oversight

Hazardous Waste: EPA's Generation and Management Data Need Further Improvement

Published by the Government Accountability Office on 1990-02-09.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

..~.__._.-___..._...,
                  “.-___..__
                           ____._   -   _-.._   -...--   _..___   -_----l-   _..._   -_--------
                                                                                                  ---




Fc~h-mry l!)!bO
                                                                             HAZARDOUS WASTE
                                                                             EPA’s Generation and
                                                                             Management Data
                                                                             Need Further
                                                                             Improvement




GAO,/ l’HMT)-!)O-:3
United States
General Accounting Office
Washington, D.C. 20548

Program Evaluation    and
Methodology  Division

B-23693 1

February 9,199O

The Honorable Mike Synar
Chairman, Subcommittee on Environment,
  Energy, and Natural Resources
Committee on Government Operations
House of Representatives

Dear Mr. Chairman:

In a February 9, 1987, letter, you asked us to investigate the efforts of the Environmental
Protection Agency (EPA) to improve the quality of information on hazardous waste
generation and management capacity, and to determine how better information could be
developed. We first briefed members of your staff on EPA'Sprogress in revising the 1987
Resource Conservation and Recovery Act reporting system in March 1988. This report
describes the final work carried out in response to your request.

As we arranged with your office, unless you publicly announce the contents of this report
earlier, we plan no further distribution of it until 30 days from the date of the report. At that
time, copies of this report will be sent to interested congressional committees and the
Administrator of the Environmental Protection Agency and will be made available to others
upon request.

If you have any questions or would like additional information, please call me at (202) 275-
1854 or Dr. Michael Wargo, Director of Program Evaluation in Physical Systems Areas, at
(202) 2763092. Other major contributors to this report are listed in appendix II.

Sincerely yours,




Eleanor Chelimsky
Assistant Comptroller General
                                                                                                         -
&ecutive Summq


                   The Environmental Protection Agency (EPA) and state environmental
PGrpose            agencies share responsibilities for the national program of hazardous
                   waste management. To perform their missions, these agencies need
                   information about the production of hazardous waste, including minimi-
                   zation efforts, and the available means for safely treating, storing,
                   recycling, and disposing of the material. However, in the past, national-
                   level information about hazardous waste has suffered from serious defi-
                   ciencies. In response to this situation, Representative Synar, Chairman
                   of the Subcommittee on Environment, Energy, and Natural Resources of
                   the House Government Operations Committee, asked GAO to evaluate
                   EPA'Sefforts to acquire information about hazardous waste and to look
                   for ways to develop better information.


                   Several interrelated federal laws have established a comprehensive
Background         national program of hazardous waste management that is implemented
                   by a partnership between the states and the federal government. The
                   Resource Conservation and Recovery Act of 1976 and the Hazardous
                   and Solid Waste Amendments of 1984 form the foundation for this pro-
                   gram, but the Comprehensive Environmental Response, Compensation,
                   and Liability Act of 1980 (known as Superfund) and the Superfund
                   Amendments and Reauthorization Act of 1986 are also important parts.

                   Between 1976 and 1986, EPA developed an information system to pro-
                   vide information on hazardous waste that relied primarily on a joint fed-
                   eral-state data collection effort, known as the biennial reporting system,
                   and on national-level sample surveys conducted by EPA. Recognizing that
                   the information obtained from these efforts was inconsistent across
                   states, flawed, and incomplete, EPA took steps to improve the quality of
                   its data. GAOevaluated EPA'Songoing efforts in order to determine
                   whether more valid and complete national information is likely to result
                   in the future. GAO also made its own determination concerning the need
                   for various kinds of information and about how best to obtain the data.


                   EPA'Sefforts have not been successful enough to ensure the achievement
Results in Brief   of important national objectives. Important information gaps remain,
                   problematic measurement and data collection procedures will limit the
                   quality of some of the information that is produced, and the internal
                   process for developing information systems does not fully assure com-
                   plete and integrated data collection. The biennial reporting system still
                   does not ensure that the states will collect or report to EPA all of the
                   necessary data in a standardized way. These remaining problems will


                   Page 2          GAO/PEMD-90-9   EPA’s Hazardous   Waste Data Need Further   Improvement
        .


                       Executive   Summary




                       significantly impair the state capacity assurances required by the 1986
                       Superfund amendments. Data from different sources and years will be
                       used by different states, and states with missing data will have to use
                       questionable assumptions provided by EPA.



Prir@pal Findings
    I

Infortmation System    EPA has implemented information system development practices that are
Devcjlopmen.t          generally consistent with existing federal guidelines. However, some
                       refinement is needed to ensure that the several data collection mecha-
                       nisms used to collect hazardous waste information are fully integrated
                       so that the data collected by each mechanism is consistent with, and can
                       be used to supplement, data collected by the others. Specifically, clearer
                       administrative responsibility for the coordination of data collection
                       efforts is needed, and each stage in EPA'Snew life cycle management sys-
                       tem needs to utilize more thorough assessments and have more complete
                       documentation of the work that was done for major system components.


Need for Information   EPA has identified most of the needed categories of information, but
                       there are three important exceptions. First, there is no provision for
                       obtaining information (required for developing regulations) about the
                       quantity and types of waste at Superfund and other similar sites that
                       will ultimately require hazardous waste management. Second, no provi-
                       sion has been made for obtaining information on the quantity and types
                       of some additional wastes that will ultimately require management,
                       including the large volumes expected to result from the cleanup of leak-
                       ing underground storage tanks. Finally, no provision has been made for
                       obtaining information on the disposal capacity of salt domes or other
                       geological formations that may be capable of preventing the migration
                       of hazardous wastes.


Measurement Problems   EPAhas improved the measurement instruments that it uses to obtain
                       information about hazardous waste. For example, the problems in mea-
                       suring the total amount of waste generated and in classifying the types
                       of storage, disposal, and recycling technologies appear to have been suc-
              v        cessfully addressed. However, ill-defined categories of waste, imprecise
                       measures, and weakly constructed questionnaire items indicate that not
                       all measurement problems have been resolved. Indeed, the remaining


                       Page 3                GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                   Executiw   Summary




-----I



                   problems may produce significant errors in measuring the amounts of
                   different types of waste generated, the capacities of various available
                   treatment technologies, and the amount of waste eliminated by minimi-
                   zation efforts. GAOidentified a four-class framework of treatment tech-
                   nologies-physical,    chemical, biological, and thermal-that  shows
                   potential for further development. EPA could use this framework to
                   develop a classification system of treatment technologies with mutually
                   exclusive, exhaustive, and hierarchical categories. The development of
                   such a system, combined with quantitative measures of waste character-
                   istics, could help EPA resolve the remaining measurement problems.


D;kta Collection   EPA has also improved its data collection procedures. National surveys
         I         conducted directly by the agency use acceptable sampling techniques
         I         and uniform measurement instruments. However, the data collected for
         1          1987 by the biennial reporting system will not yield complete and valid
                   national-level information because of the continued use among the
                   states of different data collection instruments and systems. In addition,
                   the toxic chemical release inventory reporting system (required by the
                   1986 Superfund amendments) has not been designed to complement
                   other hazardous waste data collection efforts, which means that the
                   data cannot be used to address environmental problems within the
                   Resource Conservation and Recovery Act program. Five other factors
                   are likely to adversely affect future data collection and thus jeopardize
                   information quality. First, EPA has not provided sufficient funding for
                   states to collect and verify the data. EPA provides only 25 percent of the
                   cost of the biennial reporting system rather than the 75 percent it gener-
                   ally supplies, in accordance with EPA guidelines, for other required activ-
                   ities. Second, EPA is not planning to conduct future national surveys
                   using probability sampling, even though these have been its primary
                   source of detailed national information: the biennial reporting system
                   has not produced usable national information. Instead, EPA will rely pri-
                   marily on the biennial reporting system conducted by the states. In GAO'S
                   view, combining national surveys conducted by EPA with a streamlined
                   and standardized biennial census conducted by EPA or the states would
                   be the most efficient approach. Third, federal recordkeeping and report-
                   ing regulations do not require hazardous waste handlers to provide the
                   detailed data EPA requires and, fourth, they do not require states to use
                   a specific data collection instrument to collect all necessary specific data
                   elements, or to submit the data to EPA in a disaggregated form. Finally,
                   EPA has limited authority under the Resource Conservation and Recov-
                   ery Act to require states to collect standard data. The result of the cur-
                   rent arrangements is that the federal information system must be pieced


                   Page 4               GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                      together from separate state systems rather than, as suggested by the
                      act, having the states add data to a minimum, consistent federal system.

                  A
                       In light of these findings, GAOmakes several recommendations dealing
Recjmmendations        with the internal process for developing information systems and their
   ~                   components, filling remaining information gaps, and developing mea-
   ,                   surement instruments-in      chapters 23, and 4, respectively. To correct
                       the remaining data collection problems, GAOrecommends in chapter 5
                      that steps be taken (1) to ensure that the toxic chemical release inven-
                      tory can be used to supplement other hazardous waste data collection
                      efforts, (2) to provide a level of federal support for state data collection
                      in the biennial reporting system that assures valid and complete
                      national data, (3) to modify federal regulations governing recordkeeping
                       and reporting by individual handlers and state programs to ensure com-
                      plete data, and (4) to use probability sampling more effectively in con-
                      junction with the biennial reporting system.


                      In addition to the improvements EPA can make, GAObelieves a refine-
Matters for           merit in legislation may also be necessary to improve the quality of EPA'S
Congressional         information. Under current law, responsibility for data collection, as
Consideration         well as for other regulatory activities, is shared by federal and state
                      governments, The nonuniform data and procedures across the states,
                      which are associated with a joint federal-state data collection effort,
                      diminish the quality of national hazardous waste information. This
                      problem could be corrected by separating the recordkeeping and report-
                      ing provisions of the act from other regulatory provisions and making
                      EPA solely responsible for collecting the information required for devel-
                      oping and implementing the federal program. Uniform national data
                      would then be ensured, but states would retain the authority to add data
                      elements and to use supplemental data collection mechanisms to support
                      their needs.


                      GAOdiscussed its findings with EPA officials and has included their com-
Agency Comments       ments where appropriate. However, in agreement with the requester,
                      GAOdid not obtain official comments on a draft of this report. EPAoffi-
                      cials have stated that they generally agree with our findings and noted
                      that they have already taken steps that will at least partially address
                      some of them. Since these actions were taken after we finished our field
                      work, we could not evaluate them for this report. However, they are
                      listed at the end of relevant chapters.


                      Page 5           GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
Contents


E$ecutive Summary                                                                                                2

Chapter 1                                                                                                       10
Introduction           The Nation’s Comprehensive Hazardous Waste                                               10
                            Management Program
                       The Problem of Data Quality                                                              11
                       Objectives, Scope, and Methodology                                                       12
                       Strengths and Limitations                                                                17
  I                    Report Organization                                                                      18

Cljapter 2                                                                                                      19
Initial Problems and   Problems With EPA’s Original Information System                                          19
                       EPA Actions to Improve the Information System                                            26
EPA Improvement        Evaluation of Efforts to Improve the System                                              31
Ef!forts                   Development Process
                       Summary and Conclusions                                                                  38
                       Recommendations                                                                          39
  I

Chapter 3                                                                                                       41
EPA In formation       Identification of Activities                                                             41
                       Information Requirements for Implementation,                                             43
Needs                       Administration, and Oversight Activities
                       Information Requirements for Regulatory Development                                      47
                            Activities
                       Comparison of Information Needs With Data Collection                                     60
                            Instruments
                       Summary and Conclusions                                                              65
                       Recommendations                                                                      65

Chapter 4                                                                                                   66
Assessment of          Measuring the Type of Hazardous Waste                                                69
                       Measuring the Type of Treatment Technology                                           73
Measurement            Measuring Management Capacity                                                        75
Instruments            Measuring Waste Minimization                                                         76
                       Implications for Capacity Analysis                                                   78
                       Summary and Conclusions                                                              79
                       Recommendations                                                                      80
           Y




                       Page 6         GAO/PEMD-90-J   EPA’s Hazardous   Waste Data Need Further   Improvement
                      Contenta




Chipter 5                                                                                                82
            of Data   Data Collection in the Interim Information System                                  82
                      Implications for Capacity Assessments                                              91
    ection Methods    Data Collection in the Permanent Information System                                93
andI Procedures       Summary and Conclusions                                                           101
                      Recommendations                                                                   103
                      Matters for Congressional Consideration

Appendixes            Appendix I: Panel of State Experts                                                106
                      Appendix II: Major Contributors to This Report                                    107

Globsary                                                                                                108
   ,
Bibliography                                                                                            113

Tables                Table 2.1: Problems With the RCRA Reporting System                                 23
                      Table 2.2: Problems With Special Surveys, Permit                                   24
                          Applications, and Manifests
                      Table 2.3: Problems With Individual Data Collection                                25
                          Mechanisms and the Original Information System as
                          a Whole
                      Table 2.4: Problem Areas Addressed by Each EPA                                     27
                          Improvement Effort
                      Table 2.5: Specific Activities to Improve the EPA System                           28
                          Development Process
                      Table 2.6: EPA/NGA Improvement Activities                                          29
                      Table 2.7: EPA Improvements Developed Internally                                   30
                      Table 2.8: Interim and Permanent Status of New and                                 30
                          Revised Data Collection Mechanisms
                      Table 3.1: Program Activities That Require Generation or                           42
                          Management Information, by Function and
                          Responsible Organization
                      Table 3.2: Types of Information Required for                                       46
                          Implementation Activities
                      Table 3.3: Types of Information Required for Capacity                              54
                          Analyses and Waste Minimization Policy Assessment
                      Table 3.4: Types of Information Required for Risk                                  58
                          Assessment and Remaining Regulatory Development
                          Analysis




                      Page 7         GAO/PEMD-@O-3 EPA’s Hazardous   Waste Data Need Further   Improvement
         Contents




         Table 3.6: Comparison of Information Needs With the                                    61
             Information Gathered by the New and Revised Data
             Collection Instruments, Using the 1986 RCRA
             Reporting Instrument as a Baseline.
         Table 4.1: Required Information Types for Which                                        67
             Measurement Problems Exist, Using the 1985 RCRA
             Reporting Instrument as a Baseline.

Figure   Figure 4.l:Basic EPA Measure of Waste Type From the
              National Survey of Management Facilities




         Abbreviations

         CERCLA Comprehensive Environmental         Response, Compensation, and
                       Liability Act, 1980 (also known as Superfund)
         EPA        Environmental Protection Agency
         GAO        General Accounting Office
         HSWA       Hazardous and Solid Waste Amendments, 1984
         NGA        National Governors’ Association
         RCRA       Resource Conservation and Recovery Act, 1976
         SARA       Superfund Amendments and Reauthorization Act, 1986


         Page 8           GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
Y




    Page 9   G A O /P E M D - 9 0 - 3 E P A ’s H a z a r d o u s W a s t e Data N e e d Further Im p r o v e m e n t
Chapter 1

Introduction


                        Valid national information on hazardous waste generation and manage-
                        ment is essential for EPA and state agencies if they are to properly
                        develop, implement, and evaluate the hazardous waste management
                        program mandated by the Congress.’ However, in the past, EPA was not
                        able to develop and produce all the necessary information. As long as
                        they lack this information, EPA, the Congress, and the public will remain
                        uncertain about whether laws can be implemented effectively, whether
                        progress is being made toward waste minimization, or whether gener-
                        ated wastes are being managed safely and disposed of securely.

                        On February 9,1987, Representative Synar, Chairman of the Subcom-
                        mittee on Environment, Energy, and Natural Resources, House Govern-
                        ment Operations Committee, asked us to evaluate EPA efforts to improve
                        information quality and then determine how EPA can better develop
                        valid information. This report presents the results of our evaluation of
                        the extent to which EPAefforts have improved the agency’s ability to
                        produce the necessary hazardous waste generation and management
                        information. It also presents recommendations for further
                        improvements.


                        Complex and interrelated provisions of several federal laws are the
The Nation’s            basis of a comprehensive national program for managing the threat of
Comprehensive           hazardous waste. The national program is administered through a part-
Ha$ardous Waste         nership between the states and the federal government. The federal
                        responsibilities are administered primarily by EPA'SOffice of Solid Waste
Mahagement   Progrann   and Emergency Response.

                        Subtitle C of the Resource Conservation and Recovery Act of 1976
                        (RCRA)provides the basic structure for managing hazardous waste in the
                        United States. It prescribes activities to reduce the threat of hazardous
                        waste from generation to final disposition. The Hazardous and Solid
                        Waste Amendments of 1984 (HSWA)strengthened RCRAby further
                        encouraging waste minimization, discouraging land disposal, and requir-
                        ing the regulation of underground storage tanks.

                        The Comprehensive Environmental Response, Compensation, and Liabil-
                        ity Act (commonly referred to as CERCLAor Superfund) is also an impor-
                        tant part of the nation’s comprehensive program of hazardous waste
                        management. It requires EPAto create a National Priorities List and

                        ‘We use the term hazardous waste management throughout this report to refer to the numerous
                        technologies used for hazardous waste minimization, treatment, storage, disposal, andrecycling.



                        Page 10               GAO/PEMD-90-3     EPA’s Hazardous    Waste Data Need Further    Improvement
   i


                      Chapter 1
                      Introduction




                      establishes a fund to clean up spills and uncontrolled waste sites that
                      have been identified as priority problems. The Superfund Amendments
                      and Reauthorization Act of 1986 (SARA)strengthened CERCLAby further
                      encouraging permanent cleanups, requiring the application of relevant
                      standards developed under other federal laws (such as the Safe Drink-
                      ing Water Act), and requiring that cleanups meet relevant state stan-
                      dards, SARAalso requires each state to assure “adequate capacity for the
                      destruction, treatment, or secure disposition of all hazardous wastes
                      [including CERCLAand RCFUdefined wastes] that are reasonably expected
                      to be generated within the state during the 20-year period following. . .
                      [the assurance].“2

                      The enactment of SAM indicates a strong and growing connection
                      between RCRA,Superfund, and other environmental, health, and safety
                      legislation for managing hazardous waste. Its provisions emphasize the
                      need to assure integrated and consistent protection across environmen-
                      tal programs and media (soil, water, and air). EPAhas responded to this
                      need by establishing a cross-media initiative designed to integrate envi-
                      ronmental programs across media to ensure consistent protection. Con-
                      sequently, it is important that the information collected by EPA support
                      the interrelated objectives of these statutes.


                      EPAhas not been able to develop valid information about hazardous
The Problem of Data   wastes to cover all the functions that the agency must carry out. By
Quality               1986, EPA had developed a complex information system aimed at sup-
                      porting the requirements of the laws described in the preceeding section
                      of this report. The system included several data collection mechanisms,
                      each of which was designed to collect information deemed necessary on
                      specific attributes of hazardous waste generation or management. For
                      example, EPAestablished the RCRAreporting system (referred to by EPA
                      as the biennial report), a mandatory biennial census of large quantity
                      hazardous waste generators and management facilities, using the
                      authority provided primarily by sections 3002 and 3004 of RCRA.EPA
                      also conducted special sample surveys using the authority provided pri-
                      marily by section 3007 of RCRA.These include both national surveys and
                      smaller limited surveys for specific purposes. The Congress’s concerns
                      over the issue of data quality stem essentially from the inability of this
                      information system to provide valid national data.


                      %uperfund Amendments and Reauthorization Act of 1986, Public Law 99-499, sec. 104 (K), 100
                      STAT. 1613,162l.



