were appropriate. Ultimately, however, the primary goal of these actions is to bring about compliance, and some of them are not achieving this end. As discussed in the following sections, ineffective enforcement action partially explains why some water systems have remained in significant noncom- pliance for years. One of the more striking observations to be made about the 95 enforce- Many Significant ment cases we reviewed is the length of time many of the water systems Noncompliers Have have remained in significant noncompliance. Table 3.2 shows that as of Remained in February 1990, nearly half the cases had met the SNCcriteria for over 4 years. In 31 of these 46 cases, water systems were still in significant Noncompliance for noncompliance at that time. Years Page 39 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program . Chapter 3 Enforcement Inadequate In Deterring and Correcting Noncompliance Table 3.2: Number of Months QAO Review Cases Met SNC Criteria Cases returned to compliance as of Cases still SNCs as Number of months 2/90a of 2/90b Total 6to12 13 0 13 13to18 12 0 12 19to24 7 0 7 25 to 36 2 3 5 37 to 48 4 8 12 49 to 60 -~ 8 4 12 61 to 72 6 5 11 Over 72 1 22 23 Total- 53 42 95 Tar cases included in this column, the number of months denotes the time elapsed from when the system first qualified as an SNC to when the system returned to compliance. hFor cases included in this column, the number of months denotes the time elapsed from when the system first qualified as an SNC to February 1990. Although we used compliance data from fiscal years 1983 through 1987 to determine when our review cases first qualified as SNCS, we identified a number of instances in which the water systems had serious violations that would have qualified them as SNCS even before fiscal year 1983. For example, in five cases, the water systems had been subject to “boil water orders” for 10 or more years as a result of known or suspected microbiological contamination.” In eight other instances, the case files contained evidence that chemical or radiological MCL violations, which were serious enough to warrant classifying the system as an SNC under current criteria, had existed since the late 1970s or early 1980s. There is no simple explanation for why some water systems remain in Long-term Significant significant noncompliance for years. Many of the cases involve multiple a Noncompliers Often problems, and it was often difficult to single out a particular problem as Involve Difficult the primary factor in delaying a water system’s return to compliance. Ineffective enforcement-by states and EPA-is clearly an important Compliance Issues contributing factor in the delays in resolving some of these cases. How- ever, our case reviews disclosed other issues that contributed to the dif- ficulty of achieving compliance, including (1) financing the cost of expensive corrective actions, (2) resolving difficult technical problems when water treatment alone was insufficient, and (3) sorting out legal ‘When state regulatory officials determine that there is a significant threat of microbiological con- tamination, water systems may be ordered to notify consumers that they must boil their drinking water before using it. Page 40 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program 4 Chapter 3 Enforcement Inadequate in Deterring and Correcting Noncompliance disputes over system ownership to identify parties responsible for meet- ing drinking water requirements. SomeEnforcement Actions Earlier, we noted that states have frequently not taken required Taken Are Not Effective enforcement action in a timely manner and that some actions have not met EPA'Sappropriateness criteria. In many instances, however, enforce- ment actions satisfying the EPAcriteria have been taken but have not achieved their ultimate objective -to bring violating systems into com- pliance. We found this to be particularly true of civil referrals. Under EPA'Senforcement policy regarding civil referrals, states do not have to file suit in court before reporting an appropriate action. They may take credit for an appropriate enforcement action when the referral is made to the state attorney general.” According to officials from EPA'S Office of Drinking Water, because filing the cases in court is outside the control of the state regulatory agency, program officials should not be held accountable for any delays that occur after the referral to the state attorney general. Although the EPAofficials expressed concern about the possibility of civil referrals’ “dying on the attorney general’s desk,” they had no information on the extent to which this may be occurring. We found that 7 of the 12 civil referrals in our enforcement case reviews had not been filed as of September 1989. Significantly, in only one case among the unfiled referrals had the water system returned to compli- ance.l(’ Two of the remaining six unfiled referrals involved a single case-a municipal water system that violated the microbiological MCL continuously since 1982. The state made the first referral in July 1985 after the town failed to comply with a consent order. The case was not filed in court because the state decided to negotiate a second consent order instead. However, after continued noncompliance and the town’s a failure to initiate corrective actions, the state made a second civil refer- ral in January 1987. Although that action was also not filed, town residents finally passed a bond vote in March 1987 to finance the filtra- tion plant needed to correct the problem. “While EPA’s enforcement policy states that the referral should be made to the state attorney general, some states, such as Oklahoma, require that the referral be made to a local district attorney. “‘The state dropped the civil action in this case and categorized the water system as returned to compliance when the state determined that the system no longer met the federal definition of a com- munity water system because it regularly served fewer than 25 individuals. The case was closed 30 months after the referral date. Page 41 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 3 Enforeement Inadequate in Deterring and Cwrectlng Noncompliance The other four water systems also had long-term compliance problems that had not been resolved; as of September 1989, they had had SNC- level violations for, on average, over 5 years. For these unfiled referrals, the time that had elapsed from the referral date through September 1989 was 19 months, 43 months, 53 months, and ‘75 months. Interviews with officials in five of the six states confirmed difficulties in using the civil referral process. In Oklahoma, for example, program offi- cials must work through the local district attorneys to file civil cases. The program manager told us that while some of these attorneys are cooperative, others want nothing to do with drinking water cases unless an imminent threat to public health exists. As a result, according to the state program manager, Oklahoma officials are referring fewer and fewer cases for civil action. Oregon officials told us that they are reluc- tant to use the civil referral process because state attorneys bill the drinking water program by the hour for the legal assistance, and a single case can be very expensive. In contrast, Texas has devoted sufficient resources to its civil referral process to help ensure that drinking water cases are acted upon in a timely manner. The state attorney general has an environmental protec- tion division with about 20 attorneys, including 8 to 10 who are dedi- cated specifically to drinking water cases. Program officials told us that the only limitation on the number of civil referrals is the program office’s ability to compile the information needed to support them. Dur- ing fiscal years 1987 and 1988, the program staff referred a total of 92 water systems to the attorney general’s office for civil action. As of Jan- uary 1990,45 of these cases had been filed in court, and most had been resolved with an agreed final judgment (consent judgment) and a civil penalty. Of the remaining 47 cases, 13 were closed (largely because the I, issues that initially brought about the action were resolved), and 34 were as yet unfiled and still open. State Efforts to Deal With Concerned about the effectiveness of the civil referral process, EPAcon- Ineffective EnforcementActions sidered requiring states to have the authority to issue administrative Have Only Partially Succeeded orders and penalties in order to retain primacy and requested comments on this matter in proposed regulations. However, after commenters indi- cated that their state legislatures would be reluctant to grant additional enforcement authority, EPAopted, in December 1989, to make no changes pending the results of its research on existing state enforcement authorities and their effectiveness. Officials in EPA’SOffice of Drinking Water expect to complete this study by the end of fiscal year 1990. Page 42 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 3 Eniorcement Inadequate in Deterring and Cwre&hg Noncompliance Nevertheless, 5 of the 6 states we visited obtained authority to issue administrative orders as an alternative to the cumbersome civil referral processI However, we found that the usefulness of this authority may be limited unless (1) it is also accompanied by the authority to assess meaningful penalties and (2) the procedures that must be followed can achieve results in a timely fashion. Washington program officials told us that the enactment of administra- tive order authority in 1986 was largely responsible for eliminating a large backlog of enforcement cases awaiting action. However, according to Washington officials, the real key to effective administrative author- ity is the state’s ability to assess penalties for failure to comply with an order. These officials also say they have had some success using penal- ties as a bargaining tool to force water systems to take corrective action In contrast, Vermont does not have authority to assess administrative penalties. Administrative orders may contain a threat of penalties if a water system fails to comply, but the state must go to court to