                      Page 11             GAO/PEMDBO-3
                                                    EPA'sHazardous           Waste Data Need Further   Improvement




                                                                    ,
                         Chapter J
                         Introducti 7’~~
                                      j




                         Since 1986, EPA has taken steps aimed at improving the quality of haz-
                         ardous waste generation and management information. These actions
                         have produced three general changes: (1) implementation of new proce-
                         dures for assuring that new or revised data collection mechanisms pro-
                         vide valid information, (2) implementation of an interim information
                         system, and (3) plans for a permanent information system. The interim
                         system relies on one-time-only national surveys of hazardous waste gen-
                         erators and management facilities, in addition to a partially revised RCRA
                         reporting system. The permanent system will rely primarily on the fully
                         revised RCRA reporting system and will also be able to utilize the toxic
                         chemical release inventory reporting system required by SARA.It will not
                         include major national surveys.

                         One important concern of the Congress is whether sufficient capacity
                         exists to manage the hazardous waste that will be generated in the fore-
                         seeable future. There are three crucial types of capacity analysis. The
                         first type (required by RCRA) is designed to examine whether sufficient
                         capacity currently exists to implement the land disposal restrictions
                         required by HSWA.~The second, an internal EPA initiative used for regula-
                         tory development and termed an integrated capacity analysis, is
                         designed to assess the effects of all planned regulatory actions on the
                         supply of management capacity and on all sources of demand for man-
                         agement capacity. The final capacity analysis (required by SARA) pro-
                         vides the basis for each state to assure that sufficient capacity exists to
                         manage the hazardous waste (including RCRA and CERCLA wastes) that
                         will be generated in the state for the next 20 years. Unless EPA'S hazard-
                         ous waste information system can provide valid data, it will be impossi-
                         ble to develop sound capacity assessments or to support other
                         regulatory development activities.



Objectives, Scope, and
Methodology

Objectives               The Subcommittee on Energy, Environment, and Natural Resources of
                         the House Government Operations Committee asked us to evaluate EPA
                         efforts to improve information quality and to determine how EPA can

                         “The 1984 amendments (HSWA) of RCRA banned the land disposal of hazardous waste that had not
                         been treated in accordance with standards developed by EPA to ensure that it would not contaminate
                         water supplies. HSWA requires EPA to determine that sufficient capacity of the relevant treatment
                         technologies exists to implement the restrictions or postpone the requirements for up to two years.



                         Page 12               GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further    Improvement
    .


        Chapter 1
        Introduction




        better develop valid data. We translated the request into the following
        specific evaluation questions.

        1, What problems have been identified with EPA'Soriginal hazardous
        waste management information system?

        2. What steps have been taken by EPAto improve the quality of its haz-
        ardous waste generation and management information system?

        3. Is the process EPA initiated to effect improvements consistent with
        generally accepted standards for developing information systems?

        4. Has EPAidentified the information on hazardous waste generation and
        management that is required by EPA and the states in order to support
        the program mission?

        5. How well do EPA'Smeasurement instruments actually measure the rel-
        evant attributes?

        6. Are EPA'Srevised data collection methods and procedures likely to
        result in valid national information?


Scope   Information on hazardous waste generation and management is essential
        to determine whether sufficient management capacity exists to handle
        the hazardous wastes that are being produced now and that will be pro-
        duced in the foreseeable future. However, our evaluation is not limited
        to the information needed for capacity analysis because information on
        hazardous waste generation and management is also used by EPA and
        state programs for other important activities such as enforcement. Dif-
        ferent uses of generation and management information may require dif-
        ferent information or different levels of specificity in the same
        information. For example, detailed data on the concentration of hazard-
        ous constituents in a quantity of waste may be necessary for some pur-
        poses, whereas only information on whether these constituents are
        present may be necessary for others. Since EPAuses specific generation
        and management information for multiple purposes, we could not fully
        evaluate EPA'Sneed for generation and management information without
        broadening our scope to include consideration of all uses of this type of
        information.

        Furthermore, we recognized that EPApossesses several mechanisms for
        collecting information and that states have additional mechanisms in


        Page 13         GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
              Chapter 1
              IntroducQ0 “’




              place. Specific information used for a single purpose is frequently
              obtained from multiple data collection mechanisms. Reviewing the qual-
              ity of EPA'S information on hazardous waste generation and management
              inevitably involved an examination of these mechanisms and how they
              functioned together to provide the needed information. Consequently,
              the scope of our study includes the entire system of data collection
              mechanisms and the several uses of the data by EPA and state agencies.
              Accordingly, we define valid information as information that is suffi-
              ciently accurate, detailed, and relevant for its various intended uses.

              In sum, the scope of our evaluation involves the portion of EPA'S hazard-
              ous waste information system that provides technical data on hazardous
              waste generation and management for multiple uses, including the over-
              laps between EPA and state needs, We examined the problems EPA expe-
              rienced with its original information system until 1986 and evaluated
              EPA'Songoing efforts to correct them that were carried out from 1986 to
              August 1988. We could not evaluate more recent EPA actions. We did not
              examine computerized data management systems that only store admin-
              istrative information, including the resource conservation and recovery
              information system and the comprehensive environmental response,
              compensation, and liability information system.


Methodology   To answer our evaluation questions, we applied generally accepted stan-
              dards of information system design, as well as accepted conventions for
              measurement and data collection. Taken together, these principles con-
              stitute the yardstick we used to evaluate EPA'Sefforts. We believe that
              adherence to the principles is likely to lead to high quality information
              while lapses will probably have an adverse effect. Reliance on standards
              and conventions was appropriate in this study because the data col-
              lected by the new and revised data collection mechanisms developed by
              EPA were not ready for us to examine and seek to validate during the
              time frame of our field work.

              To answer our first evaluation question concerning the problems that
              existed in the original information system (prior to interim system
              development), we reviewed the existing literature and interviewed EPA,
              state, and other experts who were familiar with the information system.
              We identified problems in the areas of systems development, informa-
              tion needs, measurement, and data collection. We also identified known
              gaps in the data that were collected. To identify the steps EPA has taken




              Page 14         GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
Chapter1
Introduction




                                                                          -
to improve the hazardous waste information system, our second evalua-
tion question, we interviewed EPAofficials and reviewed EPA
documentation.

We evaluated EPA'Ssystem development efforts (question 3) by applying
existing, relevant federal standards for information systems develop-
ment as normative criteria. These standards are contained in the
Paperwork Reduction Act and the Federal Managers Financial Integrity
Act, as well as in guidelines for implementing these laws developed by
Presidential Councils, the Office of Management and Budget, the Gen-
eral Services Administration, and GAO.~

Determining whether EPA has identified the data needed to support the
agency’s hazardous waste mission in both the interim and permanent
systems (question 4) required three steps. We first conducted a detailed
examination of EPAprogram activities that are designed to achieve the
program’s mission, We conducted a series of semi-structured interviews
with relevant division directors, each branch chief (or designated repre-
sentative), and many section chiefs and project managers in the Office
of Solid Waste and Emergency Response. Each branch prepared a list of
the activities performed in each of its sections. During the interviews,
we determined which activities used data on hazardous waste genera-
tion or management. We also identified the data that are required to per-
form the activity and any problems EPA personnel were experiencing
with available information, Finally, we obtained and reviewed available
samples of the products generated by the activities, using these data to
further specify needed information. Because the scope of our project
was broad, as previously discussed, we had to limit our data collection
to one round of in-depth interviews. This step defined EPA’Sinformation
requirements.

Second, to identify state data needs, we interviewed program officials in
a judgment sample of both large and small states using semi-structured
interview techniques. In addition, we attended meetings of the National
Governors’ Association (NGA) advisory panel devoted to the redesign of
the RCRAreporting system. We also conducted a two-day workshop with
selected state program experts from both large and small states to help
identify state activities and data needs. (See appendix 1.) Each partici-
pant identified the activities for which data are required and any prob-
lems the states have experienced with available information. The


4Seebibliographyfor detailedcitations.


Page   15           GAO/PEMD-90-3
                               EPA’sHazardousWasteData NeedFurther Improvement
Chapter 1
IntroductirHv




participants also provided documentation on the structure and organiza-
tion of their programs, examples of data collection instruments that dif-
fer from those of EPA, and examples of the products produced by the
activities that use these data. The participants also evaluated existing
EPA data sets, which varied in breadth and detail of coverage, to deter-
mine the extent to which the data sets would be sufficient for the needs
of their states. Finally, we compared the results of our analysis of EPA
and state data needs to the revised data collection instruments EPA
developed, to determine whether the agency has identified the required
information.

Turning to measurement and data collection, we evaluated the extent to
which EPA initiatives are likely to improve data quality. We also evalu-
ated the new or revised data collection instruments EPA developed by
applying generally accepted conventions of measurement as normative
criteria to determine whether the measures are likely to result in valid
and reliable data (question 5). The measurement conventions we applied
can be found in numerous sources; three of these sources (Measurement
in the Social Sciences: the Link Between Theory and Data, Measurement
in the Social Sciences: Theories and Strategies, and Measurement Theory
for the Behavioral Sciences) are listed in the bibliography.

Next, we evaluated the data collection methods and procedures that
have been or will be employed in the interim and permanent information
systems (question 6). In this analysis, we applied generally accepted
conventions of data collection to determine the extent to which valid
data are likely to result. We also examined whether the different data
collection mechanisms are integrated so that they function together to
provide valid data and whether data collection is fully supported by
federal regulations.

We also examined the likely impact of remaining problems on each of
the three types of capacity analyses discussed previously. We assessed
the extent to which the necessary data will be supplied and the extent to
which measurement and data collection problems will affect the quality
of the capacity assessments.

The answers to questions 2 through 6 are conclusions about the likeli-
hood that valid information will result from EPA improvement efforts.
By comparing these conclusions to the problems identified in the original
information system, we determined the extent to which improved data
quality is likely. Identifying continuing and new problems in the interim
and permanent information systems allowed us to identify areas where


Page 16         GAO/PEMD-99-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                Chapter1
                Introduction




    I           additional improvements are needed. We determined how EPA could
                employ generally accepted standards in these areas to further improve
    I           data quality.


                Our evaluation has two major strengths. The first is the systematic
Strengths and   application of principles about information system design, measure-
Limitations     ment, and data collection. The second is the involvement of state pro-
                gram officials, as well as EPA officials. By involving state and EPA
                officials, it is possible to incorporate user participation and secure a
                greater degree of understanding about data needs and appropriate
                mechanisms for data collection.

                Our evaluation also has two limitations. First, although it is often desir-
                able to conduct several rounds of in-depth interviews to completely
                identify data needs, because the scope of our review was broad, we had
                to limit data collection to one round of interviews. Therefore, our results
                concerning EPA and state data needs should be considered preliminary.
                The second limitation of our study stems from the recommendation in
                the literature on the analysis and design of formal information systems
                that conclusions should only be drawn about independent systems, such
                as a business that is not a subsidiary of another.5 The hazardous waste
                information system is not entirely independent in this sense, which
                made it difficult for us to determine the boundaries of the information
                system for our evaluation. We include the toxic chemical release inven-
                tory reporting system as a mechanism that can be relied on in EPA'Sper-
                manent hazardous waste information system because it deals with
                hazardous waste and contains information important to the hazardous
                waste program. However, it should be noted that the reporting system
                was not mandated by RCFLA, and the EPAoffice with the lead responsibil-
                ity for implementing RCRA did not have the lead responsibility for devel-
                oping this system. While it is appropriate for our purposes to treat this
                reporting system as part of the hazardous waste information system, it
                could also be viewed as part of other EPA information systems.

                We obtained oral comments on a draft version of this report from EPA
                officials. Their comments have been incorporated where appropriate.
                Our review was conducted in accordance with generally accepted gov-
                ernment auditing.


                “Andre Blokdijk and Paul BlokdUk, Planning and Design of Information Systems (New York:   Aca-
                demic Press, 1987), pp. 39-40.



                Page 17              GAO/PEMD-90-3
                                                EPA'sHazardous          Waste Data Need Further    Improvement
                      Chapter 11
                      IntroduclJtu 7




                      In chapter 2, we identify the problems experienced by the original infor-
Rebort Organization   mation system and describe EPA'Sefforts since 1986 to improve the qual-
                      ity of its information on hazardous waste generation and management.
                      We describe both the interim information system these efforts produced
                      by August 1988 and the planned permanent information system. We
                      also present the results of our evaluation of EPA'Sefforts to improve the
                      information system development process. Chapter 3 presents the results
                      of our assessment of the extent to which EPA has identified and made
                      provisions for obtaining information needed to carry out activities man-
                      dated by federal laws pertaining to hazardous waste. In chapters 4 and
                      6, respectively, we present the results of our evaluation of EPA'S mea-
                      surement instruments and data collection methods. In each of these
                      chapters, we also identify remaining problems, discuss our conclusions,
                      and present recommendations concerning further improvements.




                      Page 18          GAO/PEMD-99-3   EPA’s Hazardous   Waste Data Need F’urther Improvement
                            This chapter presents the problems that were identified by EPA and
                            other experts as contributing to low data quality in the original informa-
                            tion system and describes EPA'Sefforts to correct them. In addition, we
                            evaluate EPA'Sefforts to establish an improved information system
                            development process.


                            In this section, we answer our first evaluation question: What problems
      lems With EPA’s       have been identified with the original hazardous waste management
     inal Information       information system?

                            Between 1976 and 1986, EPA established an information system to sup-
                            port the hazardous waste management program required by the Con-
                            gress. The system included five data collection mechanisms, each of
                            which provided some technical data on the generation or management of
                            hazardous waste. The RCRAreporting system (also known as the biennial
                            report) was the primary mechanism for periodic data collection, but EPA
                            also relied extensively on special sample surveys for information not
                            provided by the RCRAreporting system. Additional mechanisms included
                            management facility operating and closure permit applications,
                            manifests, and notifications of hazardous waste activity. While these
                            additional mechanisms do not serve primarily to collect information on
                            hazardous waste generation and management, they contain some techni-
                            cal data and therefore should be considered in the overall estimation of
                            data availability. In the following subsections, we first describe each of
                            the five data collection mechanisms and summarize the problems EPA
                            experienced while using each of them. We then summarize the problems
                            EPA experienced with the system as a whole-that      is, problems that
                            involve the joint functioning of the separate components.


Desqription of 0 sriginal   Since its establishment in 1980, the RCRAreporting system has been the
                            principal mechanism for the periodic collection of information on haz-
Data Collection             ardous waste generation and management. In 1980, EPA published rules
Mechanisms                  establishing the forerunner of the current system. These rules created
                            an annual census of large quantity hazardous waste generators and
The FCRA Reporting System   management facilities to be conducted in each state. States that that had
                            been authorized (see glossary) by EPAto operate their own hazardous
                            waste program in lieu of RCRAcould collect their own information as
               Y            long as they met the general RCRAauthorization requirements. That is,
                            reporting requirements in authorized states had to be equivalent to, con-
                            sistent with, and not less stringent than the federal reporting require-
                            ments; however, authorized states were not required to collect identical


                            Page19          GAO/PEMD-90-3
                                                       EPA’sHazardousWasteData NeedFurtherImprovement
                                                                                              I




Chapter 2
Initial Pwblema and EPA
Improvemc :gj Effoti




information. Authorized s tates were to submit summaries of the infor-
mation they collec ted to EPA. EPA collec ted the data, using its own ins tru-
ment, in s tates that were unauthorized.

In 1982, EPA proposed to shift to a national, biennial sample survey. EPA
preferred this approach over the s tate census because the agency
thought the data would be more verifiable, contain more detail, produce
better national information, and reduce the overall paperwork burden
on indus try. EPA s tated specifically that, under the s tate census
approach, “the variety of forms and data processing s y s tems used by
the s tates would probably preclude timely and effic ient data analy s is .“1

According to the plan, EPA was to be solely responsible for the survey
 and thus would not have to rely on summaries or other information
transfers from authorized s tates . States could continue to collec t addi-
tional information from all hazardous waste handlers (see glossary), but
they were no longer required to do so. The proposal provided for s tates
to add questions, increase the sample s ize, and conduct the survey
jointly with EPA to avoid any problems that might arise from differing
s tate and federal authority .

The plan was flawed, however, in that it contained no specific provis ion
for maintaining a current lis t of all handlers or the basic information
required about each that is necessary for enforcement and other pur-
poses. In addition, many of those who filed comments with EPA believed
the plan would increase rather than decrease the overall paperwork bur-
den on indus try because, under the plan, s tates could maintain their
exis ting reporting requirements. As a result of negative comments and
the threat of litigation, EPA withdrew the proposal and ins tead issued
rules in 1983 establishing the current biennial s tate census.

As established by current regulations and EPA polic y , the RCRAreporting
s y s tem consis ts of three tiers . Firs t, federal regulations require handlers
in unauthorized s tates to report direc tly to EPA, using a specific EPA data
collec tion ins trument, Authorized s tates may use their own ins truments,
which must collec t information that is consis tent and equivalent to the
EPA ins trument (but not necessarily identical). In the second tier, federal
regulations require authorized s tates to submit summary reports to EPA.
As established by EPA polic y , the third tier of the RCRAreporting s y s tem

'EPA,“ Hazardous W aste Management System: Standards Applicable to Generators of Hazardous
W aste and Standards Applicable to Owners and Operators of Hazardous W aste Treatment, Storage,
and Disposal Facilities,” Federal Register, 47:197 (October 12, 1982), p. 44933.



Page 20              GAO/PEMD-90-3     EPA’s Hazardous W aste Data Need Further    Improvement
                        Chapter 2
                        Initial Problems and EPA
                        Improvement    Efforte




                        consists of the agency’s compiling data and then publishing a national
                        report2 EPA has completed three RCRA reporting cycles covering 198 1,
                        1983, and 1985. However, in 1988, EPA documents stated that the RCRA
                        reporting system had never produced valid national data. EPA officials
                        now believe the report on the 1986 reporting cycle published in March
                        1989 contains “reasonably valid data” on the total volumes generated
                        and managed. However, they acknowledged that the problems discussed
                        later in this chapter prevented timely data analysis and that some of the
                        problems could not be overcome.

                        Four additional data collection mechanisms in the original information
                        system (through 1986) did or could have potentially provided informa-
                        tion on hazardous waste generation and management. We briefly
                        describe each of them in the following sections of this chapter.

Speci4.lSampleSurveys   EPA'Soriginal information system, which existed through 1986, relied on
                        national probability sample surveys for more detailed and uniform
                        information than that provided by the RCRA reporting system. EPA also
                        has conducted other smaller surveys for more limited purposes. EPA con-
                        ducted a major national survey of hazardous waste generators and man-
                        agement facilities (originally intended as the first biennial survey),
                        covering calendar year 1981, to support the regulatory impact analyses
                        required by Executive Order 12291. Since then, EPA has used the survey
                        data extensively for developing regulations.

                        The remaining data collection mechanisms were not designed primarily
                        for collecting generation or management data to characterize the regu-
                        lated population. However, they do contain some information on hazard-
                        ous waste generation and management, and therefore should be
                        considered in the overall estimation of data availability.




                        2Although HSWA does not require a formal report to the Congress, the language of the conference
                        committee report strongly implies that the Congress will mandate a periodic report if the RCRA
                        reporting system does not prove satisfactory. The conference committee report concluded that since
                        EPA had begun a program (the 1983 cycle of the RCR.4 reporting system was then under way) to
                        provide the needed information to the Congress and the public, a “formal report to Congress” is not
                        needed. The report adds that the “administrator is expected to continue this program and to seek
                        more accurate data than has been available in the past” (Hazardous and Solid Waste Amendments of
                        1984 Legislative History, U.S. Code Congressional and Administrative News, 98th Cong., 2nd sess.,
                        1984. Public Law 98.616, p. 6706).



                        Page 21               GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further     Improvement
                              Chapter 2
                              Initial PnrMenw and EPA
                              Improvemc 1.” Efforta




operating and ClosurePermit   Hazardous waste management facilities are required to seek and obtain
Ap@ications                   operating and closure permits.3 Federal regulations specify a two-part
   I                          permit application, Part A of the permit application is a standard form
                              that contains specific information listed in federal regulations. For
   I                          example, Part A includes the types and quantities of hazardous wastes
                              the facility plans to manage and a description of the processes to be
                              used to manage the waste: including the design capacity (see glossary).
                              Part B of the permit application contains extensive additional, but not
                              standardized, information designated in federal regulations.

M+fests                       To track the cycle of hazardous wastes, RCRA requires each consignment
   I                          of hazardous waste shipped off-site for management (including CERCLA
                              and corrective action wastes) to be documented in a manifest. The haz-
                              ardous waste manifest contains information on the type, quantity, and
                              disposition of the hazardous wastes shipped away from the point of gen-
                              eration The receiving management facility must retain a copy of each
                              manifest and return copies to the transporter and the generator, who
                              must also retain them.

Notification                  Federal regulations require handlers of hazardous waste to inform EPA
                              or authorized states of their regulated activities. Upon initial notifica-
                              tion, EPA assigns the handler an identification number. This is the princi-
                              pal mechanism for identifying the regulated population. The federal
                              notification document contains information on the type of regulated
                              activities and the type of regulated wastes handled.


Problems With the Data        EPA experienced problems with all the previously discussed data collec-
Collection Mechanisms         tion mechanisms. Based on existing literature and interviews with EPA,
                              state, and other experts familiar with the system, we divided the prob-
                              lems EPA experienced with each data collection mechanism into three
                              general areas: (1) information requirements, (2) measurement, and (3)
                              data collection. Table 2.1 summarizes the problems EPA experienced with
                              the RCRAreporting system in each of these areas. Table 2.2 summarizes
                              the problems EPA experienced with the additional data collection mecha-
                              nisms and notes the area(s) in which problems were experienced.


                              31n addition to a permit to operate a management facility in an approved manner, each facility must
                              have an approved plan for how the facility will eventually be closed-to ensure that it does not
                              become an uncontrolled hazardous waste site-and a post-closure plan for monitoring. Facilities that
                              are closed continue to require monitoring by their owners. Facilities that recycle hazardous wastes do
                              not require operating permits, but most require storage permits. A limited number of recycling opera-
                              tions are exempt from all permit requirements.



                              Page 22               GAO/PEMD-90-3      EPA’s Hazardous    Waste Data Need Further     Improvement
                                    Chapter 2
                                    Initial Problems and EPA
                                    Improvement    Efforts




Table 2.11:Problems With the RCRA
Reporting System                    Problem type and specific
                                    problem                          Problem description
                                    Information requirement
                                    No information on                EPA and most states did not collect information on the
                                    management characteristics       characteristics of wastes that determine appropriate
                                    of wastes                        management practices
                                    No information on amount         The EPA data collection instrument did not obtain
                                    managed and disposed of by       information on the amount of each type of waste managed
        I                           each management and              and disposed of by each management and disposal
                                    disposal technology              technology, such as landfills
                                    No information on amount of      The EPA data collection instrument did not obtain
                                    management capacity              information on available management capacity
                                    No quantitative information on   The EPA data collection instrument did not obtain
                                    waste minimization efforts       quantifiable data on the extent of waste minimization or the
                                                                     specific efforts to reduce waste generation
       /
                                    Measurement
                                    inadequate measure of waste      Handlers reported similar wastes   as different EPA-defined
                                    type                             waste types (EPA waste codes);     many handlers and states
                                                                     reported wastes as mixed, which     prevents calculation of
                                                                     amounts of each type
                                    Inadequate measure of the        Some wastes were not counted,      and others were counted     -
                                    amount of wastes generated       more than once
                                    and manaaed
                                              w-

                                    lnadeauate measure of            Tvpes of manaaement technologies were not sufficiently
                                    management technologies          specified; EPAcould not determTne intermediate treatment
                                                                     steps or how much waste was managed by what types of
                                                                     technoloaies
                                    Inadequate measure of            EPA and state data collection instruments did not
                                    regulated status                 adequately update notification forms, which prevented EPA
                                                                     from developino an accurate list of active handlers
                                    Data collection
                                    Inconsistent data processing     Few states automated the RCRA reporting system, and
                                    systems across states            many had low quality control; the data received by EPA
                                                                     varied in timing, form, and quality
                                    Inconsistent data collected      Lack of a uniform data collection instrument and
                                    across states                    inconsistent EPA guidance caused the collection of
                                                                     inconsistent data because (1) some states and EPA regions
                                                                     instructed handlers to report wastes treated in exempt
                                                                     processes and some did not, (2) some states had more
                                                                     stringent definitions of hazardous waste, and (3) some
                                                                     states used differing measures; in addition, summary data
                                                                     concealed differences so that EPA could not separate state
                                                                     from RCRA regulated wastes




                                    Page 23              GAO/PEMD-90-3    EPA’s Hazardous   Waste Data Need Further   Improvement
                                    Chapter 2:
                                    Initial Pr&i   ferns and EPA
                                    Improvemer       Efforts




Tabii 2.2: Problems With Special
Surqeys, Permit Applications, and   Source                         Use                   Problem
Menjfests                           Special sample             EPA has relied on         Sample surveys do not provide the
                                    surveys                    special sample            information that is needed on all handlers
                                                               surveys for               and quickly become out-of-date (data
                                                               information not           collection problems)
                                                               provided by !he
                                                               $;FIt;eportlng

                                    Operating and closure      Management facilities     Permits are not required for generators (data
                                    permit applications        are required to obtain    collection) and are not accurate reflections
                                                               a permit detailing        of many actual activities (information
                                                               activities likely to be   requirements); some valid data are included
                                                               performed                 (such as limiting conditions of operation),
                                                                                         but the permits are not accessible because
                                                                                         they are retained by state and local offices
                                    Manifests                  In order to track its     Manifests are not uniform across states and
                                                               movement, each            contain different information in different
                                                               consignment of            states (information requirements and
                                                               hazardous waste           measurement), federal regulations do not
                                                               shipped away from         require the return of manifests to EPA or
                                                               the generating facility   state programs, and most hazardous waste
                                                               is manifested             is managed at the generating facility where
                                                                                         manifests do not apply (data collection)
                                    Notification               Hazardous waste           Notification contains limited information
                                                               handlers are required     (information requirements) and is often not
                                                               to notify EPA or states   updated to reflect current regulatory status,
                                                               of regulated activities   activities, or wastes handled (data
                                                               and type of waste         collection)
                                                               handled




Prbblems With the                   EPA experienced problems at the overall information system level. We
Information System as a             defined system-level problems as those that affected more than one
                                    component of the system or those that involved the joint functioning of
Whole                               system components. The first system-level problem was the lack of inte-
                                    gration among system components-that         is, the different data collection
                                    mechanisms failed to function together to produce information. For
                                    example, EPA officials explained that handlers often used a variety of
                                    codes from different data collection mechanisms to complete portions of
                                    the RCXAreporting system data collection instrument and that this
                                    caused problems in interpreting the responses. In addition, some states
                                    used manifest data, which was inconsistent across states, to prepare
                                    part of the information for their RCRAreporting system submissions. The
                                    problem thus created consisted of a lack of system integration, involving
                                    the use of inconsistent data collection instruments and systems among
                                    states, between EPA and states, and for both states and EPA over time.




                                    Page 24                GAO/PEMlMO-3      EPA’s Hazardous   Waste Data Need Further   Improvement
                                              chapter 2
                                              Initial Problems and EPA
                                              Improvement    Efforts




                                              The second type of system-level problem involved EPA’Sinternal process
                                              for developing and modifying its information system. Specifically, prob-
                                              lems in the area of system development included (1) the lack of a com-
                                              prehensive plan to coordinate data collection efforts, (2) the lack of a
                                              comprehensive evaluation of existing regulatory development data, (3)
                                              isolated data collection efforts that were narrowly focused and resulted
                                              in duplication of effort, and (4) data planning and collection responsibil-
                                              ities that were divided among program offices with inadequate integra-
                                              tion of data collection efforts (including cross-media data collection
                                              efforts). System development problems such as those just described are
                                              associated with individual data collection mechanisms, as well as the
                                              information system as a whole. Therefore, although it was not possible
                                              to demonstrate that system development problems caused the specific
                                              problems just discussed, such a connection is plausible.

                                              Table 2.3 indicates the major problem areas that affected each individ-
                                              ual data collection mechanism and the original information system as a
                                              whole. Through 1986, these problems had prevented EPA from determin-
                                              ing with reasonable certainty how much hazardous waste of what type
                                              was generated, how it was managed, whether sufficient management
                                              capacity existed, or whether progress was being made in reducing waste
                                              generation.


Table a.3: Problems With Individual Data Collection Mechanisms and the Original Information System as a Whole
                                                                                      Problem area
                                                        System          Information                           Data collection
Data collection mechanism                      ~~~ -.-  development     requirements         Measurement      method
RCRA reporting system                                                              X”                      X                    X
Special   sample surveys                                ~-...       -                                                           X
-.____..II_.
Permit ar3plications                                                               X                                            X
Manifegts                                                                          X                       X                    X
Notification                                                                       X                                            X
__-_._-..-,....      ..-   ~~   .~   -   --.-...-
The svstem as a whole                                           X                                                               X
                                              “An “X” indicates that a problem existed




                                              Page 25                   GAO/PEMD-90-3    EPA’s Hazardous   Waste Data Need Further   Improvement
                     Chapter 2
                     Initial Yrtiblema and EPA
                     Improvetiwnt    Efforts




                     This section addresses the second evaluation question: What steps have
EPA Actions to       been taken by EPA to improve the quality of its hazardous waste genera-
Iqprove the          tion and management information system?
Information System   EPA initiated 13 efforts to improve the information system that have
                     directly affected the quality of hazardous waste generation and manage-
                     ment information4 Four of these efforts were aimed at improving the
                     overall system development process, while the remaining nine were
                     aimed at specific components (such as data collection mechanisms) or
                     parts of components. Of these nine efforts, three were undertaken coop-
                     eratively with the National Governors’ Association (NGA), while the
                     remainder were internal EPA initiatives. These actions resulted in two
                     general outcomes: (1) the establishment of data collection mechanisms
                     to serve national objectives temporarily until permanent mechanisms
                     become fully functional and (2) the development of plans for the perma-
                     nent mechanisms. We refer to these outcomes as EPA'Sinterim and per-
                     manent information systems. The interim system includes a partially
                     revised RCRA reporting system in combination with one-time-only
                     national surveys of hazardous waste generators and management facili-
                     ties. The permanent system will rely primarily on a fully revised RCRA
                     reporting system and can also draw on the toxic chemical release inven-
                     tory reporting system required by SARA.The reporting requirements for
                     the toxic chemical release inventory reporting system are to be phased
                     in during the interim phase and will become fully functional in the per-
                     manent phase by 1991. The permanent system will not include the
                     national sample surveys contained in the interim system but will con-
                     tinue to include the more limited scope surveys included in the original
                     information system.

                     Table 2.4 depicts the original problem area(s) that each of the 13 actions
                     were intended to improve. Table 2.5 describes the four system develop-
                     ment efforts. Table 2.6 describes the improvement efforts undertaken
                     cooperatively by EPA and NGA, while table 2.7 lists the six improvement
                     efforts implemented solely by EPA. Table 2.8 shows the status of each
                     new or revised data collection mechanism in the interim and permanent
                     information systems.




                     4EPA has initiated numerous other efforts to improve its information system that are not directly
                     related to the quality of technical data on hazardous waste generation and management. An evalua-
                     tion of these efforts was beyond the scope of our report.



                     Page 26              GAO/PEMD-90-3     EPA’s Hazardous    Waste Data Need Further    Improvement
                                                 Chapter 2
                                                 Initial Problems and EPA
                                                 Improvement    Efforta




Table 2.4: Problem Areas Addressed by Each EPA Improvement Effort
                                                                                                   Problem area
                                                               System               Information
Improvdment                                                    development          requirements          Measurement          Data collection




 4. Devblooment   of a life cycle manaaement   system          X
 5. E       n and evaluation of the RCRA reporting system                           X                     X                    X

 6. Dedelopment of the SARA capacity assurance                                      X
    req irements                                                            -                                                                     -
       Y
 7. Development of the Toxic Chemical Release Inventory                             x   -                 X                    X
    SvsJem
 8. Reviiew of a sample of completed forms to assess 1985                           X                     X                    X
    RCRA reoortina cvcle
 9. Devefocment of a hazardous waste manaaement                                     X                     X                    X
    simulation model
10. Conduct of a national survey of hazardous waste                                 X                     X                    X
    manaaement facilities
11. Conduct of a national survey of hazardous waste                                 X                     X                    X
    generators                                                                  -
12. Rec/esian of the hazardous waste manifest                                       X                     X                    X
13. DeGelopment of a new measure of hazardous waste                                                       X
    WPb

                                                aAn “X” indicates that a problem area was addressed.




                                                Page 27              GAO/PEMD-90-3      EPA’s Hazardous   Waste Data Need Further   Improvement
                                             chapter 2
                                             Initial Pwblemfi and EPA
                                             Improvew~ .lt Efforts




Tat+le 2.5: Specific Activities to Improve
the; EPA System Development Process          Improvement                  Purpose                   Result
                                             1. Evaluation of data        To improve the            A complete listing of data sources relevant
                                                collection activities     coordination of future    to decision making was developed; all
                                                and data sources          data collection           activitities were assessed, as were their
                                                                          activities by providing   interrelationships with each other;
                                                                          a framework for           recommendations on how to improve data
                                                                          collecting all of the     collection activities, individually and in
                                                                          office’s regulatory       relation to one another, were developed and
                                                                          development data          proposed
                                                                          and to assess past
                                                                          and current activities
                                             2. Establishment of a        To establish an           A coordinating office for information system-
                                                central coordinating      organization that         activities was established within the Office
                                                office                    could centrally control   of Solid Waste; responsibility for all
                                                                          all information           components of the hazardous waste
                                                                          system-related            information system was transferred to this
                                                                          activities                office
                                             3. Establishment of a        To establish a            A fully functional tracking system was
                                                data collection           mechanism designed        established that lists all data collection
                                                tracking system           to monitor the            activities approved by the Office of
                                                                          relationship and          Management and Budget
                                                                          efficiency of data
                                                                          collection activities
                                             4. Development     of a      To assure that            EPA implemented a life cycle management
                                                Life cycle                information systems       system covering hazardous waste
                                                management                developed meet the        information system development
                                                svstem                    reauirements




                                             Page 29                   GAO/PEMD-90-3    EPA’s Hazardous   Waste Data Need Further   Improvement
       .

                                                                                                                               -



                                 Chapter 2
                                 Initial Problems and EPA
                                 Improvement    Efforts




Table 246: EPA/NQA Improvement
Activitids                       Improvement        Purpose                                 Result       ______
                                 1. Redesi n of     To revise the major reporting           A policy decision was made that
                                    the RC 1 A      component of the information            the revised system will be the
                                    reporting       system and to develop a                 single permanent routine data
                                    system          partnership with the states for its     collection mechanism; the effort
                                                    use                                     also resulted in the development of
                                                                                            a plan for fully revising the
                                                                                            reporting system and interim
                                                                                            procedures used for the 1987
                                                                                            reporting cycle-national      surveys
                                                                                            such as those conducted during
                                                                                            the interim phase will be
                                                                                            discontinued, but EPA will continue
                                                                                            to use limited scope surveys
                                                                                            sponsored by the substantive
                                                                                            d.ivisions
                                 2. Development     To meet the requirements of SARA,       Data requirements to support the
                                    of SARA         specifications needed to be             technical analyses were developed,
                                    capacity        developed on how the capacity           and the technical analyses to be
                                    assurance       assurance analyses were to be           conducted were defined
                                    requirements    performed and what data were
                                                    required
                                 3. Development     To develop a system which would         EPA has promulgated final rules for
                                    of a toxic      provide both public information on      the establishment of the inventory
                                    chemical        the use and management of toxic         reporting system; the system IS a
                                    release         chemicals, and data for regulatory      mandatory census of firms that
                                    inventory       decision making                         manufacture, process, or use toxic
                                    system                                                  chemicals




                                 Page 29            GAO/PEMD-90-3     EPA’s Hazardous     Waste Data Need Further Improwment
                                          Chapter 2
                                          Initial Ymblenw and EPA
                                          Improvexwt*% Efforts




Taqle 2.7: EPA improvements   Developed
Internally                                Improvement          Purpose                                    Result
                                          1. Conduct           To confirm the 1985 RCRA                   A sample of 1985 hazardous waste
                                             survey to         reporting cycle data and assess            handlers was developed; problems
                                             assess1985        their problems                             in the areas of data accuracy and
                                             RCRA                                                         data completeness were identified
                                             reporting cycle
                                          2. Develop           To develop a simulation model that         Development of an assumption-
                                             hazardous         could be used to estimate volume           driven model that determines (1)
                                             waste             managed by each technology                 appropriate treatments, (2)
                                             management                                                   sequence of treatments, (3) overall
                                             simulation-                                                  volume reduction associated with
                                             model                                                        treatment, and (4) amount of waste
                                                                                                          treated in each treatment
                                                                                                          sequence
                                          3. Conduct           To collect detailed data on 1986           Partial results of survey were used
                                             survey of         waste management activities to             to support the first third of the land
                                             hazardous         support capacity assessments               disposal restrictions: survey was
                                             waste             required by land disposal rules of         completed after our evaluation data
                                             management        HSWA                                       collection was concluded
                                             facilities
                                          4. Conduct           To collect detailed data on the            Results were not available until
                                             survey of         nation’s hazardous waste                   after our evaluation data collection
                                             hazardous         generators                                 was concluded
                                             waste
                                             generators
                                          5. Redesign          To address problems in data                Initiative postponed
                                             hazardous         collection, measurement, and
                                             waste manifest    requirements area
                                          6. Develop new       To improve reliability, validity, and      Initiative discontinued
                                             measure of        usefulness of the measure
                                             hazardous
                                             waste tvoe


Table 2.6: Interim and Permanent Status
of New and Revised Data Collection        Data collection mechanism         Interim phase                       Permanent phase
Mechanisms                                National survey of hazardous      Implemented and completed           No further national surveys
                                          waste manaoement facilities       by late 1988                        conducted
                                          National survey of hazardous      Implemented and completed           No further national surveys
                                          waste generators                  by 1989                             conducted
                                          RCRA reporting system             Partially revised system            Fully revised system to be
                                                                            implemented for 1987 cycle;         implemented
                                                                            completion originally planned
                                                                            for 1989 now planned for
                                                                            1991
                                          Toxic chemical release            Final regulation published;         System fully implemented
                                          inventorv reportina svstem        svstem partiallv implemented




                                          Page 30              GAO/PEMD-90-3      EPA’s Hazardous      Waste Data Need Further      Improvement
                                  Chapter2
                                  Initial Problem@ and EPA
                                  Improvement    Efforts




                                  This section addresses our third evaluation question: Is the process EPA
Evaluation of Efforts             initiated to effect improvements consistent with generally accepted
t0 I&prove the system             standards for developing information systems?
Devblopment Process               Of the four efforts aimed at improving information system development
                                  shown in table 2.4, the first three were aimed at increasing the level of
                                  overall information system integration and coordination among develop-
                                  ment projects, while the final effort was aimed at planning specific
                                  information system components to assure that they fully support the
                                  program mission. Our evaluation of these four efforts is given below,
                                  while evaluations of the other improvement activities are presented in
                                  succeeding chapters.