collect them. As a result, Vermont has rarely assessed any penalties. At one time, Massachusetts also had to refer cases to the state attorney general to get a civil penalty assessed, creating the same problems that adminis- trative orders were intended to solve-backlogs of low priority cases at the state attorney general’s office. The problem was alleviated after the drinking water program obtained its own authority to issue penalties in 1986. Another potential problem with administrative order authority is the process required for its use. For example, EPAregion VI officials told us that Louisiana has never issued an administrative order and, in fact, has rarely initiated the process because it is so cumbersome. The procedure requires the water system to have three opportunities to return to com- a pliance before the state can issue the order. A Louisiana program man- ager confirmed that the administrative order process is lengthy and said that the state does not have enough program staff to handle the work load. While EPAis understandably reluctant to force states to adopt specific enforcement authorities and practices, we believe the agency needs to ensure that states have some method to carry out their enforcement responsibilities effectively and to return violators to compliance. EPA ’ ‘As noted above, Texas officials believe the state has devoted the necessary resources to make the civil referral process work effectively. Page 43 GAO/RCED-90-127 Noncompliance Undermines EPA mg Water Program ‘I Chapter 3 Enforcement Inadequate in Deterring and Correcting Noncompliance does conduct an annual program review in each state to evaluate its ability to implement the requirements of the Safe Drinking Water Act. In conducting this review, EPAshould examine whether those states that choose to rely on civil referrals have access to sufficient resources within the attorney general’s office to obtain effective action. Similarly, where states have adopted administrative order authority, the agency should examine the need for (1) administrative penalty authority and (2) a workable process for issuing the orders. EPA Follow-up Enforcement According to EPA'Senforcement policy, SNCSthat have neither returned RestrictedWhen State Actions to compliance nor been subject to appropriate state enforcement actions Are Ineffective go on an “exceptions list” and become potential targets for EPA enforce- ment, Of the 95 cases included in our review, we examined the 38 cases reported to be on this list to determine whether EPAhad taken enforce- ment action. Evaluating EPAefforts in the 38 cases was somewhat prob- lematic because some cases did not actually qualify as “exceptions.“I~ Notwithstanding this classification problem, we found that EPAenforce- ment policy excludes some serious enforcement problems from this list and thus from EPAfollow-up enforcement. In a number of the long-term SNCcases, states had taken enforcement actions that were appropriate under EPAguidelines, but the actions were not effective in returning the systems to compliance-sometimes long after the violations first began. In such cases, EPAis authorized to step in and exercise its own enforcement authority to resolve the compliance problems. When asked why eight such cases were not targeted for EPA follow-up enforcement action, EPAregional officials told us that no fed- eral action was taken because (1) the state was actively tracking the case, (2) state enforcement appeared to be sufficient, (3) the state did not request intervention, or (4) they assumed that the state was han- dling the case. b However, we found that these cases were lingering on with little or no progress and, in at least one instance, presented a potentially serious health risk. In this instance, a water system had serious violations since 1980 that, under the current criteria, would warrant classification as an SNC.State officials tracked the system-issuing notices of violation, ‘%tates had taken appropriate action in 13 cases, and in another 7, the systems had actually returned to compliance. We found that EPA had initiated some type of enforcement action in 5 of the remaining 18 pending cases, and in several others, EPA’s decision to forego enforcement action appeared reasonable. For example, four pending cases involved long-term fluoride MCL violations within EPA Region VI. Although EPA regional officials did issue proposed administrative orders in two similar cases, further action had been suspended pending the results of a feasibility study on fluoride removal. Page 44 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 3 Enforcement Inadequate in Deterrlug and Correcting Noncompllance making site visits, and imposing a boil water order-for several years. However, the first state action that was appropriate under EPA guide- lines did not occur until May 1987, when the state imposed a compliance schedule. Local residents subsequently failed to approve formation of a water district, which would have made them eligible for state grant funds, and thus did not meet compliance schedule milestones. A new compliance schedule was established in June 1989, and at the same time, the state exempted the system from microbiological monitoring requirements. A state official explained that despite the apparent health risk-of the few microbiological samples that have been taken, nearly all have tested positive-the exemption was granted to show good faith and provide an incentive for the system to make improvements. He said that the users know the system is contaminated and have been told to boil their water. When asked why EPAhad not stepped in on this case, an EPAofficial told us that as long as the state has taken action and is making progress, the agency does not interfere. Discussions with EPAofficials in all three regions confirmed that despite EPA'sexpanded enforcement authority under the 1986 amendments to the Safe Drinking Water Act, EPApolicy is generally to defer to the states on enforcement matters. They told us that the states have the primary enforcement responsibility and that EPAis reluctant to intervene unless states request assistance. Region I started issuing formal notices of vio- lations to all new SNCS beginning in 1988 but otherwise relies on the states to refer cases for enforcement action. Although Region VI actively solicits enforcement referrals from its states, the region does not inter- fere when states do not request its involvement. Region X also relies on state referrals. Based on the results of our case reviews, we believe that in light of the large number of long-term SNCS and the problems with ineffective enforcement in some states, EPAshould be more aggressive in initiating enforcement actions itself when state action clearly has been delayed or ineffective. Beyond the assistance this would provide in returning some long-term SNCS to compliance, such a stance would emphasize that the ultimate goal of any enforcement action is to achieve compliance. Page 46 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 3 Edorcement h&equate in Deterring and (iMmdng Noncompliance High Compliance Costs While improving state and EPAenforcement will address some of the and Other Difficult Issues long-term compliance problems in the drinking water program, this is by no means a complete solution. Some SNCSpresent intractable problems Also Hinder Timely that enforcement actions may not cure-regardless of whether the Corrective Action actions meet EPAenforcement criteria. Even where corrective actions are underway, a major project, such as a filtration plant, can take several years to complete. In the meantime, these systems continue to be in non- compliance. Such problems, either individually or in combination, pre- sent significant obstacles to achieving full compliance and help explain why many SNCShave been out of compliance for lengthy periods. Small Water SystemsHave The difficulty of paying for needed system improvements was a recur- Difficulty Paying for Costly ring issue throughout our case reviews. Although the circumstances in Corrective Actions each case were unique, one common element was the small size of most SNCS.Two-thirds of the 75 water systems we reviewed served 500 people or fewer, and 87 percent serve 3,300 people or fewer, percentages that reflect the makeup of the SNCuniverse overall. The small size of these water systems affects their ability to finance corrective actions and, to some extent, compete for grants and loans. In some instances, the resource problems are compounded by water rates that are artificially low because users are not charged the true cost of providing drinking water. However, even substantial increases in water rates may not be enough to finance corrective actions. For exam- ple, in one of our review cases, a water system serving 126 people had to install new water lines, a storage tank, and a water treatment plant to comply with state and federal drinking water regulations. According to the program manager, the system is so poor that three members of the water board had to take out a personal loan to pay for the $2,000 con- struction permit. Total project costs, which were partially funded by the Farmers Home Administration, exceeded $200,000. A Ironically, although the smallest water systems are the least able to afford costly improvements, their size may hurt them when they com- pete for funding, According to the Oklahoma program manager, for example, applicants for state grant assistance are ranked in part on the basis of the number of people affected. Unless an imminent threat to public health exists at a small system, funding for a larger system is more likely to be approved before funding for a small one. Other state officials provided similar information. Page 46 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program /’ Chaptfz 3 Enforcement Inadequate in Deterring and (krecting Noncompliance Technical ProblemsCauseDelays States and water systems also face difficult technical issues, particu- in Achieving compliance larly when the drinking water contains chemical or radiological contami- nants. The cost to install treatment can be prohibitive, and, in some instances, the effectiveness