Ovenall Information               The EPA evaluation includes a comprehensive listing of existing data
System Integration and            bases and provides the foundation for developing a plan for coordinat-
                                  ing data collection activities. The draft report contained six recommen-
Coo+dination                      dations for improving the management and dissemination of existing
                                  and ongoing information activities. The responsible EPA official
The Evaluation of Data            explained that these draft recommendations were not final and that
Collection Activities             alternative recommendations were being developed in conjunction with
                                  the contractor. The evaluation represents a significant accomplishment
                                  both because it systematically catalogues existing data sources and
                                  because it is a step toward establishing a comprehensive plan for data
                                  collection.

The Central Coordinating Office   The establishment of a central data coordinating office in EPA'SOffice of
                                  Solid Waste, which administers RCRA,is also an important accomplish-
                                  ment because it establishes a known authority and responsibility for
                                  developing integrated information systems and fosters a supportive atti-
                                  tude. Such practices are recommended by the Office of Management and
                                  Budget, the General Services Administration, and GAO systems develop-
                                  ment guidelines for implementing the Paperwork Reduction Act. The
                                  evaluation of data collection activities, discussed previously, indicated
                                  that increased coordination was not achieved without difficulty. How-
                                  ever, staff gradually began to recognize that the existence of a central
                                  focal point facilitated the interactive agreement necessary for meeting
                                  complex data needs. All 13 of the improvement projects we discussed
                                  above had some contact with this central data coordinating office. Such
                                  contact helped reduce duplication and helped increase both the extent to
                                  which single data collection efforts met the needs of multiple users and
                                  the extent to which multiple data collection efforts complemented one
                                  another.


                                  Page 31             GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
Chapter 2
Initial Yn~blems and EPA
I.mprove:wnt   Efforts




The central coordinating office, however, has not yet completely cor-
rected the problems of the past. EPA has still not developed a comprehen-
sive data collection plan. The previously discussed evaluation of data
collection activities concluded that without such a plan to act as a “forc-
ing function” for coordination and integration, data collection efforts,
many of which are performed by EPA contractors, could continue to be
isolated.

The central coordinating office also did not have full authority to
develop new data collection efforts. This office had to rely on contribu-
tions from the budgets of the divisions to develop information sources.
That is, the budget for information system development was not sepa-
rated from the substantive divisions. Other offices continued to have
primary responsibility for developing key aspects of the information
system. In addition, there was no prime contractor with overall respon-
sibility for the hazardous waste information system. EPA officials stated
that these factors contributed to a lack of integration, including conflict-
ing plans and duplication of effort.

One important instance of a lack of integration of data collection efforts
was that the plan for the 1987 RCR.4reporting cycle was finalized before
the requirements for SARAcapacity assurance were developed, even
though one important purpose for redesigning the reporting system was
to provide information for capacity assurances. EPA officials stated that
they were forced to “jury-rig” different data collection mechanisms in
order to develop the data necessary for the capacity assurances
required by SARA.EPA officials also observed that the timing of HSWAand
SARArequirements complicated their efforts. (We discuss the relation-
ship between data collection efforts and intended uses further in chap-
ter 5.) EPA officials stated that another important instance of lack of
integration was that divisions offering funding for a specific data collec-
tion project have not always followed through with the funds to com-
plete it.

In addition, the Office of Toxic Substances and the Office of Solid Waste
did not fully integrate the toxic chemical release inventory reporting
system required by SARAwith the other reporting systems related to haz-
ardous waste. This jeopardizes EPA efforts to achieve its cross-media
objective (discussed earlier) and contributes to the continuing isolation
of data collection efforts. Officials in EPA'SOffice of Solid Waste pointed
out that this data collection mechanism is not intended primarily to
implement RCRA,although it does capture information about hazardous
waste. This is all the more reason for ensuring that the data collected


Page 32            GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                                          ,.
                             Chapter 2
                             Initial Problem8 and EPA
                             Improvement    Efforta




                             can be used to supplement efforts that are primarily intended to imple-
                             ment Rcfu.
The @ataCollectionTracking   The data collection tracking system, part of the office’s Information Col-
Systkjm                      lection Budget program, was intended to list all information requests
                             submitted for Office of Management and Budget approval, the level of
                             effort required, and the funding approved by this agency. Such a system
                             has real potential for helping to assure the integration of data collection
                             efforts. However, the system is not fulfilling this potential for three rea-
                             sons. First, it is incomplete because some projects or revisions to ongoing
                             projects are not included. Second, the system is not organized for maxi-
                             mum impact on coordination and integration because submission of doc-
                             umentation to the system occurs too late in the planning process for it to
                             be effective. Finally, the system lacks adequate coordination authority
                             in that the staff who operate the tracking system are not responsible for
                             working with the program offices to coordinate and integrate data col-
                             lection efforts.


Planning Specific            Life cycle management is a standard approach to developing and revis-
                             ing information systems or components. It is based on the concept that
Infoirmation System          all information systems progress through the same basic stages from ini-
Components: the Life         tial development to operation and maintenance and, finally, termination.
Cyc$eManagement System       The President’s Councils on Management Improvement and on Integrity
                             and Efficiency view the concept of life cycle management as the conven-
                             tional approach to developing information systems that “evolved
                             because of the need for managers to assess the totality of work to be
                             undertaken, and to develop plans accordingly.“6 EPA’S fiscal years 1987-
                             89 Information Resources Review Plan now requires life cycle
                             management.

                             Life cycle management requires specific documentation (that is, reports
                             on specific topics) for each stage of the cycle, the level of which should
                             be commensurate with the importance of the information system.
                             Although the number of stages varies in different applications, the basic
                             documentation is standard and includes a needs statement, a feasibility
                             assessment, a risk analysis, a cost benefit analysis, and a system deci-
                             sion paper. The life cycle management system is new and thus only cov-
                             ered one of the major data collection efforts we identified-the     redesign

                             “President’s Council on Management Improvement and President’s Council on Integrity and Effi-
                             ciency, Model Framework for Management Control Over Automated Information Systems (Washing-
                             ton, DC.: US. Government Printing Office, January 1988), p. 22.



                             Page 33             GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
                                                                                         -
Chapter 2
Initial Pr&lems and EPA
Improvewiur it Efforts




of the RCRAreporting system. The other major data collection efforts
developed by this office- that is, the national surveys of hazardous
waste generators and management facilities-were      not covered. The
planning documentation did, however, discuss the relationship between
the national surveys and the RCRAreporting system. We reviewed the
implementation of the life cycle management system for the RCRAreport-
ing system according to the federal guidelines discussed earlier. Since
the RCRAreporting system is crucial for providing hazardous waste gen-
eration and management information, we discuss the life cycle manage-
ment documentation for the system in detail.

The life cycle management system, as implemented, is consistent with
EPA policy and other federal guidelines. It includes all the required docu-
mentation. Although it does not include a separate risk analysis docu-
ment, risks are discussed. Risk analyses are important because they
assess the extent to which the organization will become dependent on
the system, the consequences of failure, how failure can be avoided, and
the type of backup that should be required.

The implementation of the life cycle management system has the poten-
tial to assure that information system planning supports the mission of
the organization. EPA'Suse of the system was especially strong in the
area of incorporating user participation. EPA ensured that the views of
state program users were incorporated by working with an advisory
council, composed primarily of state program personnel, that was estab-
lished by NGA to assist in the revision of the RCU reporting system.

The documentation of the analyses performed, however, was limited,
given the importance of the system and the extent to which the agency
will depend on it. Even though the system is crucial, summary documen-
tation (brief statements indicating what analyses were performed and
the conclusions) would have been satisfactory if more in-depth work
had been cited. However, the additional work that was cited contained
only general statements and, in some cases, EPA officials stated that no
more in-depth work had been conducted. In the following paragraphs,
we summarize the limitations of the documentation provided for the
1987 revision of the RCRAreporting system.

The needs statement, which documents problems in the existing infor-
mation system and the need to redesign it, did not include a complete
analysis of existing problems. Since the document states that the RCRA
reporting system has never produced valid national data, it would be
expected that the existing problems would be thoroughly assessed.


Page 34            GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
chapter 2
Initial Problems and EPA
Improvement Efforts




Neither the report on the evaluation of a sample of the completed forms
from the 1986 RCRAreporting cycle nor the NGA reports cited in the
needs statement contained a systematic assessment of previous
problems.

The RCRAreporting system was vulnerable to problems caused by the
differences between the various state data collection systems and proce-
dures. And, although the many differences between state procedures
remain a feature of the revised system, they (and their implications for
the revised system) were not thoroughly assessed. Since the revised sys-
tem relies on the same existing state systems that have produced signifi-
cant problems in the past, the lack of a careful analysis of how the
revised system would solve past problems is a notable omission. In addi-
tion, there was no assessment of the extent to which EPA will be depen-
dent on the adequate functioning of the RCRAreporting system. This
indicates that the risks associated with the revised system were not
completely addressed.

EPA'S cost benefit analysis considered only two alternatives to the cur-
rent approach to collecting generation and management information for
regulatory development purposes: (1) relying entirely on national
surveys sponsored by EPA, and (2) relying entirely on an expanded RCRA
reporting system. The possibility of using a more effective combination
of these strategies than was used in the past was not examined in the
feasibility study.

EPA chose to eliminate national surveys-such    as the surveys of genera-
tors and management facilities employed in the interim phase-primar-
ily to save costs. No weight was given to the fact that the RCRAreporting
system had not produced valid national information in a timely manner
or that the plan would make EPA totally dependent on an untested entity.

Perhaps most importantly, the life cycle management documentation did
not contain a systematic analysis of the required information. The docu-
mentation lists 17 standard management reports that are needed, but it
does not link the reports to either program needs or specific data ele-
ments. The requirements analysis report cites NGA background reports
as providing additional information on the development of the data col-
lection instrument. However, while the NGA reports do contain addi-
tional information, it is general information rather than a detailed
analysis of EPA or state program information needs. And, while the




Page 36            GAO/PEMD-90-3
                              EPA'sHazardous   Waste Data Need Further   Improvement
Chapter 2
Initial P1*Mems and EPA
Improvea+* -rt Efforta




Office of Management and Budget approval package for the RCRA report-
ing system included somewhat more detail, the discussion there is also
general in nature.

The NGA advisory council repeatedly requested that EPA prepare an anal-
ysis of the needs to be served by each proposed data element. EPA offi-
cials promised to prepare such a report for the final meeting of the NGA
advisory council (in October 1987) before the beginning of the 1987
reporting cycle. According to EPA officials, however, time pressures pre-
vented its accomplishment.

The documentation focuses on EPA'Sneed for regulatory development
data but does not analyze state data needs. EPA recognized that states
have different information needs and that the revised RCRA reporting
system would require states to collect information they do not need. The
system concept report, part of the life cycle management system, states
that

“Core data items [required from states] are those items which are collected by imple-
mentors and transmitted to oversight [EPA]. Implementors generally, but not neces-
sarily, have interest in these items. Implementors have a responsibility to validate
core data items” [emphasis added].”

Requiring the states to collect information essential to EPA but not
needed by the states is not consistent with the two-domain concept that
was incorporated into the plan for revising the RCRA reporting system.7
The concept is based on the premise that the information transferred to
EPA is a subset of the information needed by the states. Despite this rec-
ognition and the known potential for conflict over this point, no system-
atic assessment of the specific differences between EPA and state data
needs or their implications for the RCRA reporting system was conducted.
In addition, the incentive for states to maintain data integrity, enhanced
by state ownership and control of the data (an essential part of the two-
domain concept), is jeopardized if states do not need the data for their
own purposes. EPA officials pointed out that the data to be included in
the implementor domain was subject to negotiation between EPA and
state representatives and thus did not need to be a subset of the data
needed by the states.

“Office of Solid Waste, EPA, “Biennial Report Information System, System Decision Paper I: Concept
Development,” unpublished (November 30, 1987), p. 6.

7The two-domain concept was developed specifically for a new overall data management system,
termed the resource conservation and recovery information system, and then applied to the RCRA
reporting system.



Page 36              GAO/PEMD-90-3     EPA’s Hazardous    Waste Data Need Further    Improvement
Chapter 2
Initial Problems and EPA
Improvement    Efforb




The life cycle documentation did not address the question of whether
appropriate legal authority was in place to support the revised system
or whether additional regulations were needed. Yet authorized states
remain uncertain about whether they have the legal authority to collect
additional data from handlers using the RCRAreporting mechanism. EPA
remains uncertain about whether authorized states can be required to
collect the data deemed necessary in a standard form or provide disag-
gregated data to EPA.

Authorized states are uncertain whether RCRA'Ssection 3007 authority
is applicable to states and whether it is sufficient to enable them to col-
lect the new data included on EPA’S 1987 revised data collection instru-
ment. We discussed this issue with an EPA official in the Office of the
General Counsel. He indicated that section 3007 provides the authority
for EPA and authorized states to obtain any information that handlers
possess as long as it used for the purposes specified in RCRA,although he
acknowledged that this application of section 3007 might have to be
tested in lengthy litigation. However, their authority under section 3007
cannot be used by EPA or authorized states to require that handlers
develop new information in order to complete reporting instruments
(such as performing tests on waste streams that are not already specifi-
cally required by applicable federal or state recordkeeping regulations).

EPA’Splan for the revised RCRAreporting system includes requiring
states to obtain data that are identical to that included on EPA'S revised
data collection instrument (but not to use the instrument itself) and to
submit data to EPA in a disaggregated form. We asked the EPA official in
charge of the RCRAreporting system whether authorized states could be
required to provide uniform disaggregated data to EPA. He expressed
concern about whether EPA could require authorized states to provide
uniform data in a disaggregated form. He stated that EPA was currently
working on this issue but would not have additional authority in place
for the 1989 reporting cycle. As described previously, existing federal
regulations specifically require authorized states to submit summary
reports (rather than disaggregated data) but do not specifically require
authorized states to collect data that are identical to that contained in
EPA'S data collection instrument. Without these requirements, the same
problems of inconsistent measures and definitions across states can
recur in the new, revised system. We believe these legal questions could
and should have been resolved prior to the start of the 1987 reporting
cycle.




Page 37            GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
              Chapter 2
              Initial Prwtdema and EPA
              Improverno    *.Efforta




              Before we began our evaluation, EPA had developed an information sys-
Shnmaxy and   tern that included five data collection mechanisms that provided some
Ctbclusions   information on hazardous waste generation or management. EPA expe-
              rienced specific problems with the individual data collection mecha-
              nisms or with the information system as a whole in four areas: (1)
              system development, (2) information requirements, (3) measurement,
              and (4) data collection, Through 1986, these problems had prevented
              EPA from determining with reasonable precision or certainty how much
              of what type of hazardous waste was generated, how it was managed,
 I            whether sufficient management capacity existed, or whether progress
 I            was being made in reducing hazardous waste generation.

              EPA initiated 13 efforts to improve the hazardous waste information    sys-
              tem that directly affected the quality of its information on hazardous
              waste generation and management. In this chapter, we presented our
              evaluation of EPA efforts to improve the information system develop-
              ment process; efforts aimed at improving specific information system
              components are discussed in subsequent chapters. Four information sys-
              tem development problems have been identified: (1) lack of a compre-
              hensive plan to coordinate data collection efforts; (2) no comprehensive
              evaluation of existing regulatory development data; (3) isolated data
              collection efforts that were narrowly focused and resulted in duplication
              of effort; and (4) data planning and collection responsibilities that were
              divided among program offices, with inadequate integration of data col-
              lection efforts (including cross-media data collection efforts).

              EPA'Sevaluation of data collection activities has fully corrected the sec-
              ond of these problems. However, the improvement efforts we evaluated
              are not likely to prevent the problem of the isolated development of data
              collection efforts needed to ensure integrated data collection that fully
              supports the program mission. No overall data collection plan has been
              developed. The central coordinating office does not have full authority
              to develop data collection efforts and did not use a central prime con-
              tractor for developing systems. The coordinating office had to rely on
              contributions from the budgets of program offices, other offices contin-
              ued to have responsibility for improvement projects, and the tracking
              system does not provide a mechanism for ensuring consistency. These
              factors also contribute to the possibility both of failure to collect needed
              information in some areas and of redundant data collection in others.

              The principal planning mechanism (the new life cycle management sys-
              tem) is a major improvement, but it needs refinement to ensure that it
              accomplishes its intent. In the one applicable case where it was used, the


              Page St3           GAO/PEMDBO-3   EPA’s Haza.rdous Waste Data Need Further   Improvement
                  Chapter2
                  Mtial Problems and EPA
                  Improvement   Efforts




                  analysis and documentation were inadequate given the importance of
                  the system. Important potential data collection alternatives were not
                  considered, the analysis of costs and benefits was inadequate, previous
                  problems with the data collection mechanism were not systematically
                  examined, and specific information needs were not systematically iden-
                  tified. Finally, the legal authority for collecting new information was not
                  clarified.

                  Our findings show that EPAhas established an improved information
                  system development process that is largely consistent with federal
                  guidelines and EPA policy. Our findings also show that further improve-
                  ments are needed. EPA should develop a comprehensive plan for data
                  collection, The life cycle management system should require complete
                  and detailed documentation for major information system components.
                  Placing the final authority for developing information system compo-
                  nents in the central coordinating office, which would have control over
                  the information system development budget and use a prime contractor
                  when appropriate, is an additional step EPA should consider in order to
                  achieve an overall information system that effectively supports the pro-
                  gram mission.

                  EPA officials indicated that in general they agreed with our findings and
                  are already taking steps that at least partially address some of them.
                  They stated that a survey of the states to determine what state officials
                  believe their data needs are has now been completed, that the central
                  coordinating office now has its own budget that is supposed to be ade-
                  quate for system development (although the office may still obtain addi-
                  tional funding from substantive divisions), and that they now have a
                  prime contractor for the RCRAreporting system. We did not, however,
                  evaluate EPA’srecent activities.


                  We recommend that the Administrator of EPA direct that the Assistant
Recommendations   Administrator for Solid Waste and Emergency Response take appropri-
                  ate steps to enhance its information system development process and
                  fully ensure that data collection efforts complement each other and sup-
                  port the program mission. Specifically, a comprehensive data collection
                  plan should be developed. Steps should be taken to improve the assign-
                  ment of responsibilities for planning and directing the development of
                  information system components by increasing the authority of the cen-
                  tral coordinating office to develop data collection efforts and ensure
                  consistency. Finally, the life cycle management system should be refined



                  Page 39           GAO/PEMD-90-S   EPA’s Hazardous   Waste Data Need Further   Improvement
                                                                                        ‘

Chapter 2
Initial Pwblema and EPA
Improveuwrt    Efforts




to ensure the complete and detailed analysis and documentation of each
stage of the cycle for major system components.




Page 40             GAO/PEMD-90-S   EPA’s Hazardous   Waste Data Need Further   Improvement
       .

Chapter 3

EPA Information Needs


                    This chapter addresses our fourth evaluation question: Has EPA identi-
                    fied the information on hazardous waste generation and management
                    that is required by EPAand the states in order to support the program
                    mission? We identify those activities conducted by EPA and the states
                    that need generation and management information, present our assess-
                    ment of the information needed for each activity, and compare these
                    needs to EPA'Snew and revised data collection instruments. The compari-
                    son shows the extent to which the agency has provided for the obtaining
                    of necessary information, including the information that was identified
                    as missing in the original EPAinformation system.


                    EPA'Sneed for information on hazardous waste generation and manage-
Idedtification of   ment is broadly determined by statutory goals, objectives, and require-
Actibities          ments. Specific information needs are a function of the activities
                    conducted by EPA and the states to accomplish this mission. To identify
                    these activities, we first asked each of the 38 potentially relevant EPA
                    branches to enumerate its activities. This resulted in a universe of 6 16
                    distinct activities. We then asked each branch chief to identify the activ-
                    ities that require information on either hazardous waste generation or
                    management. In addition, we reviewed the activities not identified in the
                    second step to guard against oversight and misunderstanding, and
                    reviewed reports and documents that resulted from the activities to fur-
                    ther specify information requirements. This “bottom-up” approach ena-
                    bled us to obtain a detailed understanding of specific program activities
                    and their information needs. We also interviewed directors of key divi-
                    sions with policy responsibility for the development of generation and
                    management information.

                    With respect to state data needs, it was necessary to identify only those
                    activities with consistent information needs across all states, since
                    states and EPA share the results of national data collection efforts for
                    these activities, While authorized states use generation and management
                    information to support many activities, their activities and information
                    needs are not generally consistent. This is so because (1) states may
                    implement federal requirements in various ways as long as the results
                    are consistent with and equivalent to federal requirements and (2) in
                    addition to implementing federal requirements, states may develop addi-
                    tional requirements and establish activities with their own unique infor-
                    mation needs. We included only those activities with consistent
                    information needs across all states because these can be used to form
                    the foundation of a national information system, while activities with
                    inconsistent information needs cannot.


                    Page 41         GAO/PEMD-90-S   EPA’s Hazardous   Waste Data Need Further   Improvement
                                     Chapter fi
                                     EPA Infoh , w tlon Needs




                                     EPA and state agencies have different responsibilities, which are divided
                                     along functional lines. All authorized states have primary responsibility
                                     for implementing the federal program (such as issuing permits and con-
                                     ducting inspections). EPA regional offices assume this responsibility in
                                     unauthorized states and for Hazardous and Solid Waste Amendments of
                                     1984 (HSWA)provisions that become effective in authorized states with-
                                     out the modification of state laws and regulations. EPA headquarters has
                                     the primary oversight responsibility for monitoring the performance of
                                     the state agencies, administering the overall national program (such as
                                     allocating funding to states), and developing federal regulations pertain-
                                     ing to hazardous waste management. The state activities we identified
                                     as having consistent information needs are those implementation activi-
                                     ties required of all states by EPA.

                                     We identified 10 activities performed by all states or EPA that require
                                     periodic collection and maintenance of hazardous waste generation or
                                     management data. Table 3.1 lists the activities we identified in each
                                     functional area and shows the organizational responsibility for each.

Tab14 3.1: Program Activities That
Reql)ire Generation or Management                                                                                         Responsible
Infotjnation, by Function and                                                                                             organization
Responsible Organization             Function and activity                                                              State        EPA
                                     Implementation
                                       1. Prioritize inspections                                                        Xa
                                       2. Prepare enforcement cases                                                     X
                                       3. Provide technical assistance                                                  X
                                     Administration and oversight
                                       4. Develop SARA capacity assurance                                               X            X
                                     Regulatory development
                                       5. Regulatory policy assessment                                                               X
                                       6. Technology assessment                                                                      X
                                       7. Capacity assessment for implementing           land disposal restrictions                  X
                                       8. Integrated capacity analysis                                                               X
                                       9. Risk assessment                                                                            X
                                       10. Regulatory impact analysis                                                                X
                                     aAn “X” indicates the location of primary responsibility




                     Y




                                     Page 42                 GAO/PEMD-90-S       EPA’s Hazardous   Waste Data Need Further   Improvement
                             Chapter 3
                             EPA Information    Needs




Infoimation
Req ’irements for
Imp 7ementation,
Awinistration, and
Ove7’sight Activities

Implfmentation Activities    Implementation activities are those that are conducted to carry out
                             existing state and federal laws and regulations concerning hazardous
                             waste management.’ Our panel of state experts identified five implemen-
                             tation activities that require information on hazardous waste generation
                             or management: (1) prioritizing inspections, (2) preparing enforcement
                             cases, (3) providing technical assistance, (4) assessing fees for hazard-
                             ous waste handlers, and (6) issuing permits for management facilities.
                             Of these implementation activities, assessing fees and issuing permits do
                             not require national information. Fees are not required by the Resource
                             Conservation and Recovery Act of 1976 (RCRA) and are assessed at the
                             state level. Fees are set on the basis of the amounts and types of waste
                             generation and w%te management activities, but there is extreme varia-
                             tion in how states assess these fees. Permits consider facility and waste
                             characteristics but are issued on a facility-specific basis. Consequently,
                             permit issuance and fee assessment do not require information on the
                             regulated population as a whole and therefore do not contribute to the
                             foundation of a national information system.

Prioritizing   Inspections   State agencies do not have the resources to inspect every hazardous
                             waste handler. Therefore, agencies must decide which facilities to
                             inspect. EPA has recently begun to establish general guidelines for states
                             to use in prioritizing inspections. The major priority is environmental
                             significance-that     is, the potential for harm resulting from the activities
                             conducted at each facility. For example, EPA has specifically identified
                             management facilities with incinerators, facilities with land disposal
                             units (especially those nearing closure), and facilities with ongoing cor-
                             rective actions as high priority facilities for the purpose of inspection.
                             EPA has determined that these attributes of management facilities post
                             high risks to human populations and the environment. RCRA also
                             requires annual inspections of commercial facilities receiving wastes

                             ‘To facilitate recordkeeping relevant to all activities, EPA assigns an identification number to each
                             handler. Similarly, all activities require that EPA describe the regulatory status of each handler and
                             the reason for nonregulated status.



                             Page 43               GAO/PEMD-90-3      EPA’s Hazardous     Waste Data Need Further     Improvement
                                Chapter 3
                                EPA Iuforwrtion   Needs




                                defined by the Comprehensive Environmental Response, Compensation,
                                and Liability Act of 1980 (CERCLA),and thus these facilities receive a
                                high priority. Once the priorities are established, states need to target
                                the high priority handlers for inspections. States need information on
                                each handler about the types of management technologies used, the
                                source of wastes (such as whether the waste originates at a CERCLAsite),
                                whether the technology is commercially available, the permit status
                                (interim, final, or nearing closure), whether ongoing corrective actions
                                exist, and whether facilities are government owned.

                                EPA enforcement officials explained that states will also have to priori-
                                tize inspections of generators and management facilities in terms of the
                                environmental significance of large quantities of wastes, particularly
                                hazardous wastes, and wastes that require specific treatments according
                                to the land disposal restrictions. Thus, the states require from each facil-
                                ity information on waste characteristics of sufficient detail to identify
                                the treatment technologies required by the land disposal restrictions.
                                This required information includes the identity of the regulated waste
                                streams (as defined by EPA waste code), information on the physical and
                                chemical form of the waste, and the concentrations of hazardous and
                                nonhazardous constituents that determine the way a waste is treated-
                                that is, its treatability. (See glossary entry for Treatability Analysis.) It
                                also requires information on the presence of other particularly hazard-
                                ous constituents that do not affect the selection of management technol-
                                ogies (including metals and nonmetals). Although state and EPA officials
                                could not identify all these constituents, they would include chemicals
                                such as methylene chloride, a known carcinogen.

Preparing Enforcement   Cases   Information on hazardous waste generation and management is required
                                for prioritizing and preparing enforcement cases. First, state agencies
                                use information on the types of wastes generated and managed to priori-
                                tize their responses to emerging enforcement actions, which may result
                                from public complaints or inspections. For example, if a complaint is
                                received indicating the possibility of illegal dumping at a facility, state
                                agencies use information on the type of wastes handled at the facility
                                (and the degree of hazard they represent) to prioritize the enforcement
                                case. With respect to CERCLAand corrective action sites, enforcement
                                officials identify responsible parties by reviewing information on the
                                origin, volume, and treatability characteristics (as defined by EPA waste
                                code) of specific regulated wastes received at a particular site. Informa-
                                tion on management technologies is also necessary to determine how the




                                Page 44             GAO/PEMD-99-3   EPA’s Hazmdous   Waste Data Need Further   Improvement
                             Chapter 3
                             EPA Information   Needs




                             wastes were handled at the site. The most detailed information neces-
                             sary for this activity is the identity of major hazardous constituents pre-
                             sent in a quantity of waste. This information is necessary to enable
                             states to identify parties that sent waste with specific constituents to
                             particular facilities.

I*ovic$ng Technical&z&tame   States have increasing responsibilities for providing technical assistance
                             to waste generators as a result of the alternative management required
                             by the land disposal restrictions. States are beginning to provide techni-
                             cal assistance to generators in locating appropriate facilities for manag-
                             ing (including recycling) specific types of waste. In order to provide this
                             assistance, states require information on the specific management tech-
                             nologies available at each facility that provides commercial waste
                             management.

                             Table 3.2 summarizes the data needs we identified for each implementa-
                             tion activity and shows that the data requirements across the activities
                             are quite similar. These data are needed on every handler because
                             implementation activities involve all handlers; thus, the use of
                             probability samples would not be possible.




                             Page 48             GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement




                                                          ’             J
    ,
                                          Chapter a
                                          EPA InfwmMion       Needs




Tab& 3.2: Types of information Required
for Ihplementation Activities                                                                                   Activity
                                                                                                              Prepare            Provide
                                                                                            Prioritize        enforcement        technical
                                          Type of information                               inspections       cases              assistance
                                          Regulatory status
                                          EPA ID number                                     Xa                X                  X
                                          Type of status                                    X                 X                  X
                                          Reason for nonregulated       status              X                 X                  X
                                          Wastes and waste characteristics
                                          Each regulated waste stream (as defined           X                 X
                                            by EPA waste code) present in a
                                            quantity of waste
                                          Quantities                                        X                 X
                                          Chemical and physical form for                    X                 X
                                            treatability analysis
                                          Concentrations of hazardous and                   X                 X
                                            nonhazardous constituents for
                                            treatability analysis
                                          Major additional hazardous metals                 X                 X
                                            present
                                          Major additional hazardous nonmetals              X                 X
                                            present
                                          Waste
                                          .-.___ management
                                          Treatment technologies                            X                 X                  X
                                          Storage technologies                              X                 X                  X
                                          Disposal technologies                             X                 X                  X
                                          ._____----
                                          Types                                             X                 X                  X
                                          ~~--___of recycling
                                          Commercial status                                 X                                    X
                                          Permit status                                     X                                    X
                                          Waste   source
                                          ..-..-____
                                          Originating facility                                                X
                                          CEPCLA, corrective action                         X
                                          aAn “X” indicates that information is required.




Administrative and                        Administrative activities are those required to operate EPA or state
Oversight Activities                      agency programs but not directly associated with implementing existing
                                          laws and regulations. Oversight activities are those conducted by EPAto
                                          determine whether the states are implementing the federal program ade-
                                          quately and whether the overall national program is effective. The
                                          activities conducted by EPA or states in these areas that use generation
                                          or management data are (1) allocating resources required for program
                                          activities, (2) monitoring state agency performance, (3) evaluating the
                                          effectiveness of the national program, and (4) developing the capacity



                                          Page 46                 GAO/PEMD-90-3      EPA’s Hazardous   Waste Data Need Further   Improvement
                   Chapter3
                   EPAInformation Needs




                   assurances required from each state under the Superfund Amendments
                   and Reauthorization Act of 1986 (SARA).At the present time, however,
                   only the capacity assurances each state must develop require a national
                   information system. State data needs for allocating resources appear to
                   be minimal and inconsistent across states, while EPA does not use genera-
                   tion or management data for this activity. EPA and most states have not
                   conducted program evaluations. Finally, EPA does not use generation or
                   management information for monitoring state agency performance.

                   SARAcapacity assurance does require consistent national information.
                   EPA is responsible for establishing consistent data and analysis require-
                   ments across all states, but the states are responsible for developing
                   their own assurances. The states require an extensive amount of data to
                   conduct the required analysis. Our panel of state experts indicated that
                   SARAcapacity assurance is the major driving force behind the increased
                   need for information on hazardous waste generation and management.
                   Since these data needs are both extensive and similar to those of other
                   capacity analyses conducted by EPA for developing federal regulations,
                   we discuss the information needs for SARAcapacity assurance in the fol-
                   lowing sections of our report in conjunction with the other capacity
                   analyses EPA conducts to support the development of federal
                   regulations.


                   Developing federal regulations calls for different types of analysis that
Information        use information on hazardous waste generation and management. Five
Requirements for   types of analysis are used in the process of developing federal regula-
Regulatory         tions, although not all these types are required in every case: (1) regula-
                   tory policy assessment, (2) capacity assessment (land disposal and
Development        integrated), (3) risk assessment, (4) technology assessment, and (5) reg-
Activities         ulatory impact analysis. Usually, regulatory policy analysis is a qualita-
                   tive consideration of legal and regulatory issues that does not require
                   generation or management information. However, the assessment of reg-
                   ulatory policy concerning waste minimization does require generation
                   and management data. Capacity assessments, policy assessment for
                   waste minimization, and risk analyses require extensive generation and
                   management information. Technology assessment, used to identify best
                   demonstrated available technologies, requires less extensive data, which
                   are fully provided by the data needed for capacity assessment and risk
                   assessment. The cost benefit analyses required for the regulatory impact
                   analysis under Executive Order 12291 use the same generation and
                   management data as those required for risk assessments. Thus, from an
                   information requirements perspective, the essential types of analysis to


                   Page47           GAO/PEMD-90-3
                                               EPA’sHazardousWasteData NeedFurther Improvement
                           Chapter 3
                           EPA Inforr,w   lion Needs




                           examine are capacity assessment, waste minimization, and risk assess-
                           ment. Since waste minimization data are also required for capacity
                           assessment, we discuss them together. In the following section, we pro-
                           vide a detailed discussion of their data needs.


Cagjacity Assessment and
Wa$te Minimization
Capacity Assessment        Three types of capacity assessment are currently conducted. First, HSWA
    I                      requires EPA to determine whether sufficient capacity exists to imple-
                           ment the land disposal restrictions required by HSWA.If sufficient capac-
                           ity does not exist, the regulations must be postponed. Second, EPA is
                           conducting an integrated capacity analysis to determine the effects of
                           proposed regulations on waste generation and management capacity
                           over the next three to five years. This is an internal analysis that is used
                           in risk assessments and regulatory impact analyses but will not result in
                           a specific published report. Finally, SARArequires each state to assure
                           that sufficient capacity will exist to manage the hazardous wastes pro-
                           duced in the state for the next 20 years. In each case, whether sufficient
                           capacity exists is determined by comparing the amount of capacity
                           required to manage hazardous wastes to the amount available for that
                           purpose. We contrast the data needs for each type of capacity analysis
                           according to the three major stages of any capacity analysis: (1) the
                           determination of required capacity, (2) the determination of total capac-
                           ity, and (3) the determination of whether available capacity is
                           sufficient.

                           The Determination of Required Capacity. This involves two basic steps:
                           (1) identification of the volumes of waste covered or affected by the
                           capacity assessment and (2) analysis of the treatability of the affected
                           waste and determination of the required capacity for each type of man-
                           agement technology.

                           It is first necessary to determine the volume of waste covered by the
                           analysis. For the land disposal restrictions, these are termed “affected
                           wastes” because only certain wastes (as designated by EPA waste codes)
                           are affected by each requirement. This requires data on the volume of
                           each waste stream currently managed that would be affected by the
                           land disposal restrictions.




                           Page 48               GAO/PEMDdO-3   EPA’s Hazardous   Waste Data Need Further   Improvement
,



    Chaptm3
    EPAInformation   Needs




    Only the proportion of the affected waste streams that is actually land-
    disposed is subject to the requirements. This includes residuals from
    prior treatment (secondary generation) of affected waste streams and
    agents added during management, as well as wastes that were land-dis-
    posed without prior treatment (primary generation). Thus, information
    on the volume of each affected waste stream disposed of by each cur-
    rent land disposal technology is essential. Land disposal technologies
    include landfills, surface impoundments, and waste piles.

    Finally, since some affected wastes are already treated to a level that
    would meet the treatment requirements, it is necessary to determine the
    proportion of land-disposed, affected wastes that already meet the
    treatment standard. This determination requires information on each
    type of management technology the affected waste streams were sub-
    jected to prior to land disposal, in order to identify the proportion that
    would not have met the requirement.

    The analysis for the land disposal restrictions requires a “snap shot”
    analysis. Information on the volumes of wastes that will need to be
    treated in the foreseeable future is not required. Thus, information on
    currently managed CERCLA,corrective action, and closure site wastes is
    required, but information on the total volume of these wastes that will
    ultimately require treatment is not.

    The integrated capacity analysis must identify the volumes of primary
    and secondary generation of all wastes from all sources that are cur-
    rently treated and that will require treatment in the foreseeable future.
    Thus, the integrated capacity analysis requires information beyond that
    required for implementing the land disposal restrictions. In addition to
    currently managed wastes, the analysis requires information on quanti-
    ties of waste at CERCL.Asites, corrective action sites, and closure sites
    that will require treatment over the time frame of the analysis. The inte-
    grated capacity analysis also requires information on the volumes of
    waste that will become subject to management requirements as a result
    of new regulations currently under development. Additionally, informa-
    tion is needed on potential forms of waste management-including
    underground vaults, salt domes, and other geological formations used
    for disposal-that prevent the migration of wastes.

    Finally, the integrated capacity analysis requires information on the
    extent of waste minimization. This information is necessary to deter-
    mine the impact that waste minimization efforts will have on the volume



    Page 49             GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Purther   Improvement
Chapter 3
EPA Infwwation   Needs




of waste that will require additional capacity. We discuss the data needs
for waste minimization in the next section of our report.

SARAcapacity assurance requires all the information necessary for the
integrated capacity analysis. In addition, it requires information on
state-level imports and exports in order to determine the quantities of
waste managed within each state, the volumes of wastes exported to
specific states, and the volumes imported from specific states.

The second major stage of the analysis of required capacity, termed
treatability analysis, is the same for all types of capacity analysis. The
purpose of the treatability analysis is to determine the required capacity
of each type of management technology by assigning each quantity of
affected waste to appropriate management technologies. In this way, the
required capacity of each type of technology is defined.

The type or types of treatment technology required by a specific volume
of waste is determined by its physical and chemical characteristics,
including the presence and concentration of some hazardous and some
nonhazardous constituents. Treatability analysis sorts wastes into
groups based on these characteristics and assigns the volumes to spe-
cific treatment technologies or combinations of technologies. In order to
perform the analysis, information is required on the physical and chemi-
cal characteristics of each volume of waste, including the concentration
of some hazardous and nonhazardous constituents that determine appli-
cable treatment technologies. Technical reports prepared for EPA list
approximately 60 characteristics that are used to determine accurately
the type of treatment technology applicable to a given quantity of
waste.

Determination of Total Management Capacity. This stage is the same for
each of the three types of capacity assessment. It requires information
on (1) current total capacity and (2) future total capacity.