of available treatment technologies is ques- tionable. For example, according to a Texas program manager, there is no viable treatment for fluoride contamination. In other instances, treating the existing source will solve one problem but may create another. For example, Texas officials said although cer- tain radiological contaminants can be treated, the water system then has the problem of disposing of the radioactive waste resulting from the treatment process. We learned of a similar problem sometimes caused by chlorination, the process used to disinfect drinking water. Chlorine may react with organic matter in the water to produce total trihalomethanes in excess of the MCL.According to the Oklahoma program manager, changing the primary disinfection agent from free chlorine to chlora- mines solves the trihalomethane problem, but produces a more persis- tent chlorine residual, which requires extra precautions from those using kidney dialysis machines and other users. When treating the existing water source is financially or technically infeasible, a water system may have to locate an alternative water source and pay high costs for drilling new wells or for connecting to neighboring systems. Here, too, the system may not be guaranteed that its drinking water problems will be resolved. In one of our case reviews, a water system contaminated with selenium attempted to resolve the problem by completing the development of a partially developed well site it owned. However, this tactic was not successful-the new well not only exceeded the selenium limit, but also violated the MCI23for arsenic and radiological contaminants. 1, Difficult Legal IssuesSometimes Another difficult issue arises when state regulators cannot identify a complicate Enforcement system owner against whom to take enforcement action. In some instances, the original owner is no longer present, and the system users are unwilling to take responsibility for the operation and maintenance of the facility or for meeting drinking water requirements. Compounding the problem is the fact that virtually all of the ownership disputes seem to occur at the very smallest systems, where financing needed improve- ments presents a major obstacle, even if the users do agree to take responsibility. In Washington, program officials told us that many of the ownership disputes occur at systems that were built by developers. The problems Page 47 GAO/RCED-90-127 Noncompllance Undermines EPA DrMdng Water Program Chapter 3 Enforcement Inadequate in Deterring and Correcting Noncompliance often begin when the developer sells the properties. In some cases, the home owners form an association to operate and maintain the water sys- tem. Over the years, however, these associations tend to break down, while at the same time, the water system deteriorates. Sometimes, users deny that there is a “water system” or that they are hooked up to it. Another problem is that associations of homeowners are not eligible for most of the grant and loan programs available to finance improvements. Our case reviews disclosed some unusual tactics employed by state pro- gram officials in trying to establish the ownership of systems and to achieve compliance. In Washington, for example, the lack of a responsi- ble party at one water system led the state to act to put the system into receivership. According to state officials, they did so because it was highly unlikely that the system owner would be willing or able to make the needed improvements. The water system was in poor condition and the owner was in jail on an unrelated matter. Although the state obtained a default judgment against the system, it is considered uncol- lectible because the owner’s assets are negligible. Moreover, the court has been unable to appoint a receiver because the fees a receiver could collect by operating the system are too small for the system to be profitable. counts on its enforcement program to deter violations of drinking Conclusions EPA water standards and other requirements and to return water systems to compliance when such violations do occur. Citing limited EPAand state resources, the agency has largely restricted potential targets for timely and appropriate enforcement action by states to the most serious viola- tors-significant noncompliers. We found, however, that of the 95 cases we reviewed, EPA'Scriteria for timely and appropriate enforcement were met as required in 24 cases. 6 While extenuating circumstances exist in a number of these cases, some involve records of chronic violation by water systems that are capable of returning to compliance. To deal with such problems, the EPAAdmin- istrator announced, in April 1989, that the agency will encourage states to increase their number of enforcement actions. Moreover, the Office of Drinking Water developed a number of initiatives for fiscal year 1990 to strengthen its enforcement program, including a model for escalating state and EPAenforcement actions in cases involving water systems with chronic violations. While we support these actions, we believe EPAneeds to make several fundamental changes in its approach toward enforce- ment for these efforts to have their intended effect. Page 48 GAO/RCED-99-127 Noncompliance Undermines EPA Drinking Water Program Chapter 3 Enforcement Inadequate in Deterring and Cwrecting Noncompliance First, EPAneeds to emphasize to states that actions taken need to be complete and effective. For example, the large majority of bilateral com- pliance agreements examined during our file reviews were either not signed by both parties, as required, or did not contain a compliance schedule with interim milestones and a final date when compliance will be achieved. These requirements help ensure that the agreements will be honored and that corrective actions will proceed as planned. In some cases, EPAcriteria specify that an enforcement action stronger than a compliance agreement should have been used because the system in question had a poor compliance history. Perhaps of greater concern, a number of enforcement actions that did meet the EPAcriteria had little or no effect on system compliance. We found this to be particularly true for civil referrals, which EPAcounts as appropriate regardless of whether they are filed in court. We believe that EPAis responsible for helping to ensure that the actions are not only appropriate according to its guidelines but also effective. Where states choose to enforce drinking water requirements through civil referrals, we believe EPAneeds to determine, as part of its annual program review with the state, whether the state attorney general’s office is willing and able to act on them. Many states rely on administrative orders as a more feasible enforce- ment action to take than civil referrals. However, some state programs have had limited success in using these actions because the programs do not have the authority to levy administrative penalties or because the orders involve cumbersome administrative processes. As part of its annual program review, EPAneeds to help ensure that states relying on administrative orders can implement them in a timely manner and implement them with the necessary “teeth” to be effective. Second, we believe that EPAneeds to broaden the universe of compliance problems for which it is willing to take follow-up enforcement action. At least some regional offices will only follow up on SNCcases when asked to do so by the state and will assume that a problem is being handled as long as the state has taken an appropriate action. Our review identified a number of such cases, however, that were making little or no progress toward resolution. At least one case presented a potentially serious health risk. Still, it is important to realize that improving state and EPAenforcement is not a complete solution to the program’s compliance problems. As revealed in our case reviews, some water systems clearly face problems Page 49 GAO/RCED-99-127 Noncompliance Undermines EPA Drinking Water Program . “1 chapter 3 Enforcement Inadequate in Deterring and tirrecthg Noncompliance that cannot be resolved simply through stronger EPAor state enforce- ment. Chief among these problems are the staggering costs some systems face in resolving serious contamination problems-particularly in the wake of new contaminant limits and other challenges posed by the strengthened requirements of the 1986 amendments to the Safe Drink- ing Water Act. These problems and EPA’Sefforts to help states and water systems deal with them are discussed in the following chapter. To help ensure that state and EPAenforcement actions meet program Recommendations requirements and are effective in encouraging water systems’ compli- ance with drinking water program requirements, we recommend that the Administrator do the following: . Direct the Office of Drinking Water to help ensure that state enforce- ment actions meet EPA’Sappropriateness criteria. Specifically, the Office should issue guidance to EPAregions emphasizing that bilateral compli- ance agreements must be signed and must include a compliance schedule with interim milestones, if applicable, and a final date when compliance will be achieved. The guidance should also reiterate that a bilateral com- pliance agreement may not be an appropriate action when the violating water system has had a poor compliance history. . Take steps increasing the prospect that appropriate state enforcement actions will return violating systems to compliance. Specifically, the Administrator should direct EPAregions to examine, as part of their annual program reviews, whether (1) states relying on civil referrals have the resources and commitment needed within the state drinking water program office and the attorney general’s office to ensure that such referrals will be acted upon and (2) states relying on administra- tive orders have a workable procedure to implement them in a timely b manner and have sufficient authority to assess penalties as part of the order. l Direct the Office of Drinking Water to revise its enforcement guidance to regions to encourage them to more actively consider EPAfollow-up enforcement action beyond cases referred to EPAby state authorities. This expanded universe of enforcement targets should include cases where state action may have been taken but was not effective in achiev- ing compliance. Page SO GAO/RCED-90-127 Noncompliance Undermines EPA Drinkhg Water Program New Drinking Wakr RequirementsWill Pose Additional Complianceand Enforcement Challenges As mentioned in chapter 1, the Safe Drinking Water Act, enacted in 1974, required EPAto promulgate and periodically revise national drink- ing water regulations for public water systems. The 1986 amendments to the act significantly enhanced EPAand state drinking water program responsibilities, requiring standards for 83 additional contaminants, stringent filtration and disinfection requirements, and increased moni- toring for regulated and unregulated contaminants. Although it is too early to determine total programmatic impacts of the 1986 amendments, we found that EPAofficials, state program managers, and representatives of industry and state associations expect the new requirements to have tremendous impacts on local, state, and federal drinking water programs. These individuals pointed out that small water systems will be particularly affected because they already lack the financial and technical resources necessary to implement the existing drinking water requirements. The same people also agreed that without additional resources, compliance and enforcement problems will increase dramatically. The 1986 amendments to the Safe Drinking Water Act include several Major Drinking Water new statutory mandates. Among other things, the amendments require Requirements EPAto Mandated by the 1986 set nonenforceable health goals, commonly called maximum contami- Amendments ’ nant level goals, and enforceable maximum contaminant levels or treat- ment techniques for 83 specific contaminants; . establish criteria by which states determine which surface water sys- tems must install filtration; . promulgate treatment technique regulations that will require all public water systems to use disinfection; . establish requirements for water systems to monitor for unregulated contaminants; . publish a list of contaminants, which are known to occur or anticipated to occur in public water systems and which may require regulation, and set, every 3 years, maximum contaminant level goals and MCLs for at least 25 contaminants on the list; and . develop corrosion control treatment requirements to minimize lead and copper deposits from plumbing materials such as lead pipes, solder, flux, or fittings. Page 61 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program , Chapter 4 New Drinking Water Requirements Will Pose Additional Compliance and Enforcement Challenges EPAhas thus far issued final or proposed regulations for most of the new EPA Has Proposed or drinking water requirements. Agency officials expect all of the new Promulgated drinking water regulations to be promulgated and in effect by mid- to Regulations for Most late 1994. New Drinking Water As of December 31, 1989, EPAhad issued final regulations for 15 of the Requirements 83 contaminants, the surface water filtration criteria (also called the surface water treatment rule), disinfection treatment of surface water systems, and the monitoring of 51 unregulated contaminants. The agency had also issued in final form, the list of contaminants that may require regulation (also called the drinking water priority list). EPAhad also proposed regulations for an additional 40 contaminants, treatment techniques for controlling lead and copper corrosion in plumbing materi- als, and monitoring requirements for an additional 100 unregulated contaminants. At the completion of our review, EPAofficials were developing regula- tions for the remaining 28 of the 83 contaminants, disinfection treat- ment of groundwater systems, and monitoring requirements for other unregulated contaminants. They were also developing regulations to control disinfection by-products, which can result from the disinfection of surface water and groundwater supplies. EPAplans to issue proposed regulations for these activities sometime in 1991 and final regulations in 1992. These regulations will significantly increase program responsibilities for ResourceConstraints water systems, states, and EPA.According to information obtained from Will Increase EPAand state officials, EPA'Spublished cost analyses, and the results of Compliance and studies conducted by representatives of water systems and states, the new requirements will cost millions of dollars to implement in coming a Enforcement Problems years. As discussed below, these new costs will place considerable finan- cial strain on many water systems and states, and will ultimately affect EPA'Sability to implement the program. Impacts on Water Systems Under the 1986 amendments, water systems must adhere to more strin- gent water treatment, monitoring, and reporting requirements. Accord- ing to EPAofficials, many systems will have to install new equipment or modernize their infrastructure (i.e., distribution, storage, treatment, lab- ” oratory, and monitoring facilities) to comply with some of the new stan- dards, particularly the new filtration requirement. Also, some systems will have to contract with certified laboratories to perform complicated Page 52 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 4 New Drinking Water Requirements Will Pose Additional Compliance and Enforcement Challenges analytical services and hire skilled operators to take water samples and operate their facilities. EPAofficials told us that these changes or improvements will be very expensive to many systems. Table 4.1 summarizes EPA'Slatest available estimates of the number of community and noncommunity water systems affected by the new drinking water requirements and the total annual costs for implement- ing these requirements. It shows that compliance by water systems is projected to cost about $2.5 billion annually.1 EPAofficials stated that the numbers of water systems affected by each requirement and the annual compliance costs will probably be higher than the agency’s esti- mates because the estimates (1) assume systems affected are in compli- ance with existing regulatory standards; (2) do not include costs for removing lead pipes and other plumbing fixtures, which is required under the lead and copper corrosion rule; and (3) do not include costs for regulating all 83 contaminants or costs for disinfecting surface water and groundwater supplies and controlling disinfection by-products. Table 4.1: Estimated Annual Costs to Water Systems for Implementing 1986 1986 Dollars_______- in millions Amendments to the Safe Drinking Water Annualized Average Act Number of annual Total annual systems %cti monitoring compliance -Rule .-- -~---_I .- affected cost9 cost cost Volatile organic chemicals 1,824 $32.7 -_____ $23.1__- $55.8 --___ ..-----_- Filtration 10,228 511.6 17.1 528.6 -__--..-. Total coliforms 200.183 0 75.2 75.2 Synthetic organic chemicals 2,284 45.4 32.2 77.5 .___ Inorganic --.--___._-chemicals 1,896 123.2 12.4 135.6 - Lead/coPoer corrosion control 43.927 302.2 32.9 335.2 ..--z-___-- 790.3 -____ Radionuclides-_~-___-- 22,867 -____ 2.6 792.9 Disinfection 103,354 474.8 12.8 --__-- ---487.7 b ____-- b Total $2.280.2 $208.3 $2.488.5 Note: According to an economist in EPA’s Office of Drinking Water, EPA presented its estimates in 1986 dollars to ensure consistency. Most estimates presented here result from regulatory and economic anal- yses EPA conducted in late 1986 and early 1987. Dollars are rounded to the nearest 100,000. “Figures in this column include the estimated annualized costs over 20 years at a lo-percent discount rate and 1 year of annual operation and maintenance (O&M) expenses. “The number of water systems affected can not be added together because some systems WIII be affected by multiple rules. ‘Figures presented for water suppliers were taken from EPA’s report Estimates of the Total Benefits and Total Costs Associated With Implementation of the 1986 Amendments to the Safe Drinking Water -Act (Mar. 16, 1989). Page 63 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 4 New Drinking Water Requirements Will Pose Additional Compliance and Enforcement Challenges EPAofficials project that nearly all of the nation’s 68,000 community water systems and the over 219 million customers they serve will be affected by one or more of the new regulations. The impacts on individ- ual systems and consumers will vary depending on the applicable regu- lations and treatment requirements. The new filtration requirement is expected to be particularly costly to many water systems and, therefore, can cause increases in consumers’ monthly water bills. For instance, EPA currently estimates that an average household receiving water from a small community water system that does not filter its water will see its monthly water bill increase by $30 to $50 when filtration costs are added. For the same size household served by a large system, EPAesti- mates that filtration will add $6 per monthm2 Largest Impacts on Small Water A 1988 survey by the League of Women Voters Education Fund” found SJYWIIS that many systems currently have serious financial problems that pre- vent adequate maintenance and treatment of their drinking water sup- plies. In most states, more than 90 percent of these troubled systems are classified as small systems. EPA'Scost estimates show that compliance with the new drinking water requirements will affect water systems of all sizes; however, small systems will have greater difficulties meeting the new challenges. EPAofficials explain that these small systems, which already account for more than 90 percent of current drinking water vio- lations, lack the financial and technical resources needed to manage a water system. The EPAofficials expect that the addition of new drinking water requirements, many of which will pose increased technical chal- lenges, will only exacerbate problems for small systems. EPAofficials also suggested to us that such difficulties will inevitably increase compliance problems among small water systems. The problems officials are expecting include water systems that exceed MCLstandards, a