The determination of current total capacity requires three types of
information: (1) the capacity of individual units of equipment, (2) the
management technology or technologies used in each unit, and (3) sche-
matic diagrams showing how the units are linked together into process
systems. A process system is a number of linked units performing one or
more management technologies in series. The analysis also requires
information on those units of equipment that are shared by one or more
process systems. This requires detailed design and operational informa-
tion on unit process capacities (including ancillary equipment),


Page 50             GAO/PEMD-903   EPA’s Hazardous   Waste Data Need Further   Improvement
Chapter 3
EPA Information   Needa




throughput operations, and amount of downtime per operating period.
This information allows the capacity of each management technology
and each system to be calculated.

The determination of future total capacity requires information on
future plans and an assessment of regulatory changes. Current manage-
ment facilities may plan to increase or decrease total capacity, and par-
ties that do not currently manage hazardous waste may plan to.
Moreover, these plans are themselves contingent on regulatory changes.
In addition, some types of equipment can be quickly converted to
another management technology.

The analysis of future changes also requires an assessment of the
effects of regulatory changes. For example, changes in minimum tech-
nology requirements have led to the closure of many surface impound-
ments, which decreases total hazardous waste management capacity.
The location standards and other minimum technology requirements
may also decrease total management capacity. Information on design
characteristics of management units and permit status is required to
determine the capacity that currently meets each alternative proposed
standard.

Determination of Available Capacity. Available capacity is the differ-
ence between the total and the utilized amount of management capacity.
Utilized capacity includes that portion used by nonhazardous wastes
managed in hazardous waste facilities and by agents added to hazardous
waste during management (such as reagents and stabilizers). Hazardous
waste management capacity is sufficient if available capacity is equal to
or greater than required capacity.

For the land disposal restrictions, available capacity is calculated by
first determining the proportion of the total capacity of each alternative
technology that is already utilized for the treatment of other hazardous
and nonhazardous wastes. Then, the amount of waste restricted from
land disposal is apportioned to determine if sufficient capacity exists to
implement the regulation, The analysis must be conducted on a facility
basis for on-site and captive systems (that is, those that are not commer-
cially available), Any portion of the affected waste that exceeds the
capacity available at that site or those of subsidiaries must be assigned
to commercial facilities. The data must be of sufficient detail to identify
captive facilities (owned by the same company) and units of equipment
that are partly or completely available for commercial waste
management.


Page 61              GAO/PEMDQO-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                    Chapter 3
                    EPA Information   Needs




__-..
   _.___-
                    For the integrated capacity analysis, the changes in total capacity of all
                    types over the time frame of the analysis are compared to the changes in
                    total demand from all sources (including nonhazardous waste and
                    agents added during management) for the same period. The analysis
                    must also avoid including capacity that is only available to a single firm
   I                or a limited number of firms with capacity that is available to any firm.

                    SARAcapacity assurance requires the same comparison as the integrated
                    capacity analysis discussed previously, except that it is required on an
                    individual state basis rather than a national basis. This means that in
                    addition to the analytic techniques already discussed, the analysis
                    requires information on the origins and destinations of the types and
                    amounts of wastes imported into and exported from each state. This
                    information is required to show whether each state has within its bor-
                    ders sufficient capacity to manage the waste imported into the state and
                    the waste both generated and treated within its borders. This informa-
                    tion is also required as a basis for agreements with other states concern-
                    ing imports and exports.

Vu’hstk!
      Wnimization   EPA is currently assessing existing waste minimization efforts in order to
                    establish national policy-that   is, whether and what type of regulations
                    may be needed. As previously discussed, data on waste minimization are
                    also required as a component of some capacity assessments in order to
                    project future demand for waste management capacity. The information
                    requirements for assessing waste minimization efforts can be divided
                    into two categories: (1) the extent and (2) the determinants of waste
                    minimization.

                    Extent. To assess the extent of waste minimization that has been
                    achieved requires knowledge of both the absolute and relative change in
                    the quantity and toxicity of waste generated. Information on the abso-
                    lute reduction requires data on the total quantity and toxicity for at
                    least two years to observe any change. An important consideration here
                    is that the quantity of waste may be reduced by removing water or
                    other harmless constituents while the quantity of hazardous constitu-
                    ents remains the same, thus increasing the toxicity of the waste. Sound
                    information on the change in toxicity requires detailed quantitative
                    information on the change in concentration of each hazardous
                    constituent.

                    Since the overall volume of waste generated varies with the volume of
                    useful production, information on the absolute change in waste genera-
                    tion alone is insufficient. It is also necessary to know the relative change


                    Page 62             GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
Chapter 3
EPA Information   Needs




in waste generation per unit of production. This requires information on
the quantity of production in each year. However, the types of units
produced may result in different quantities of hazardous waste. There-
fore, useful information on relative waste generation requires data on
the amount and toxicity of the waste generated for each type of product
in each year.

Determinants. Information on the determinants of waste minimization is
necessary to assess areas where interventions might further reduce
waste generation, In addition to the waste stream (as defined by EPA
waste code) involved, this requires detailed information on the charac-
teristics of the waste and the production process that generated it
because the feasibility of waste minimization varies with the type of
waste and the specific industrial process. Waste minimization practices
are also more feasible for recurrent industrial hazardous waste genera-
tion than for other sources of waste. Additional sources of waste for
which waste minimization practices are less applicable include one-time-
only generation (such as decommissioned equipment that is contami-
nated); off-specification chemical products that require disposal; and
wastes from CERCLAsites, corrective action sites, closure sites, or those
produced by the cleanup of leaking underground storage tanks.

Table 3.3 summarizes the information we identified as needed for capac-
ity assessment and waste minimization policy assessment. The just-con-
cluded discussion shows that the information needed for capacity
assessment and waste minimization policy is more extensive than that
required for implementation activities. However, since the information
is only used to characterize the regulated population and is not needed
for every handler, sample surveys are possible.




Page 53              GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                                           chapter 3
                                          EPA InOwmat.ion     Needs




Ta$e 3.3: Types of Information Required
for /Capacity Analyres and Waste          Type of information                                      SARA’        lNTEGb    HSWA=       WMPAd
Mitiimization Policy Assessment           Reaulatorvw status
                                          EPA identification number                                Xe           X         X           X
                                          Type of status                                           X            X         X           X
                                          Reason for nonregulated status                           X            X         X           X
                                          Wastes and waste characteristics
                                          Each regulated waste stream (as defined by EPA           X            X         X           X
                                            waste code) present in a quantity of waste
                                          Quantities                                               X            X         X           X
                                          Physical form data for determining treatability of       X            X         X           X
                                            wastes
                                          Chemical characteristics for treatability analysis       X           X          X           x          -
                                          Concentration of hazardous and nonhazardous              X           X          X           X
                                            waste for treatability analysis                                                                      -
                                          Waste management
                                          Treatment technoloaies                                   X           X          X
                                          Storage technologies                                     X           X          X
                                          Current disposal technologies                            X           X          X
                                          Potential disposal technologies (geological              X           X
                                              formations)
                                          Types of recycling                                       X           X          X
                                          Residual waste generation                                X           X          X
                                          Type of equipment                                        X           X          X
                                          Ancillarv eauipment                                      X           X          X
                                          Capacity of each unit of equipment                       X           X          X
                                          Design characteristics affected by proposed              X           X          X
                                              reaulations
                                          Svstem diagrams                                          X           X          X
                                          Type of management system (technologies and              X           X          X
                                              equipment)
                                          Capacity of management system                            X           X          X
                                          Quantity hazardous waste managed by each                 X           X          X
                                              technology or system of technologies
                                          Quantity of nonhazardous waste managed by                X           X          X
                                              each technology or system of technologies
                                          Quantity of agents added during management           X               X          X
                                          Planned capacity changes
                                          ~--                                                  X               X          X
                                          Commercial status                                    X               X          X
                                          Permit status                                        X               X          X
                                          Imports and exports
                                          Originating facility identification                  X
                                          Destination facilitv identification                  X
                   Y

                                          Waste source
                                                                                                                                   (continued)




                                          Page 54                GAO/PEMD-90-3   EPA’s Hazardous       Waste Data Need Further   Improvement
                  Chapter 3
                  EPA Information     Needs




                  Type of information                                      SARA’       INTEGb     HSWA=      WMPAd
                  Routine industrial oroduction                            X           X                     X
                  One-time-only generation, including                      X           X                     X
                    decommissioned equipment, off-specification
                    product, CERCLA, corrective action, closure,
                    and other remedial action
                  CERCLA volumes requiring management                      X           X                     X
                  Corrective action volumes requiring management           X           X                     X
                  Additional sources affected by pending                   X           X
                    regulations
                  Waste minlmlzation
                  Waste stream affected                                    X           X                     X
                  Specific industrial orocess                              X           X                     X
                  Total waste volume change                                X           X                     X
                  Total production change                                  X           X                     X
                  Soecific oroduct chanae                                  X           X                     X
                  Specific waste volume channe                             X           X                     X
                  Concentration of each hazardous constituent                                                X
                  %ARA = SARA capacity assurances

                  blNTEG = integrated capacity analysis

                  CHSWA - HSWA capacity analysis for land disposal restrictions
                  dWMPA = Waste Minimization Policy Assessment
                  eAn “X” indicates that information is required.


Risk Assessment   EPA currently uses comparative risk assessment techniques to analyze
                  potential regulatory requirements by determining whether land disposal
                  is actually more hazardous than the demonstrated alternatives. For
                  example, air emissions from incineration of a waste could be more haz-
                  ardous than the soil and groundwater contamination from land disposal.
                  EPA'Scurrent use of risk assessment, however, has a limited role in risk
                  management decisions. Continued land disposal of untreated waste
                  would not be permitted even if all demonstrated alternatives to land dis-
                  posal were identified as more hazardous. As already mentioned, the data
                  requirements for risk assessment also provide the data needed for the
                  remaining regulatory development activities, including technology
                  assessments and regulatory impact analyses.

                  Complete risk assessments generally consist of four analytic stages: (1)
                  hazard identification, (2) dose-response estimation, (3) exposure assess-
                  ment, and (4) risk characterization. Basically, the first two stages deter-
                  mine the probability of an adverse health or environmental incident
                  from exposure to precise amounts of substances identified as hazardous.



                  Page 55                 GAO/PEMD-30-3      EPA’s Hazardous   Waste Data Need Further   Improvement
Chapter iI
EPAInfw~ation   Needs




Exposure assessments determine the amount of the hazardous sub-
stances to which populations are actually exposed. The final stage char-
acterizes the risk of individuals or populations based on the work
completed in the first three stages.

Information on the generation and management of hazardous waste
must be collected and maintained by EPA to support the exposure assess-
ment portion of these risk assessments. The other risk assessment stages
rely on different types of data. In order to estimate exposure, EPA must
know the volumes of waste that are land-disposed and the volumes that
would be treated by each alternative technology or combination of tech-
nologies for each proposed alternative regulation. In addition to the
information needed for capacity analysis, more detailed data on the
characteristics of the wastes are required. Information is required on
waste chara.cteristics that determine the mobility or rate of migration of
hazardous constituents through a medium and on the concentration of
each hazardous constituent. Whereas some chemical and physical form
parameters (such as solubility) that determine the appropriate treat-
ment also affect mobility, EPA documents note that additional parame-
ters (such as the rate of biodegradation) are also required. While
knowing the concentration of some hazardous and nonhazardous con-
stituents is necessary to determine the appropriate alternative treat-
ment, the concentration of -all hazardous constituents is required for
complete exposure assessments. When EPA does not have the necessary
data, assumptions based on engineering judgments are used to complete
the exposure analysis.

In addition to the amounts and characteristics of the wastes involved,
releases to the environment and subsequent exposure are determined by
characteristics of management technologies and equipment. This
requires much of the same data necessary for the capacity assessment
because it is necessary to assign the wastes that are currently land-dis-
posed to types of alternative technologies. In addition, detailed data are
required on the design of equipment and materials of construction in
order to estimate the magnitude of releases from different types of
equipment under different regulatory scenarios. Information on the
types of monitoring methods are also required since releases are
affected by the stringency of monitoring.

Finally, detailed site-specific geohydrologic and other environmental
data pertaining to waste management are necessary for determining the
speed and concentration of wastes that move through different environ-
mental media and ultimately reach human populations. These data


Page 56           GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
.
    Chapter 3
    EPA Information   Needs




    include, for example, actual releases, the types of soil underlying facili-
    ties, the height of seasonal water tables, groundwater flow rates, lati-
    tude and longitude, and the distance of waste handling areas from the
    nearest residence, surface water, wells, or property boundary. These
    data are used in conjunction with fate and transport models to estimate
    likely human and environmental exposures.

    Table 3.4 summarizes the data requirements for risk assessment and the
    additional remaining regulatory development activities. The foregoing
    analysis shows that risk assessments, which also support cost benefit
    analyses, require extensive information in addition to that required for
    capacity assessments. Because these data are used to characterize
    aspects of the regulated population as a whole, probability sample
    surveys are possible (since information is not necessarily required for
    every handler).




    Page 57             GAO/PEMB90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                                           Chapter 3
                                           EPA Infcwmation   Needs




Table 3.4: Types of information Required
fOr’Ri8k ASSeSSment and Remaining                                                                                              Regulatory
Rebulatory Development Analysis                                                                 Technology     Risk            impact
                                           Type of information                                  assessment     assessment      analysis
                                           Regulatory status
                                           EPA identification number                            Xa             X               X
                                           Type of status                                       X              X               X
                                           Reason for nonregulated status                       X              X               X
                                           Wastes and waste characteristics
                                           Each regulated waste stream (as defined by           X              X               X
                                             EPA waste code) present in a quantity of
                                             waste
                                           Quantities                                                          X               X
                                           Physical form data for determining treatability      X              X               X
                                             of wastes
                                           Additional physical form data for determining                       X               X
                                             mobility of wastes
                                           Chemical characteristics for treatability analysis   X              X               X
                                           Additional chemical characteristics for                             X               X
                                             determinina mobilitv
                                           Concentration of hazardous and nonhazardous          X              X               X
                                             waste for treatability analysis
                                           Concentration of all hazardous constituents                         X               X
                                           Waste management
                                           Treatment technologies                               X              X               X
                                           Storage technologies                                                X               X
                                           Current disposal technologies                                       X               X
                                           Potential disposal technologies (geological                         X               X
                                               formations)
                                           Types of recycling                                                  X               X
                                           Residual waste generation                                           X               X
                                           --..-
                                           Tvpe of eauioment                                                   X               X
                                           Ancillary ___-
                                                       equipment                                               X               X
                                           Capacity of each unit of equipment                                  X               X
                                           Design characteristics affected by proposed                         X               X
                                               regulations
                                           Design characteristics for estimating releases                      X               X
                                           Types of monitoring methods                                         X               X
                                           Construction material                                               X               X
                                           System diagrams                                                     X               X
                                           Type of management system (technologies and                         X               X
                                           - equipment)
                                           Capacity of management system                                       X               X
                                           Quantity hazardous waste managed by each                            X               X
                                               technology or system of technologies
                                                                                                                                (continued)




                                           Page 58               GAO/PEMD-90-3    EPA’s Hazardous    Waste Data Need Further   Improvement
    Chapter 3
    EPA Information     Needs




                                                           Technology      Risk
    Type of information                                    assessm6irt     assessment      analysis
    Quantity nonhazardous waste managed by                                 X               X
      each technoloav or svstem of technoloaies
    Quantity of agents added during management                             X               X
    Planned capacity changes                                               X               X
    Commercial status                                      X               X               X
    Permit status                                                          X               X
    CSeohvdrolonical and environmental data
    Soil types                                                             X               X
    Groundwater flow rates                                                 X               X
    Proximitv to water source                                              X               X
    Height of water table
    ---                                                                    X               X
    Di;Flce    from management       units to property                     X               X

    Distance from property line to nearest potential                       X               X
      human exposure
    Actual releases to environment                                         X               x
    Latitude and longitude                                                 X               X
    Waste source
    koutine                                                                X               X
    --____I_ industrial production
    Soecific industrial orocess                                            X               X
    One-time-only generation, including                                    X               X
       decommissioned equipment, off-specification
       product, CERCLA, corrective action, closure,
       and and other remedial action
    CERCLA volumes requiring management                                    X               X
    Corrective action volumes requiring                                    X               X
       management
    _--.--
    Additional sources affected bv_ pendinn
                                    .                                      X               X
       regulations
    Waste minimization
               ___-
    haste stream affected                                                  X               X
    Total  waste volume change                                             X               X
    --______-
    Total
    -.--- production change                                                X               X
    Specific product change                                                X               X       .__
    -.--___..
    Specific waste volume change                                           X               X
    Change in concentration/amount         of each                         X               X
        hazardous constituent
    aAn “X” indicates that information is required


    The foregoing analysis shows that data with different levels of detail,
Y   which correspond to different sets of activities, are needed. The least
    detailed level of data is required for implementation activities, but this
    information is needed for every handler because the information is used


    Page 59                GAO/PEMD-90-3       EPA’s Hazardous   Waste Data Need Further   Improvement
                       Chapter 3
                       EPA Information   Needs




                       to monitor individual handlers. The data required for capacity analyses
                       are much more extensive than those required for implementation activi-
                       ties. However, sample surveys are possible because the information is
                       needed to characterize aspects of the regulated population, not to moni-
                       tor each handler. Sample surveys are also possible for risk analyses, the
                       most detailed level of necessary information, for the same reason. We
                       discuss the feasibility and design of sample surveys further in chapter 5
                       when we discuss data collection issues.


                       Our first concern was to determine whether the EPA improvements
Comparison of          addressed the information gap problems identified in the original sys-
Ilhformation Needs     tern. Consequently, we compared the four new and revised EPA data col-
With Data Collection   lection instruments to the data gaps identified in chapter 2 to determine
                       if they had been closed. To determine whether EPA had provided for the
Itistruments           collection of the information we have identified through our require-
                       ments assessment, we compared the results of our assessment with the
                       four new or revised data collection instruments. We defined a data gap
                       as any situation in which EPA had not made provisions to collect the data
                       that we identified as necessary.


Overall Assessment     The four new or revised data collection instruments developed by EPA
                       include those developed for (1) the revised RCXA reporting system, (2)
                       the generator survey, (3) the management facility survey, and (4) the
                       toxic chemical release inventory reporting system. Table 3.5 combines
                       the information needs we identified in the above information require-
                       ments analysis across all activities and indicates the information that is
                       provided by each of the new or revised data collection instruments. The
                       information included on the 1985 RCRAreporting system instrument is
                       included as a baseline.