fail to install required filtration equipment, do not perform required monitoring, or fail to take measures to prevent lead or copper corrosion. “In contrast, EPA predicts that the monthly water bill for the same size household served by a small system that already filters its water will increase by an average of $2 to $6per month, and house- holds served by a large system that already filters its water will see their bills increase by about $1. EPA’s analysis assumes that the average household contains 2.8 people and that each person uses 100 gallons of water per day. %onducted between December 1987 and June 1988, the League of Women Voters’ survey was designed to identify the impact that complying with the 1986 amendments will have on water sys- tems and states. The respondents included 672 local water officials in 49 states and state drinking water administrators in all 60 states and the Virgin Islands. The survey sample was composed mostly of larger water systems. Page 54 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chaptar 4 New Drhkiug Water Requirements Will Pose Additional Compliance aud Enforcement Challenges In addition, our contacts with state and EPAprogram managers suggest that the new requirements may lead some financially pressed water sys- tems to find ways to avoid incurring additional expenditures. As noted in chapter 2, officials in all of the states we visited cited instances in which they detected or strongly suspected that water systems had falsi- fied compliance data to make it appear that the systems were complying with drinking water requirements. While these officials believe such practices are not widespread, the incentives to engage in them will increase when expensive new regulations such as the filtration require- ments take effect. Impacts on States Just as the 1986 amendments increased responsibilities for water sys- tems, they also increased responsibilities for state drinking water pro- grams. Among these new responsibilities is the authority to decide (1) the amount of monitoring water systems must conduct for regulated and unregulated contaminants, (2) which water systems must install filtra- tion, (3) the vulnerability of water systems to certain types of contami- nation, and (4) when to issue variances and exemptions to the new requirements. In 1988 EPAand the Association of State Drinking Water Administrators conducted a joint survey to obtain estimates of staff and funding resources states will need to implement the existing requirements and those established in the 1986 amendments.4 Twenty-one of the 36 responding states reported that their resources are inadequate to meet current program requirements. Thirty-three of the states said that they have had to limit their drinking water program activities over the past 6 years because of insufficient funding and/or staffing. According to the survey, it would cost states approximately $129 million annually to fully implement existing drinking water program require- ments.” Of this total, $32 million is currently provided through federal grants and $63 million through state funding. The remaining $34 million represents the funding shortfall states said they have trying to comply with existing program requirements. 4A detailed questionnaire was sent to all states and territories. The 36 states and one territory that responded regulate approximately 78 percent of all community water systems. “Because all U.S. states and territories did not respond to the survey, EPA and the Association ex- trapolated the results from the responding states and territory to obtain national estimates. Page 56 GAO/RCED-90-127 Noncompliance Underminea EPA DrMdng Water Progrnm Y ,:.I f ;. . ‘- 1 Chapter 4 New Drinking Water Requirements Will Pose AddItional Compliance and Enforcement Challenges As table 4.2 indicates, the survey estimates that states will need over $186 million between 1987 and 1992 for onetime start-up costs to begin implementing many of the new requirements. After 1992, they will need approximately $152 million per year to meet the new requirements. According to EPAdrinking water officials, states’ initial and annual costs will actually be much higher because states’ estimates, like those of the water systems discussed earlier, only include estimates for regulations EPAhas promulgated or proposed and, therefore, do not include the cost of regulating all 83 contaminants or complying with the new disinfection treatment and disinfection by-product requirements. Table 4.2: Estimated Initial and Annual Costs to States for Implementing 1988 1988 Dollars in millions Amendments to the Safe Drinklng Water -__ Initial implementation Total annual costs Act Rule costs (1987-l 992) (after 1992) Volatile organic chemicals $18 $17 Filtration 39 14 Total coliforms 18 18 Inorganic and synthetic organic - chemicals 21 17 Lead/copper -- __- corrosion control 47 38 Radionuclides 20 13 --- Others (sanitary surveys/inspections and data entry/reporting) 22 35 Total $185 $15i Note: Dollars are rounded to the nearest million According to the survey results, state officials expect the lead/copper corrosion control and filtration requirements to be the most costly and to absorb the largest portion of their financial resources during the ini- tial years. Other major program expenditures will include costs for con- 6 ducting sanitary surveys, identifying and classifying water systems requiring filtration, performing assessments of systems’ vulnerability to contamination, expanding laboratory capabilities, and taking formal enforcement actions. Also, state officials indicated that more resources will have to be dedicated to enforcement during the later years if water systems do not get additional resources to implement existing and new program requirements. Important State Activities Could Faced with resource shortages of this magnitude, some states may have Be Reduced y to shift their work priorities or further limit some program activities to implement the existing and new requirements. According to EPAofficials and the results of the study conducted by EPAand the Association of Page 66 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 4 New Drinking Water Requirements Will Pose Additional Compliance and Enforcement Challenges State Drinking Water Administrators, the activities most states identi- fied as “very likely to be limited” include staff training and develop- ment, special studies, and public participation and education. States also indicated they may limit their enforcement efforts, laboratory capabili- ties, and inspections/sanitary surveys because of scarce resources. Such forecasts are particularly disturbing in light of our findings in chapter 2 that more consistent use of such activities is central to any effort to improve compliance and better protect public health from contaminated drinking water. According to EPAheadquarters and regional officials, too, resource shortages will exacerbate the kind of enforcement problems we identi- fied in chapter 3. Many states will be forced to target a smaller percent- age of violators and will be less able to take appropriate enforcement actions against known violators in a timely manner. Impacts on EPA According to EPAofficials, the 1986 amendments had a tremendous impact on EPAimmediately after Congress passed the law, requiring the agency to prepare and implement regulations for 83 contaminants and other requirements. They said that the amendments’ future impacts depend on whether states and water systems get the additional resources needed for their drinking water programs. If states and sys- tems obtain sufficient resources, EPA'Srole will be limited to overall pro- gram management and oversight activities. If they do not, EPAwill have to get involved more directly in program activities such as developing state regulations and guidance, monitoring water systems’ compliance efforts, and taking appropriate enforcement actions when necessary. According to EPAdrinking water officials, a further complication, reported by some state officials, is that overwhelming program costs may lead their states to return primacy to EPA.Although no state has officially initiated such action, the EPAofficials indicated that this possi- bility is a growing concern within the agency-particularly with regard to states that are already having serious financial problems implement- ing existing drinking water requirements. The officials assert that if EPA has to carry out all or portions of the states’ program responsibilities because the states either return primacy or cannot implement their pro- grams fully, EPA,too, will face severe implementation problems. Page 67 GAO/RCED90-127 Noncompliance Undermines EPA Drinking Water Program Chapter 4 New Drinking Water Requirements WiJl Pose Additional Compliance and Enforcement Challenges Spurred by the agency’s growing concern with the resource shortfall EPA Efforts to Help (both within the drinking water program and agencywide), EPAcon- Water Systems and ducted two studies to identify alternative financing mechanisms states States Ok&in and water systems can use to generate additional funds for their pro- grams. The first study, conducted in 1988 by EPA’SOffice of Policy, Plan- Additional Resources ning, and Evaluation, identified states’ current use of such methods to support a variety of environmental programs.” The second study, initi- ated in May 1988 by EPA’SOffice of Water and completed in late 1989, identified financing alternatives that can be used to meet resource needs for complying with existing and new drinking water requirements. Both studies showed that states are increasingly using alternate financ- ing mechanisms-such as fees, taxes, bonds, fines, and penalties-to generate additional revenues for their environmental programs. The revenues derived from the options varied in amounts from state to state and program to program. Nevertheless, fees were found to be the most widely used option. Complementing these analyses is a more direct effort by EPA’S Office of Drinking Water to mobilize state and local governments, water systems, and private organizations to use creative approaches to find additional resources for state and local drinking water programs. Under its “Mobil- ization Strategy,” the Office plans, among other things, to . educate state decision makers (i.e., governors, legislators, public health and environmental officials, and budget officials) about the need for additional resources; l help operators of small systems understand the new drinking water requirements, provide training and technical assistance through a vari- ety of existing networks, and assist the systems in obtaining additional a resources from larger systems and private organizations; l identify readily available low-cost technological solutions for water sys- tems, particularly small systems, to use in order to comply fully with the new regulations; and l better inform the general public of health risks associated with contami- nated drinking water and the importance of maintaining safe drinking water supplies (to generate support for higher water rates). “The study included eleven states: Florida, Georgia, Iowa, Louisiana, Missouri, Nebraska, New Jersey, Ohio, Oregon, Pennsylvania, and West Virginia. The final report, entitled States Use Of Alternative Financing Mechanisms In Environmental Programs, was issued in June 1988. Page 59 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Chapt8r 4 Nem lhhking Water Requirements Wffl Pose Additional Compliance and Jhforcement Challenges Now that most of the new regulations have been either proposed or promulgated, EPAofficials said that they have begun shifting their efforts more toward the mobilization strategy. They added that EPAis strongly encouraging states and water systems to use alternative financ- ing mechanisms and the mobilization initiatives, respectively, to meet their resource needs. Representatives of water systems and states told us, however, that while they agreed that financing alternatives and EPA'Smobilization strategy will help to generate additional revenue for environmental pro- grams, they are skeptical about whether these activities will generate enough funds to meet future program needs. These representatives told us that even under the best of circumstances, alternative financing mechanisms and the mobilization initiatives will not generate enough revenue to cover all existing and new drinking water program expenditures. Nevertheless, while such activities may not offer a complete solution to the financial challenges facing the program, our fieldwork confirmed the usefulness of alternative financing in supporting vital program ele- ments. Texas, for example, which, as noted in chapter ‘2, has the most comprehensive sanitary survey program among the states we visited, pays for the surveys through an annual fee charged to water systems. The fee also covers other state services such as the collection and analy- sis of chemical samples and technical assistance. Fees range from $50 for water systems with 1 to 49 service connections to $5,000 for systems with 200,000 or more connections. According to the Texas state program manager, because systems pay for the surveys as part of the annual fee, they actually press the state to conduct them in a timely fashion to be sure they get their money’s worth. Alternative financing also benefits Texas’ enforcement program. As dis- cussed in chapter 3, Texas chooses to rely on resource-intensive civil referrals rather than administrative orders. The state attorney general’s office, staffed with 8 to 10 attorneys specifically dedicated to handle drinking water cases, has been highly successful in bringing civil actions in drinking water cases. Under a 1983 state law, the attorney general’s office is credited with a portion of the income it generates in fines and penalties to cover reasonable attorneys’ fees and court costs. In the Page 69 GAO/RCED-90-127 Noncompliance Undermines EPA D&king Water Program :., ..: ..*’ , : ,.’ Chapter 4 New Drinking Water Requirements Will Pose Additional Compliance and Enforcement Challenges drinking water program, penalty assessments have been substantial, totalling over $165,000 in fiscal year 1989 alones In contrast, states without such financing schemes have had to cut back on vital program elements. As noted in chapter 2, funding constraints have forced Washington to discontinue sanitary surveys. A 1988 con- sultant report also warned that the state’s drinking water program staff will need to double to carry out its responsibilities adequately after the regulations under the 1986 amendments take effect. In addition to rec- ommending increased support from the state general fund, the report recommended that program staff investigate “the development of user funding” to support various program activities. As required by the 1986 amendments to the Safe Drinking Water Act, Conclusions EPAhas issued or proposed many new regulations that will significantly affect EPA,states, and nearly all of the nation’s 58,000 community water systems. Although actual impacts of the new drinking water require- ments will not be known until all new regulations become effective, water systems (particularly small systems) are expected to incur enor- mous financial costs and face difficult new compliance challenges. Already facing huge gaps between program costs and available funding, states will see their own regulatory costs increase by hundreds of mil- lions of dollars annually. With water systems and states both experienc- ing increased difficulties implementing the program, EPAalso expects a correspondingly greater burden on its own resources. While we emphasize in chapter 3 that improved enforcement is needed to encourage better compliance among water systems, EPA'Salternative financing efforts reflect the reality that better enforcement alone will 8 not do the job. Additional resources will be needed to increase water testing, perform sanitary surveys, train operators, and perform a vari- ety of other activities needed to ensure the safety of the nation’s drink- ing water. While EPA'Salternative financing efforts are by no means a complete answer to the resource question, our own fieldwork suggests that these efforts offer some hope that vital program activities can be funded. 7Through April 1989, Texas had assessed$168,660 in penalties. According to program officials, the attorney general’s office historically has collected approximately 90 percent of the amounts assessed. Page 60 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Page 01 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program __i Ppe k$krse Health Effects AssociatedWith ’ ‘* Ingesting ContaminatedDrinking Water According to officials from EPA and the Centers for Disease Control (CDC),over 600 contaminants have been detected in the nation’s drinking water supplies. Some of the contaminants (such as fluoride, cadmium, and arsenic) occur naturally in drinking water, while others (such as alachlor and chlordane) are man-made. Generally man-made contami- nants get into drinking water supplies through improper waste disposal, broken or faulty plumbing fixtures, agricultural runoff, leaks from underground gasoline and petroleum storage tanks, and discharges from power plants or medical facilities. According to EPA and CDC officials, the type of health problems that can result from consuming contaminated drinking water depend on the con- taminant, level of contamination in the water, susceptibility of the per- son consuming the water, and length of time a person consumed the contaminated water. Overall, EPA officials classify known and/or sus- pected health problems into three broad categories: acute, chronic, and carcinogenic effects. Acute health effects result mainly from microbiological contamination. EPA officials told us that the effects usually appear from 1 hour to sev- eral days after ingestion. Gastroenteritis is the most common acute ill- ness associated with ingesting contaminated drinking water. Other acute effects include headaches, vomiting, mild stomach cramps, mild cases of diarrhea, fatigue, and nausea. These symptoms generally last only a few hours or days and, for the most part, disappear without professional medical treatment. Chronic health effects generally appear after longer incubation periods. EPA and CDC officials told us that these effects may not appear until years later and still not be attributed to contaminated drinking water. The most commonly known chronic health effects include hepatitis, and . damage to the liver, kidneys, heart, and other body organs/systems. The most dangerous potential health effects involve contaminants that cause carcinogenic effects. These effects are the most difficult to detect and attribute to contaminated drinking water. Most information avail- able on the chronic and carcinogenic health effects is based on the results of laboratory tests performed on animals. The following table lists 66 of the 83 contaminants EPA must regulate in accordance with the 1986 amendments to the Safe Drinking Water Act, and the known or suspected adverse health effects associated with Page 62 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program :‘,. .,, .,’ ~, Appendix 1 Adverse Health Effects Associated With Ingesting Contaminated Drinking Water them. At the time of our review, EPA had not identified the adverse health effects associated with the remaining 17 contaminants. Table 1.1: Adverse Health Effects A88OCiatad With Drinking Water Contaminant8 Contaminant Adverse health effects tiolatile organic chemicals _-.- -...--.- ._-- 1. Benzene Central nervous system effects, nausea, dizziness, and vomiting; also increases the risk of leukemia 2. Carbon Tetrachloride Liver, kidney, and lung effects; -.-- also is a possible cancer-causing agent 3. 1,2.Dichloroethane Heart, liver, kidney, and central nervous and circulatory system effects; also is a possible cancer-causing agent 4.. Para:Dichlorobenzene Kidney, liver, and central nervous, pulmonary, and circulatory system effects; also is a possible cancer-causing agent 5. l,l-Drchloroethylene -. .- Heart, liver, kidney, and central nervous system effects; also is a possible cancer-causing agent _ -..- ..--.~-.- 6. Tetradhloroethytene Liver, kidney, and central nervous system effects; also is a possible cancer-causing agent 7. 