                       Page 60             GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                                                        Chapter 3
                                                        EPA Information     Needs




Table 315: Comparison of Information Needs With the Information Gathered by the New and Revised Data Collection Instruments,
Using t$e 1985~RCRA Reporting Instrument as a Baseline.
       /                                                                                 Data collection instrumenta         1.
Type 01 information                                                               1985    1987       TSDR      GEN      TCRI
Regula ory status                   __..-.__I__.-.                                                             ___--.
       t
EPA idehtification   number                                                                              Xb          X          X            X             X




                                                                          present in a quantity of       X           X          X            X
                                                                                                                                         ___.-.-        ~~---~
                                                               ___.-_                                    X           X          X            X
                                                                                                         c           X          X            X             X
                                                                                                                                             X
                                                                                                         c           X          X            X
Additional chemical characteristics for determining mobility
                                                       --..                                                                                  X
Concedtration of hazardous and nonhazardous waste for treatability                                       c           X          X            X
                                                                     .___~analysis
Major additional hazardous metals present             ..__--- --.-.                                                  X          X            X
Major-additional hazardous nonmetals present            ---        ______-                                           X          X            X
Concer$ration/amount    of all hazardous chemicals present                                                                                   __--._        x
                                                                                                                                                         -~~
Managpment data                           --                                                                                           .-__~~     ..- ---~-
Treatment technologies                                                                                   X           X          X           X               x
Storage technologies                                                                                     X           X          X           ...~_ .._ - ~~ X-
                                                                                                                                            X
Curreni drsposal technologies                                                                            X           X          X      ---- X            ~~-X
                                                                                                                     d          d            d             (1
PotentiLl disposal technologies (geologic formations)
Types of recycling         ~-                                                                                        X          X            X __---~      X
Residual waste generation                                                                                            X          X            X
Type oi equipment                                                                                                               X            X
                                                                            -__
Ancrllary .equ;pment
Capacrty of each unit of equipment .-- .._.__- .-_.--____.                                               c                      X               X
                                                                                                             -
Design’characteristics     affected-by proposed regulations                     ___                                             X               X
Desrgn: characteristics for estimating releases                           .-___-                                                    ..~.._. X          ~~
Type of monitoring methods                                                                                                      X               X
Constrkon       material                                                                                                        X               X
System diagrams                                               ..--.. .---                                                       X    -___       X ~     ~~.~~~
Type oi management system (technologies and equipment)                                             -                 X          X               X
Capacity of management system                                                                            c           X          X               X ..~~_ .~
                                                                                                                                      .I_
Quantky’                                                                                                                        X          ___.X
        + .hazardous-_-...
                        waste. managed        by each--...technology or system 07 technologies
                                -. .._.----- --.._.-                                                     c           X                                ~~~     -
Quantity nonhazardous waste    -_--- ._.._- .--___-by each
                                      managed                  technology or system of technologies                  X          X               X
                                                           -_--_.
Quanti’ties of agents added during management                                                                                   X         -..__ X ..-~ ~--.
        r                   P--- ---------.- --- .-.
                                                                                                                                                      (continued)




                                                        Page 61                GAO/PEMD-99-3       EPA’s Hazardous   Waste Data Need Further       Improvement
                                                 Chapter 3
                                                 EPA Information    Needa




-
                                                                                                              Data collectlon instrumenta
Tvbe of information                                                                                  1995      1987      TSDR       GEN         TCRI
Plahned capacity changes                                                                                      X          X          X
Commercial status                                                                                             X          X          X
Perjmit status          -~~~ ..---                                                                            X          X          X
Imbts        and exDorta
Oribinating facility identification                                                                  X        X
Destination facility identification                                                                  X        X
Gebhydrologic and environmental data
Soil types                                                                                                                          X
                                                                                                                                    X
                                                                                                                                    X
                                                                                                                                    X
                             units to property line                                                                                 X
Didtance from property line to nearest potential human exposure                                                                     X
Actual releases to environment                                                                                                      X
La&de and longitude                                                                                                                 X
Wbte source
Routine industrial production                                                                                 X          X          X
O&-time-only     eneration, including decommissioned equipment, off-specification                             X                     X
  product, CE w CLA, corrective action, closure, and other remedial action
Total CERCLA and total corrective action volumes that will require management                                 d          a          d           d

                                                                                                              d          d          d           d
Additional sources affected by pending regulations (underground storage tanks)
WEiste minimization
Waste stream affected ~-. ~~~ ..-~                                                                   c        X                     X
Spkcrfic industrial process                                                                          c        X                     X
Waste volume change                                                                                  c        X                     X
Production change                                                                                    c        X                     X           X
Chanae in concentration/amount  of each hazardous constituent                                        c                              X           X
                                                3985 = 1985 RCRA reporting instrument
                                                1987 = 1987 RCRA reporting instrument
                                                TSDR = National Survey of Management Facilities (TSDR survey)
                                                GEN = National Survey of Generators
                                                TCRI = toxic chemical release inventory reporting instrument
                                                bAn “X” indicates the presence of the required information.
                                                ‘Information gap orginally identified in chapter 2
                                                dRemaining information gap

                                                In chaDter 2, four information gaps were identified in the original EPA
                                                information system: { 1) physic&l &id Ch6WYkl ~am\~g~~eflt cRaracWi$-
                                                tics of a waste, (2) amount of waste actually managed or diapos&i 6f By
                                                each management technology or series of technologies, (3) available
                                                management capacity, and (4) the extent of waste minimization. Each of
                                                these gaps is indicated in table 3.6 by a superscript C. There are more


                                                Page   02              GAO/PEMD-90-3       EPA’s Hazardous    Waste Data Need Further   Improvement
                        Chapter 3
                        EPA Information   Neede




                        than four information types thus indicated because knowledge of more
                        than one attribute is necessary to fill each gap, Table 3.5 shows that EPA
                        has eliminated all of the gaps that were identified in chapter 2. Table 3.5
                        also shows that the vast majority of the information we identified as
                        necessary in our complete information requirements analysis is
                        addressed in one or more of the data collection instruments.2 If the infor-
                        mation is included on one or more of the new or revised instruments, EPA
                        has identified the required information.

                        Table 3.5, which is based on our complete information requirements
                        analysis, also shows that some information we identified as necessary is
                        not included in any of the data collection efforts. Each area where an
                        information gap remains is indicated by a superscript D in table 3.5. The
                        three areas where information is lacking are (1) the quantities and types
                        of waste present at CERCLAand corrective action sites that will ulti-
                        mately require treatment and secure disposal at RCRAregulated facili-
                        ties, (2) the quantities and types of waste from the cleanup of leaking
                        underground storage tanks that will be subject to the requirements of
                        Subtitle C under the expanded organic toxicity regulation, and (3) the
                        management capacity of salt domes and other geological formations.
                        Complete information in these areas is not available in any other
                        existing sources, will be difficult to obtain, and will seriously jeopardize
                        the achievement of important national objectives until it is obtained.


Implications for the    The lack of information in these areas in the interim information system
Interim and Permanent   will primarily affect SARAcapacity assurances, EPA'S integrated capacity
                        analysis, and EPA regulatory development activities, including risk
InfoSmation Systems     assessments and regulatory impact analyses. As discussed previously,
                        this information is necessary to fully determine whether sufficient
                        capacity will be available in the foreseeable future and the risks, costs,
                        and benefits posed by alternative regulatory approaches. In addition,
                        the planned permanent information system contains no mechanism for
                        providing this information.

                        Currently, the primary source of data on CERCLA site wastes is the
                        records of decision developed immediately before the start of a cleanup.
                        Records of decision are agreements between the federal government and
                        the other relevant parties concerning a specific site. These documents,

                        “The four initial information gaps were identified before our complete information requirements anal-
                        ysis was conducted. Therefore, there could have been-and in fact were-more information gaps
                        than were identified in chapter 2.



                        Page 63               GAO/PEMD-90-3     EPA’s Hazardous    Waste Data Need Further    Improvement
Chapter 3
EPA Inkmation   Needs




produced in the final stage of investigating sites for cleanup, contain
estimates of the volumes of waste to be managed and the management
technologies required. However, records of decision only exist for 100 of
the 890 sites on the National Priorities List, and this list does not include
additional sites that states have designated for cleanups. Other data are
available for many additional sites from other stages of the investiga-
tion of specific sites, such as the regulatory investigation and feasibility
study stage.

Adequate data on the volumes of waste from corrective action sites that
will need treatment are also lacking. The only data available for correc-
tive action sites are found in corrective measure studies. Similar to
records of decision, these studies are only available for a limited number
of potential corrective action sites.

According to the EPA official we interviewed on this subject, wastes gen-
erated from the cleanup of underground petroleum storage tanks are not
currently regulated as hazardous wastes. Accordingly, no effort has
been made to assess the volumes of soil contaminated with substances
such as benzene and toluene. The revised organic toxicity regulation
under development at EPA, however, will capture much of these wastes
as hazardous waste. Although the process would be difficult and expen-
sive, the volumes of these wastes could be estimated by a survey.

No data currently exist on the potential capacity of geological forma-
tions that could be used for waste disposal. If it can be shown that waste
will not migrate from these formations, their use could help alleviate
potential capacity shortfalls. EPA officials stated that one geologic for-
mation (a demonstration project operated by the Department of Energy)
is currently receiving low level radioactive wastes mixed with hazard-
ous waste, that permit applications have been submitted for two geologi-
cal formations to operate as disposal facilities with no migration
variances (see glossary), and that there are numerous other promising
geological formations. The potential capacity of these geologic forma-
tions could be estimated on the basis of a sample of known formations.

The data just discussed could be used to significantly improve EPA'S
understanding of the amounts of waste that will require management in
the future and when it is likely that these wastes will need treatment.
However, we recognize that it is not possible to develop information on
future waste management needs with the same precision as is possible
for currently generated wastes and present management practices,



Page 64           GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                   Chapter3
                   EPAInformationNeexb




                   Our analysis shows that EPA has extensive needs for hazardous waste
                   generation and management information. The information requirements
                   can be divided into three groups based upon the amount of detail
                   required.

                   All the data gaps identified in the original information system are filled
                   by one or more of the new or revised data collection instruments. The
                   vast majority of the additional data needs we identified have also been
                   addressed by one or more of the instruments developed by EPA. This rep-
                   resents a significant accomplishment by EPA. However, we also found
                   that three information requirements have not been satisfied. Without
                   this information, EPA regulatory development research and national
                   capacity analyses, including SARAcapacity assurances, will be signifi-
                   cantly flawed.


                   We recommend that the Administrator of EPA direct the Assistant
Recc)mmendations   Administrator for Solid Waste and Emergency Response to take appro-
                   priate and feasible steps to fill remaining information gaps, including (1)
                   the volumes of waste located at CERCLAand corrective action sites that
                   will ultimately require management capacity, (2) the volumes of waste
                   that will require management capacity under proposed regulations
                   (including the large volumes of waste expected from the cleaning up of
                   leaking underground storage tanks), and (3) the potential disposal
                   capacity of salt domes and other geological formations that are capable
                   of preventing the migration of wastes.




                   Page65                GAO/PEMD-9S3EPA'sHazardousWasteDataNeedF'urtherImprovement
Chapter 4

Assessmentof MeasurementInstruments


             This chapter addresses our fifth evaluation question: How well do EPA'S
             measurement instruments actually measure the relevant attributes? We
             evaluated the measures used in the four new and revised instruments in
             order to determine whether they are likely to produce the needed infor-
             mation identified in chapter 3. We applied relevant, generally accepted
             measurement conventions developed to help assure that measures are
             reliable and valid. (The reliability of a measure is the extent to which it
             produces the same result when repeatedly applied to a characteristic of
             an object. The validity of a measure is the extent to which it actually
             measures the attribute about which information is needed.) We also dis-
             cussed potential measurement problems with EPA and other experts
             familiar with the hazardous waste system and the activities for which
             the data will be used.

             Since the instruments EPA developed use similar or identical measures of
             basic attributes, we reviewed each type of measure across all four
             instruments. Table 4.1 summarizes the overall results of our analysis of
             the measures used to provide the information identified as necessary in
             chapter 3, and it also shows the areas where measurement problems
             persisted. (The measurement problems originally identified in the 1985
             RCFWreporting instrument are included as a baseline.)




             Page 66         GAO/PEMD-80-3   EPA’s IIaaudous   Waste Data Need F’urther Improvement
                                                        Chapter 4
                                                        Aaseesment   of Meseurement    Inetrumente




Table 411: Required Information Types for Which Measurement Problems Exist, Using the 1985 RCRA Reporting instrument as a
Baselin/e.
                                                                                         Data collection instrumenr
                                                                                 1985      1987      TSDR      GEN    TCRI




                                                                                                       X           X          X          X
                                                                                                       X
                                                                                                                   X          X          X           X

                                                                                                                   X          X          X



Major additional hazardous metals present
Major additional   hazardous
              -.- __..
                     -           nonmetals present
                         ..-.-.----.
Concentration and range of all hazardous chemicals -- present
Management   data ~-.~~ .
     ‘.._ ~. __.---.
Treatment                                                                                              X           X          X          X
    . , .~technologies
              . .-.-.---- ---        -~~--
Storage technologies                                                                                   X
Dkposa~. technologies
            . .._...__...~~ ~-..-...-           .-                                                     X
Types Qf recycling                                                                                                 X          X          X
                                                                       -.___
Residu’ai waste generation                         .___-
Type of equipment
Ancillary equipment
Capac/ty of each-unit
                _._..-~ of .--..-
                                equipment
                                       .- ----       ---                                                                      X          X
Design’ characteristics~~~affected
                              -I__.-- by proposed regulations
Design’ characteristics for estimating releases
Consiroction’~material                                   __.
System diagrams               . .._.~ --___--___-                                                                                                 __-
Type of management system (techniques and equipment)         --
Capacity of management            system                                                                           X          X          X               -_
                       _.-.. - .--____------.---
Quantity hazardous waste managed by each technique or system of techniques
Quant/ty nonhazardou~-wastemanaged             ____- by each technique or system of techniques
Plannedcapacity changes
Commercial status                           ___~--                                                                                                            -
Permit status                               ..-. _~.
Imports and
          -. _.exports
                -~ . . --_--_L ---..
Originating facility identification                    -__-
                                                                                                                                              (continued)



                                                        Page 67              GAO/PEMD-903        EPA’s Hazardous   Waste Data Need Further   Improvement
                                                  Chapter 4
                                                  Assesement   of Measurement   Instruments




                                                                                                            Data collection instrument”
Ty#e- of- information
            ..___...
                -- --.. -__-.-.-_.....-.-- -.--                                                  1985        1987       TSDR      GEN         TCRI
De tinatlon facility identification
  1
Qsohydrologic      and environmental
                                 .---__data
Sod types
Gr undwater flow-rates
Pr ,dximity -;vY
            to water--Yiir~
                      source                    ...-_
Airlemissions                               -.-
Dikance from management units to property line                                                                                           -
Digtance from property line to nearest   potential human exposure
                                    --.-__---
A&al releases to environment
W&e     source
RoCtine induskiai product&
Spkcific industrial process _~ ~~.-----
Onlee-time-oky eneration, including decommissioned equipment, off-specification
  product, CE i CLA, corrective action, closure, and other remedial action
C$RCLA v&mes requiring management ____--             ._
Coirrective ktion volumes requiring management
Adktiohal sources affected by pending regulations
Wkte minimization
V&ste stream affected
Specific industrial process                            -___
Wiste volumechange          -. -~                   ~_-.-.-                -                                                           --__
PrOduction change                                                               ____                         X                     X           X
Change in concentration/amount    of each hazardous constituent                                              X                     X
                                                  V985 = 1985 RCRA reporting instrument
                                                  1987 = 1987 RCRA reporting instrument
                                                  TSDR = National Survey of Management Facilities (TSDR survey)
                                                  GEN = National Survey of Generators
                                                  TCRI = toxic chemical release inventory reporting instrument
                                                  bAn “X” Indicates that a measurement problem exists.

                                                  Table 4.1 shows that the vast majority of measures developed by EPA
                                                  contained no identifiable measurement problems.’ Two major problems
                                                  in the baseline 1985 RCRA reporting instrument have been corrected in
                                                  the new and revised measures. Over- and undercounting of total waste
                                                  quantity have been dealt with by employing several corrective tech-
                                                  niques, including dealing with generated wastes and managed wastes in
                                                  different parts of the instruments. The revised instruments also contain
                                                  items that thoroughly address regulatory status and the reason (if any)
                                                  for nonregulated status. The less significant original problems with the


                                                  ’Measurement problems may exist that cannot be detected prospectively but only by empirical tests
                                                  of reliability and validity once the results of the data collection efforts are complete.



                                                  Page 68              GAO/PEMD-90-3      EPA’s Hazardous   Waste. Data Need Further   Improvement
                       Chapter 4
                       Assessment   of Measurement   Lnstrumenta




                       measures of disposal technologies have also been corrected. As indicated
                       in Table 4.1, we found persisting measurement problems in the areas of
                       waste type (presence of regulated waste, physical form, chemical char-
                       acteristics, and concentration), treatment technology, capacity (total
                       possible capacity of a management technology), and waste minimization
                       ( for example, production change). These measures were inadequate in
                       each instrument in which they were included. The measures of waste
                       type and treatment technology had been identified as problematic in the
                       original 1986 RCRAreporting instrument, while those of management
                       capacity and waste minimization involve areas where no quantitative
                       data had been previously collected.


                       EPA'S new and revised measurement instruments continue to use the
Mez$ming the Type of   RCRAwaste codes as a measure of waste type. As indicated in chapter 2,
Hazkdous Waste         this measure had suffered from misclassification and lack of informa-
                       tion describing waste characteristics. To correct these problems, EPA
                       developed additional measures of waste type that attempt to obtain
                       information on the specific attributes of a quantity of waste that charac-
                       terize its type.

                       The measures are intended to provide data on waste characteristics that
                       are not systematically addressed by the RCRAwaste codes, including the
                       following:

                       chemical constituents, such as the specific hazardous or nonhazardous
                       substances that determine the appropriate type of treatment
                       technologies;
                       amounts of specific chemicals found in a quantity of waste-that is, the
                       concentration of specific hazardous or nonhazardous substances that
                       determine the appropriate type of treatment technologies;
                       chemical form, such as degree of acidity (pH); and
                       physical form, such as whether the waste is a solid or liquid.

                       EPA has developed two new types of measures of waste characteristics.
                       The first type consists of qualitative variables or classification systems
                       that include categories such as metals and nonmetals, The second type
                       consists of quantitative or continuous variables such as the concentra-
                       tion of water, solids, or specific hazardous chemicals. The basic qualita-
                       tive measure of waste type was developed by EPA for use in the national
                       survey of management facilities. This measure attempts to classify
                       wastes according to their treatability. The basic measure is shown in
                       figure 4.1.