1,l ,l ,-Trichloroethane Liver, central nervous and circulatory system effects 8. Trichloroethylene Heart, liver, kidney, and central nervous system effects; also is a possible cancer-causing ._....-. .-. .- - -.-__- -._---. agent.._. ._..-- .-.---.-~-..~--_- 9.. Vr.nyl &&ride Kidney, liver, and central nervous, cardiovascular, and pulmonary system effects 10. CIS-1 ZDrchloroethylene Central nervous system, liver, and kidney effects .._I_-.. 11. Trans-1,2, Dichloroethylene . - -.-. -- .._- __---_ _._._. Central nervous system, liver, and kidney effects 12. Chlorobenzene “.. __~. .” .._.._---.---.-.-. .--... Central nervous system, _.-...-.._..--- --.- ~-.- -.-...-. ..._.~~. liver, and kidney effects Microbiological contaminants and turbidity 13. Total Coliforms Although not necessarily disease-producing organisms, coliforms can be indicators of other organisms that cause assorted gastroenteric infections, dysentery, hepatitis, typhoid fever, _ and cholera. 14. Grardia Lambliaa Gastrointestinal disorders such as diarrhea and abdominal cramps 15. Viruses,’ Gastroenteritis, diarrhea, ..-- meningitis, -.___ and paralysis 16. Turbrdity” Interference with disinfection 17. Standard Plate Count” (Heterotrophic Diarrhea, cramps, nausea, headaches, and fatigue ....bactena count) -.. _ ..- ._ .. _.-.. -- . ._-.._-..-~... ~- --._- ~. -..... - -... - --..-..... 18. Legionella*’ Diarrhea, cramps, nausea, headaches, and fatigue Inorganic chemical8 - 19. Asbksios Benign tumors and, possibly, cancer of the stomach and pancreas 20. Barium .- Gastrointestinal distress, hypertension, neuromuscular and cardiovascular system effects 21, cadmium Kidney and liver damage, gastrointestinal distress, anemia, hypertension, renal dysfunction, _ __ and bone damage 22. Chromium Respiratory disorders, internal hemorrhaae (bleeding), liver and kidney damage, nausea, vom’iting, and gastrointestinal effects - . -’ 23. Mercury Gastrointestinal distress; kidney failure; and central nervous system effects, such as hearing impairments and speech and mental disturbances 24. N&ate “Blue Baby Syndrome” (i.e., asphyxiation by altering the oxygen-carrying capacity of the blood system), neuromuscular~..~~.. effects, kidney and central nervous system effects (continued) Page 63 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program -\ -* Appendix I Advert Health Effects Associated With Ingesting Contaminated Drinking Water Contaminant Adverse health effects ---___ 25 Nitrite”. Same adverse health effects as for nitrate 26 Selenium Lesions of the heart, kidneys, and spleen; heart, liver, kidney, gastrointestinal and -. ~-~~__.-_ neurological effects 27 Arsenic Heoatic and kidnev damaae and nervous svstem effects 28 FkJOrlde Skeletal damage and dental ____--- fluorosis, i.e., brown staining or pitting of permanent teeth 29 Anttmony Nausea, vomiting, and abdominal cramps ~- 30. Nickel Gastrointestinal oroblems 31 Sulfate Diarrhea; gastroenteritis; and dehydration, particularly in infants 32 Lead Delays in neurological and physical development, brain damage, peripheral nerve dysfunction, increased blood pressure, low birth weight, and kidney and central nervous system effects 33. Copper Gastrointestinal disturbances; anemia; and liver, kidney, and renal damage 34 Cyanide Hyperventilation, vomiting, tremors, convulsions, and death. - Organic chemicals 35 Endnn Kidney and nervous system effects, convulsions, headaches, dizziness, sleeplessness, weakness, and/or loss of appetite 36 Alachlor Liver effects and tumors in lungs, stomach, and thyroids; also is a possible cancer-causing agent 37 Atrazlne Nervous system, liver, and heart effects -- 38 Chlordane Liver and central nervous svstem effects: also is a oossible cancer-causina aaent 39 Dibromochloropropane (DBCP) Kidney, liver, and antifertility effects; also is a possible cancer-causing agent 40 1,2-Dlchloropropane (DCP) Liver, lung, and kidney effects; also is a possible cancer-causing agent 41 2,4-D Liver and kidney effects, skeletal muscular changes, and muscular incoordination 42 EplchlorohydrW Kidney, central nervous system, lung, and liver effects; infertility; possibly cancer 43 Ethylene dlbromide Lung, liver, spleen, kidney, and central nervous system effects; also is a possible cancer- causing agent 44 Llndane c&tral nervous system, liver, and kidney effects 45 Methoxychlor Central nervous, liver, kidney, and reproductive system effects 46 Polychlorinated Biphenyls (PCBs) Liver and reproductive and nervous system effects; also is a possible cancer-causing .~agent 47 Pentachlorophenol Liver, kidney, and central nervous and reproductive system effects -- 48 Tolucne Speech, v&ion, andh~~blems; impaired memory; and kidney, lung, liver, and central nervous system effects - 6 49 Toxaphene Central nervous system, liver, and kidney effects; also is a possible-~ cancer-causing ---. agent ____-. 50 Acrylamlde,’ Peripheral nerve and muscular damage, tumors; also is a possible cancer-causing agent 51 Ethylbenzene” Liver, kidney, and nervous system effects/disorders 52 Heptachlor epoxlde” Central nervous system disturbances and altered liver functions; also is a possible cancer- causing agent 53 Carbofuran Drow&ess, c%%&~%xiety, vomiting, and nervous and reproductive system effects 54 Aldlcarb Sulfoxlde” N&ous system effects 55 Aldlcarb Sulfone” Nervous system eifkcts -.- 56 2,4,5-TP (Silvex) Stomach irritation, depression, and kidney and liver effects 57 Xylene Central nervous system and liver effects 58 Styrenc” il Central nervous system and liver effects; also is a possible cancer-causing agent , .. a\ (conrinuea) Page 64 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Appendix I Adveme Health Effecta Asao&t,d With Ingesting Contaminated Drinking Water Contaminant Adverse health effects ______-___ 59. “l.lll 2,3,7,&TCDD (Dioxin) I .I ._“__.._.. - ..-.._-.-..-.-- Liver damage ---- 60. Heptachlor” Central nervous system and liver effects; also is a possible cancer-causing agent 61. Aldicarb .-~..~__.-. Central nervous system effects Radionuclides --l-l.-“------.-. ~-- 62. Radium 226 and 228 -_---------~-- Bone cancer 63. ...--~Beta particle and photon Cancer of the lungs, bones, head cavities, and leukemia 64. _~--Uranium Kidney effects and cancer in skeletal tissues 65. Gross aloha particle activitv Cancer 66. Radon Lung cancer aTreatment techniques have been proposed in lieu of an MCL for this contaminant. bThe 1986 amendments allow EPA to substitute up to seven contaminants if regulation of the substi- tuted contaminants would achieve greater protection of public health. On January 13, 1988, this con- taminant was added as a substitute to the list of 83 contaminants. Y Page 66 GAO/RCED-99-127 Noncompliance Undermines EPA Drinking Water Program , Appe ndix II Scopeand Methodology for Review of Stat6’ “ Enforcement Cases As discussed in chapter 1, we reviewed a judgmental sample of 96 SNC violations at 75 community water systems located in six states. The objective of our review was to examine the states’ enforcement perform- ance. Accordingly, we designed our review to include all possible catego- ries of cases. Individual review cases within each category were selected randomly. The following sections provide more detailed information on the selection of specific cases for our review. Our results represent only the cases reviewed and not all significant noncompliance cases nationwide. Because a nationwide sample was not feasible given time and resource Six States Selected constraints, we visited six states to gain insights into drinking water program enforcement. We selected two states in each of three geographi- cally dispersed regions-Massachusetts and Vermont, in Region I; Oklahoma and Texas, in Region VI; and Oregon and Washington, in Region X-and planned to review at least 15 cases within each state. We chose 1987 as the base year for our review because we wanted to Universe for Our include cases that were initiated after EPA'Senforcement policy was Review Included All implemented in October 1986. In addition, we wanted SNCviolations that Types of SNC were as recent as possible without being so new that one would not expect the states to have responded. We decided that selecting SNCviola- Violations tions identified during 1987 would allow sufficient time for state and EPAenforcement actions to occur. Our universe of SNCviolations was compiled from September 30, 1987, SNClists that EPAgenerated from its data management system for all six states. These lists included SNCviolations involving (1) monitoring and a MCLviolations for turbidity and microbiological contaminants, (2) moni- toring violations for total trihalomethane contaminants, and (3) MCL vio- lations for chemical and radiological contaminants (including total trihalomethanes). The EPAdata were incomplete, however, because the agency’s data management system cannot identify SNCviolations result- ing from noncompliance with chemical and radiological monitoring requirements. (See chapter 2 for further discussion.) Thus, to include these SNCSin our universe, we relied upon data that the states supplied to EPA.EPAcompiled this data and determined that these SNCSwere iden- tified as of January 1987. Although this SNClist was developed 9 months earlier than the September 30, 1987, SNClists generated by EPA,using it was the most feasible way to ensure that our universe included all types of sNc violations. Page 00 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program (F . Appendix II Scope and Methodology for Review of State Rnfomment Cases We also wanted to ensure that our enforcement reviews included all Reporting Categories types of cases. Based on the reporting categories used in EPA'Stracking for Enforcement system, several types of cases were possible: Review . The state could have taken one of four enforcement actions deemed appropriate under EPA'Senforcement policy: bilateral compliance agree- ment (EGA),administrative order (AO), civil referral (CR), or criminal filing. . The water system could have returned to compliance (RTC).' . The case could be classified as “none of the previous” (or pending), denoting that the water system had neither returned to compliance nor been subject to an appropriate state enforcement action. The review cases selected are presented in Table II. 1 by reporting cate- gory and by state. For each reporting category, “U” designates the uni- verse of SNCS,and “R” represents the number of SNCSreviewed during our analysis. 'UnderEPA'sdefinition, a system may be categorized as returned to compliance if, in the case of the microbiological and turbidity contaminants, the system has no monitoring or MCL violations for a period of 6 consecutive months. For the chemical and radiological contaminants, a system may be categorized as returned to compliance if it has conducted the required monitoring and if the test results show that the water quality is within the MCL. Page 67 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Appendtx D[ I -, &ope and Methodology for Review of State Enforcement Cases Table 11.1:Review Cases by Reporting Category and by State Number of cases by reporting category State RTC BCA A0 Civil Pending Total URURURURUR UR MA 0 0 0 0 2 2 0 0 2 2 4 4 OK 8 3 3 3 1 1 1 1 26 7a 39 15 OR 215 2b 40 4a 15 5 0 0 16 5c 286 16 TX 20 4 0 0 1 Od 2 2 105 12a 128 18 VT 4 4 6 6 0 0 0 0 5 9 15 15 WA 206 2b 6 4 27 11 3 3 313 7a 555 27 Total 453 15 55 17 46 19 6 6 467 38 1.027 95 “We initially selected an additional case for review in this category, but eliminated one case because it did not qualify as a SNC violation as of the end of fiscal year 1987. ‘We initially selected two additional cases for review in this category, but eliminated two because they did not qualify as SNC violations as of the end of fiscal year 1987. “We rnrtially selected seven cases for review, but eliminated two because the water systems did not qualify as community water systems. “We selected thus case for review, but found that the administrative order was issued by EPA and not the state. We actually revrewed thts case as a pending case, and it is included in the total reviewed in that category. “We rnrtially selected an additional case for review, but eliminated one case because the water system did not qualify as a community water system. For each of the six states selected for review, we (1) obtained state follow-up reports2 applicable to the SNC lists in our universe, (2) recorded the status of each case as reported by the state, and (3) sorted this data base by reporting category. Since none of the six states reported any criminal actions during our sample period, we sorted the SNC violations into the other five categories and then selected our review cases. a Our plan was to select at least 15 cases in each of the six states we vis- SNC Violation ited. The 15 cases selected within a state were to include three randomly Selection Method selected cases from each of the five available reporting categories. How- ever, Massachusetts had a total of 4 SNC violations in the universe, and Vermont had 15; in ea.ch case, we reviewed all SNC violations for those states. In some instances, the other states had fewer than three cases ‘To determine whether states are taking timely and appropriate action, EPA periodically requires the states to submit status reports on each SNC violation, To the maximum extent possible, we recorded the status of each case using the state follow-up reports that were submitted when the period allowed for timely enforcement action had elapsed-8 or 14 months after the applicable SNC lists were issued, depending on the type of violation. EPA’s timely and appropriate enforcement policy is dis- cussed in greater detail in chapter 3. Page 68 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program . Appendix II Scope and Methodology for Review of State Enforcement Cases within a particular reporting category. Whenever this occurred, we selected additional cases from the remaining categories to bring the total cases reviewed within a state to 15, whenever possible. Our first choice in making these additions was to select cases for which the states reported enforcement actions, but when that option was not available, we added cases from the pending category. (In most states, pending cases represented the largest category.) We eliminated 11 cases from our review because, upon examining the case files, we learned that (1) the water system responsible for the viola- tion did not qualify as a community water system or (2) the case did not actually qualify as an SNC violation as of the end of fiscal year 1987. Whenever a state had more than three cases within an individual reporting category, we randomly selected our review cases. When a state had three or fewer cases within a particular reporting category, we reviewed all cases in that category. Some of the cases selected for our review involved water systems that had multiple SNCviolations. Whenever an SNCviolation was selected in one reporting category, we reviewed all other SNCviolations committed by the same community water system during the review period. If a sub- sequent random case selection involved another SNCviolation committed by a previously selected water system, a new selection was made. In other words, once a water system was selected, it could not be selected again. In total, we reviewed 95 SNCviolations involving 75 community water systems. Sixty-one of these systems had a single SNCviolation dur- ing the sample period, and 14 systems had multiple violations. Of these 14 systems, 11 had two violations, 1 had three violations, 1 had four violations, and 1 had five violations. For each of the selected water systems, we conducted a detailed file review and interviews with state program managers. As necessary, we also discussed the cases with EPA regional officials. The results of our case reviews and our analysis of state enforcement performance are presented in chapter 3. Page 69 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program II Appendix III / Summary . of EPA’s Efforts to IssueNew *’’ .’ ’ Drlnkln ’ g Water RegulationsRequiredby the 1986 Amendmentsto the SafeDrinking Wakr Act (As of December 31, 1989) Requirement Statutory deadline statur of EPA’8 efforts 1. Set maximum contaminant 6/19/87 for 9 contaminants 4/2/86-final regulation levels or treatment issued for one inoraanic techniques and chemical (fluoride)? Final monitoring/ reporting regulation became effective requirements for 83 1O/2/87. contaminants 7/8/87-final regulations issued for eight volatile organic chemicals (VOCs). Final regulations became effective 1/g/89. 6/19/88 for 40 contaminants 8/18/88-proposed regulations issued for two inorganic contaminants (IOCs): lead and copper. Final regulations scheduled for promuloation in fall of 1990. 5/22/89-proposed regulations issued for 36 contaminants: 30 synthetic organic chemicals (SOCs) and 8 IOCs. Final regulations scheduled for promulgation sometime in 1990. 6/19/89 for 34 contaminants 6/29/89-final regulations issued for total coliforms contaminants. Final regulations will become effective 12/31/90. 6/29/89-final regulations issued for remaining five microbiological contaminants. Final regulations to become effective 12/31/90. EPA plans to issue proposed regulations for the remaining 28 of the 83 contaminants by b September 1990. 2. Issue filtration criteria for 12/19/87 6/29/89-final filtration rule surface water systems issued. Final rule to become effective 12/31/90. 3. Issue first regulations for 12/19/87 7/8/87-final monitoring monitoring unregulated requirements issued for 51 contaminants unreaulated contaminants. 5/22/89-proposed monitoring requirements issued for 100 additional unregulated contaminants. 4. Publish first Drinking Water l/1/88 and every 3 years l/22/88-first priority list Priority List thereafter oublished. (continued) Page 70 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program . ** ! Appendix III Summaw of EPA’s Efforts to Issue New Drhldng Water Regulations Required by the 1996 Amendments to the Safe Drinking Water Act Requirement -~-- --...- .- Statutory deadline Status of EPA’s efforts 5. Develop lead/copper 6/l 6188 8/l B/88-proposed corrosion control treatment regulations issued for regulation treatment of lead/copper corrosion control. Final reoulations scheduled for prijmulgation in fall of 1990. 6. Issue disinfection 6/l 9189 6/29/89-final disinfection treatment regulations, regulations issued for surface including criteria for water systems. Final granting variances, for all regulations become effective public water systems 12/31/90. EPA plans to issue proposed regulations for disinfection treatment of groundwater systems and disinfection by-products sometime in 1991. Final regulations scheduled for promulgation in 1992. i. Promulgate regulations for l/1/91 and every 3 years Office of Drinkina Water staff at least 25 contaminants thereafter plan to issue pr;posed on the Drinking Water regulations for these Prioritv List contaminants in earlv 1990 CIFluorlde was 1 of 26 contaminants regulated before enactment of the 1986 amendments. EPA issued revised fluoride regulations in 1986 to comply with the new drinking water requirements. Page 71 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program Appendix IV Major Contributors to This Report Peter F. Guerrero, Associate Director, (202) 252-0600 Resources, Steven L. Elstein, Assistant Director Community, and Wanda T. Hawkins, Evaluator Judy K. Pagano, Operations Research Analyst Economic Development Division, Washington, DC. Ellen M. Cracker, Evaluator-in-Charge Boston Regiona1 Office Toy Har Chin, Evaluator Tracey G. Westbrook, Evaluator Kathy R. Alexander, Evaluator (0894337) Page 72 GAO/RCED-90-127 Noncompliance Undermines EPA Drinking Water Program 1 .. (III u I _ _“I ._.._ ,.._ .._.. _-. . ^-..“_ ___ ..__.,l.~ll. ^,_,,,.,,, _ .., _._I .......l_,._ “-._ _ ._.“-~. _.,._ .- .__.l__-“l-.-..l.l --.-- -..--.-___- --I----- -- I I ‘tlittd Stiltt’S First-(%tssl%hiI ‘; (it~trt~rirl Acx*oilflt irig Of’l’ict~ Post ag,rc XL Ftws hit1 U’;l~ilill~t 011, I).( ‘. 2054H (;A() I+rmit. No. (;I 00 Of’fkial Hrlsiritw9 i h~rr;ift y for- h-is at t’ I Jw $1~00
Drinking Water: Compliance Problems Undermine EPA Program as New Challenges Emerge
Published by the Government Accountability Office on 1990-06-08.
Below is a raw (and likely hideous) rendition of the original report. (PDF)