                       Page 69              GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
                                                                Chapter 4
                                                                Asseeemwrt        of Measurement              Instruments




                    I
                                                                          WASTE
Figure 4.1:Baaic EPA Measure of Waste Type From the National Survey of Management Facilities


                                                                                       DESCRIPTION CODES

                    These waste description codes were developed specifically for this survey to supplement the descrip
                    tions listed with the ACRA waste codes.
                                                                                                                                                                                     t




                    RCRA F, K, P, and U WASTE CODES
                    A01    K ~8810, exactly as described’                 A05    Wastewatar by mixture, rule contain.            A09    Incinerator ash from the treatment of
                    A02    F001.F005, as a spent solvent (organ-                 Ing F, K, p. or U                                      F, K, p, or U
                           ic liquid)                                     A06    Soil or cleanup residue contaminated            Al0    Solidification residual from the treat-
                    A03    FOOI-FOOL!. as a still bottom (organic                with F, K. p, Or U                                     ment of F, K, P, or U
                           sludge)                                        A07    P or U, as a concentrated off-                  All    Wastewater treatment residual from
                    A04    FOOB-FO26 wasta, exactly as                           specification or discarded product                     the treatment of F. K. F or U
                           described’                                     A06    Empty containers that held a P or U             Al2    Other
                                                                                 waste
                    ‘“Exsctly as described”     means that we need no further clarification   cl the description   provided I” 0 w lkst of RCRA waste codes in Appendix C 01
                    the Instructions booklet


                    RCRA D WASTE CODES
                    INORGANIC SOLIDS-Waste               that la          830    Still bottoms of nonhalogenated    sol-         INORGANIC SLUDGES-Waste               that Is
                    prlmarlly     lnorgenlc and solid, with low                  vents or liquid                                 prlmarlly    InorganIc, with moderate organ-
                    organic content snd low water content                 031    Oily sludge                                     Ic content and/or water content; poten-
                    801     Soil or debris contaminated      primarily    832    Sludge with PCBs                                tlally settlee Into two phases
                            with solvents, oil, or other organics         833    Sludge with other toxic organics                861     Inorganic sludge contaminated
                    802     Other contaminated      soil or debris        834    Organic paint sludge                                    primarily with solvents, oil, or other
                    I303 Salt of a strong acid                            835    Sludge with petroleum distillates                       organics
                    804     Salt of a strong base                         636    Reactive or polymerlzeable    organic           862     Highly acidic sludge with metals
                            (SolId NaOH, KOH. etc.)                              sludge                                          863     Other highly acidic sludge
                    805     Sulfate or sulfite                            837    Resins or viscous. nontarry organics            864     Metal hydroxide sludge
                    BOB Cyanide salt                                      836    Tars or tarry sludge                            865     Sulfide sludge
                    807     Chloride, fluoride. bromide salts             839    Biological sludge                               B66 Sulfate or sulfite sludge
                    SO6 Nitrate, phosphate, or urea salts                 840    Other organic sludge                            867     Cyanide sludge
                    809     Other metal salt                                                                                     866 Other caustic sludge
                    810      Strong oxidizer salt                         INORGANIC LIGUIDS-Waste             that la            869     Sludge with strong oxidizers
                    611      Strong reductant salt                        prlmarlly aqueous and Is highly Ibid,                  870     Sludge with strong reductants
                    812      Solid explosive or propellant                wlth low-to-modemte        suspended     InOr.         871     Sludge with explosives
                    013     Solld spent filters                           gank solids md oganlc          content                 872     Brine sludge (with hlgh chloride, flue-
                    814      Dry fly ash, metal oxide, or ores            641     Solvent-water mixture                                  ride, or bromide)
                    815      Solid metal scale, filings. or scrap         842 Oil-water emulsion or miXtura                      873      Nutrient sludge (with high nitrate,
                             (crushed drums)                              843     Concentrated water solution of or-                     phosphate, or urea)
                    816      Inorganic paint or pigment solids                    ganics                                          074     Spent filtering aids
                    817      Batteries and battery parts, casings,        844     Wastewater with trace organics                 B75 Wet scrubber sludge (fly ash), metal
                            cams. etc.                                    845 Concentrated spent acid with no                            oxides. or ores
                    816      Other inorganic solids                               metals                                         876     Sludge of metal Scala, filings, or
                                                                          846     Spent acids with metals                                scrap (crushed drums)
                    ORQANIC LIOUIDS-Waate              that Is prlmab     847     Concentrated   noncorrosive, aqueous            877     Inorganic paint or plgmant sludges
                    Ily organic and Is highly tluld, wlth low In-                 solution of metals                             876     Other inorganic sludges
                    omanlc sollds content end water content               846 Wastewater or dilute solution with
                                                                                                                                 ORGANIC SOLIDS-Waste              that is prlmar-
                    819     Halogenated solvent                                   metals
                    820     Nonhalogenated     solvent                    849 Caustic aqueous WSSte with metals                  Ily organic and solld, wlth low InorganIc
                    821     Waste oil                                             Only                                           content and water content
                    822 Any organic liquid with PCBs                      850     Caustic aqueous waste with cyanides            B79 Solid waxes or polymerized OrganiCs
                    823     Anv organic liquid/solution     of other              and matals                                     860     Spent carbon contaminated with toxic
                            toxic oiganlcs                                851     Caustic aquaous waste with cyanides                    organics
                    824 Organic paint or coating (lacquer, var-                   only                                           B61 Reactive organic solid
                            nish, epoxies)                                052 Caustic aqueous waste with sulfides                862     Halogenated off-spec or discarded
                    825     Palnt thinner or spent petroleum dis.         053 Concentrated waste caustic                                 solid organic chemical
                            tlllates                                      654 Aqueous waste wlth strong oxidizers                863     Organo-nitrogen     organic chemical
                    828     Reactive or polymerizeable       organic      855     Aqueous waste with strong reductants                   (nitrogen pesticide)
                            liquid                                        858 Aqueous waste with explosives                      864     Phosphorothioate      organic chemical
                    827     Other combustible organic liquid              057 Waste liquid mercury                               865     Miscellaneous    off-spec Organic
                    826     Other organic liquid                          656 Other aqueous waste with high dis-                         chemical
                                                                                  solved solids (brine)                          666     Other organic solid
                    ORGANIC SLUDGES-Warts               that la           859     Other aqueous waste with low dis-
                    prlmsrlly     orgsnlc, with modente      Inorgan-             solved solids content
                    Ic sollds content and water content:                  060 Other inorganic liquid
                    potentially     settlea Into phases
                    829     Still bottoms of halogenated solvents
                            or llauld




                                                               Source: EPA, National Survey of Hazardous Waste Treatment, Storage Disposal, and Recycling Facili-
                                                               ties OMB No. 2050-0063 (expired December 1987).
                                                               -’



                                                               Page 70                         GAO/PEW90-I                 EPA’s Hazardous             Waste Data Need Further           Improvement
Chapter 4
Assessment   of Measurement   In&nunenta




Initially, this measure was intended to be used in conjunction with the
RCRAwaste codes. That is, it was intended to fill information gaps left by
the RCRAwaste codes; it was not intended to be a separate exhaustive
measure of waste characteristics. This approach is a significant
improvement because it does provide more information than the RCRA
waste codes alone. However, since it was used in conjunction with the
RCRA waste codes, it is not fully satisfactory because it remains vulnera-
ble to the misclassification resulting from use of the RCRAwaste codes,
which have not been corrected. Later versions of the measure are
designed to be used independently of the RCRQwaste codes. This solves
the problem of contamination by misclassification of wastes resulting
from use of the RCRAwaste codes. However, the measure remains inade-
quate in other respects.

All versions of the basic measure of waste type assume that treatability
is a single attribute. As discussed in chapter 3, the treatability of a
waste is the result of the joint occurrence of categories or values of sev-
eral independent attributes, including physical form, chemical form, and
the concentration of specific hazardous and nonhazardous constituents.
That is, waste with a specific combination of values of each of these
independent attributes is amenable to specific types of treatment
technology.

Combining independent attributes leads to complex and redundant clas-
sification systems. For example, combining the measures of three inde-
pendent attributes with 10 values each would produce a measure with
1000 categories, each representing a unique combination of the ten val-
ues of each attribute (10 x 10 x 10). Only 30 categories would be neces-
sary if each attribute were measured separately. The separate simple
measures can be crosstabulated to reproduce the full 1000 unique
combinations.

The basic EPA measure contains 98 categories that mix chemical and
physical form, and type and concentration of constituents. Whether a
quantity of solvent waste is halogenated is addressed five different
times because it must be repeated in combination with other attributes
such as organic liquids, solids, and sludges that may contain haloge-
nated solvents. The pH of a quantity of waste is addressed by 12 sepa-
rate categories because it is first broken down by whether the substance
is acidic or caustic and then cross classified by whether it is a liquid or
solid, whether it contains metals, whether it contains cyanide, or
whether it contains both cyanide and metals.



Page 71              GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
Chapter 4
Aseessmwnt of Measurement   Instruments




A long, complex, and redundant measure is burdensome to respondents
and can be expected to produce low reliability because respondents are
not likely to interpret the measure in the same way. Separate simple
measures of independent attributes help reduce respondent burden and
increase reliability.

When complex measures result from mixed attributes, efforts are some-
times made to collapse or eliminate some combinations that are not
deemed relevant for a specific use of the measure or other practical rea-
sons. This can be a problem in that different versions of the measure
may not be comparable. The various versions of the EPA measure are not
comparable because they have been shortened in different ways. For
example, the basic measure distinguishes between liquids, solids, and
sludges according to whether they are primarily organic or primarily
inorganic, but does not include gases. The version used in the national
survey of generators adds organic and inorganic gases. The version in
the 1987 RCRA reporting instrument differentiates between wastes that
are primarily organic or inorganic but combines solids and sludge into
one category and does not differentiate between organic and inorganic
gases in the same way as the generator survey. The measure used in the
toxic chemical release inventory instrument does not distinguish
between organic and inorganic substances (although the chemical names
will help in this regard), does not distinguish between solids or sludges,
and does not ask for total quantities of each waste stream. The result is
that the different versions of the measure are not comparable. Quanti-
ties of organic solids and sludges cannot be compared because solids and
sludges are not consistently distinguished in the different versions of
the measure. In addition, the subcategories and descriptions differ and
combine noncomparable wastes in residual categories. Simple, separate
measures of each attribute would facilitate standardized measures and
assure comparability because there would be less pressure to simplify
the categories further.

Even well designed qualitative measures of hazardous waste types,
however, will result in the assignment of quantities of waste to inappro-
priate treatment technologies because the categories of qualitative
measures (such as “primarily organic,” “highly fluid,” or “moderate
inorganic solids content”) cannot capture the necessary detail. For
example, some wastes may be classified as a liquid that actually are too
viscous to be subjected to treatments designed for liquids, while some
wastes may be classified as a sludge that could be treated as liquids if
they become more fluid when heated. More accurate quantitative meas-
ures or continuous variables, such as the total suspended solids, are


Page 72            GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
                       Chapter 4
                       Assesement   of’ Measurement   Inatrumenta




                       required to precisely assign quantities of waste to appropriate
                       treatments.

                       EPA has included some quantitative    measures of waste characteristics in
                       the 1987 RCRA reporting instrument and more on the instrument used for
                       the national survey of generators, These measures can be used to char-
                       acterize the type of waste without relying on the RCRA waste codes or
                       the qualitative measures of waste type. They are the most appropriate
                       measures of waste characteristics and are significant improvements
                       over earlier RCRA reporting instruments and the new qualitative meas-
                       ures discussed previously because this type of measure significantly
                       increases reliability and accuracy. However, if handlers are not required
                       by federal regulations to conduct the tests and keep records of the
                       results, it is likely that many respondents will not be able to provide the
                       requested information, We discuss this as a data collection problem in
                       chapter 5.


                       The EPA measures of the type of treatment are qualitative measures that
Me&wring the Type of   classify the type of technologies applied to treat a waste prior to final
Triatment Technology   disposition (that is, disposal, recycling, or discharge to publicly owned
                       treatment works or surface water under National Pollutant Discharge
                       Elimination System permits).2 Quantitative measures cannot be used to
                       measure treatment type because there are unavoidable qualitative dif-
                       ferences between treatment technologies. Thus, qualitative measures or
                       classification systems are appropriate.

                       The new and revised measures of treatment types are a significant
                       improvement over those used in the 1985 RCRA reporting instrument
                       because the categories are much more detailed. However, the measures
                       are not fully satisfactory because they mix independent attributes in the
                       same way as the measures of waste type already discussed. Specifically,
                       they mix aspects of the type of waste and the disposition (for example,
                       recycling) of a waste with the technology used to treat it. As previously
                       discussed, this increases the likelihood of misclassification, reduces the
                       comparability of the separate versions of the measure, and leads to low
                       reliability.

                       Our analysis of EPA technical documents indicated that treatment tech-
                       nology is a good candidate for a general classification system with a

                       ?n some cases, such as the use of hazardous waste as a fuel, the treatment (incineration) itself repre-
                       sents recycling (a form of disposition).



                       Page 73               GAO/PEMD-90-3      EPA’s Hazardous    Waste Data Need Further      Improvement
Chapter $
Assessmwt   of Measurement   Instruments




mutually exclusive and hierarchical structure. Our review also revealed
that, through repeated efforts, EPA is slowly moving in this direction.
Successive versions of the measure have come closer to being a true gen-
eral classification system, and the 1987 RCRAreporting system instru-
ment measure is the closest yet,

Technical studies of the types of hazardous waste treatment conducted
for EPA show that there are more than a hundred different technologies
that can be used to treat hazardous wastes. Based on our review of this
technical literature, we identified four generic classes of treatment tech-
nology: (1) physical, (2) chemical, (3) biological, and (4) thermal. Indeed,
our conclusion that treatment type is amenable to the construction of a
general classification results from this four-class framework-that     is,
no treatment technology can be a member of more than one of these four
classes, and every treatment technology must belong to one of them.

In addition, there are subtypes of each major class of treatment technol-
ogy, and no technology can be a member of more than one subtype. For
example, chemical precipitation is a subtype of chemical treatment that
consists of several specific techniques, such as lime and soda ash treat-
ments. None of these treatments can be a member of any other subtype.
This allows a potential classification system with several orders or
levels of detail. Aggregation to a higher level when appropriate for prac-
tical purposes is straightforward. The resulting measure could become
standard across all EPA data collection efforts because the categories
would be mutually exclusive and hierarchical, and all attributes of treat-
ment technology would be ordered. In efforts where less detail is
required, any of the more detailed categories of the measure could be
omitted as desired. The results would remain comparable with other
applications at the next higher level of aggregation.

The current measures of treatment technology partially reflect this
structure, but they also include categories based on the type of waste
treated (wastewater treatment) and the disposition of waste (recycling
as opposed to disposal). This introduces mixed attributes in the same
way as the measures of waste type discussed previously. Although some
treatments tend to be used more frequently for wastewater treatment or
when wastes are recycled, EPA officials stated that the same technologies
can be used for other purposes. In addition, whether a waste is a waste-
water or something else has (or should have) already been determined
as part of the measure of waste type (for example, concentration of
water). There is no reason to introduce the question of waste type again
as part of the measure of how the waste was treated. Whether and how


Page 74            GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
                      Chapter 4
                      Aeeesoment of Meemuement   Inhumente




                      a waste is recycled or disposed of after treatment should also be mea-
                      sured separately as a type of disposition, and not mixed with the type of
                      treatment technology.



                      treatment technologies. As discussed in chapter 3, this involves measur-
Man$gement Capacity   ing characteristics of the equipment in which a technological process is
                      conducted. These characteristics include feed rates, ancillary equipment
                      (such as pipes and storage areas), necessary downtime, and other fac-
                      tors, depending on the specific type of equipment and process. In all but
   ,                  two cases, we did not identify any significant threats to validity or relia-
                      bility. First, the measure of the total capacity of some types of incinera-
                      tion does present a problem because it did not take into consideration
                      the effects of different uses of incineration. For example, cement kilns
                      and industrial boilers use hazardous waste as a fuel, The primary pur-
                      pose is to produce heat for an industrial operation, not to destroy haz-
                      ardous waste. Thus, the amount of hazardous waste necessary to
                      produce the needed heat may be less than the total amount of waste that
                      could be incinerated, because demand for the primary product may be
                      less than the maximum output. Wastes with different heating values
                      would also need to be used in different amounts to produce the required
                      heat, This lack of specificity means that respondents can easily interpret
                      the measure differently, which reduces reliability. In addition, if respon-
                      dents assume either unlimited demand for their principal product or low
                      heat-producing wastes, capacity could be significantly overestimated.
                      This potential problem would be eliminated by making the measure
                      more specific.

                      Second, it is important to distinguish between the maximum physical
                      capacity to treat wastes and the maximum amount allowed by the EPA
                      operating permit, which may be significantly less than the physical
                      maximum. Respondents to the survey of management facilities were
                      asked to report the total amount of waste that could have been treated
                      during the year, assuming existing operating constraints, equipment,
                      and an unlimited supply of waste. EPA officials indicated that constraints
                      specified in the operating permit were supposed to be included by
                      respondents, although they were not specifically listed in the survey
                      instrument. However, they acknowledged that some respondents based
                      their answers on maximum physical capacity rather than the maximum
                      allowed by their operating permit. For example, one facility reported a
                      large capacity to stabilize wastes prior to land disposal even though the



                      Page 76           GAO/PEMD-30-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                  Chapter 4
                  Assessmenl; of Measurement   Instrumenta




                  operating permit restricted the amounts that could actually be land-dis-
                  posed. EPA officials also stated that this overreporting of capacity could
                  not always be detected because in many cases there would be no identi-
                  fiable inconsistency in the responses. EPA officials acknowledged that
                  future efforts should specify that available capacity includes the con-
                  straints imposed by the operating permit, as well as purely physical or
                  technical constraints.

                  While this specific problem was addressed in the 1987 RCRAreporting
                  instrument, the measure suffers from another problem that can produce
                  overreporting of capacity. Specifically, respondents are asked to calcu-
                  late the total capacity of treatment systems rather than (as in other ver-
                  sions) to provide information on each unit of the treatment system,
                  which permits EPA to calculate the capacity of treatment systems. The
                  danger here is that different facilities may calculate system capacity in
                  different ways, thus reducing the reliability and validity of the measure.


                  Three of the four instruments we evaluated attempt to measure the
Mdasuring Waste   amount of hazardous waste minimization that has occurred. Each of
Minimization      these attempts to measure the amount generated per unit of production
                  over time in order to develop information on progress toward minimiz-
                  ing waste generation. Relative waste minimization is important because
                  the absolute volume of waste generated over time varies with the
                  amount of production. Therefore, in addition to the absolute amount
                  generated, it is useful to measure the amount generated per unit of pro-
                  duction to control for annual variations in the amount of production.
                  However, the problem we find with such relative measures of waste
                  reduction is that they produce misleading results.

                  The first step in measuring relative waste minimization is to calculate
                  the ratio of production quantity in one year to that of the preceding
                  year. For example, if a tool manufacturer produced 1,200 tools in 1986
                  and 1,000 tools in 1987, the production ratio would be 1,200 to 1,000 or
                  1.2. The next step is to calculate the ratio of the amount of hazardous
                  waste produced during the production of the tools over the two years. If
                  the firm produced 12 tons in 1986 and 10 tons in 1987, the ratio would
                  be 12 to 10 or 1.2. The ratio of the two ratios represents the firm’s prog-
                  ress toward minimizing waste generation per unit of production. In this
                  case, the ratio is 1.2 to 1.2 or 1, representing neither an increase nor a
                  decrease in the amount of waste generated per unit of production
                  between 1986 and 1987 (even though there was an absolute decrease of
                  16.6 percent).


                  Page 76             GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
    chapter4
    As#e#ment   of Measurement   Instrumenti




    If the firm had produced only 9 tons of waste in 1987, an absolute
    decrease of 26 percent, the result would have been

.   a production ratio of 1,200 (1986) to 1,000 (1987) = 1.2
.   a waste generation ratio of 12 to 9 = 1.333
.   a waste reduction ratio of 1.2 to 1.33 = .90.

    Accordingly, this result would be said to represent a 10 percent reduc-
    tion in the amount of hazardous waste generated per unit of production.

    The foregoing measure is inadequate because it does not account for the
    specific types of tools (the production mix) produced during a given
    year. For example, the company in the preceding example may have lost
    a contract for a specific type of tool in 1987. If the production of that
    tool generated more hazardous waste per unit than the production of
    other tools, the results of the production ratio method will be mislead-
    ing. It will appear that relative waste minimization has occurred when
    in fact it did not. In the preceding example, the appearance of waste
    reduction was produced because in 1987 the firm did not produce the
    product that generated hazardous waste at a greater rate than the other
    products. During the next year, if the company returned to its original
    production level and mix by producing that tool again, the production
    ratio method would produce the following results:

    a production ratio of 1,000 (1987) to 1,200 (1988) = .833
    a waste generation ratio of 9 to 12 = .750
    a waste reduction ratio of .833 to .750 = 1.111.

    This result would create the appearance of an 11.1 percent increase in
    the amount of hazardous waste generated per unit of production when
    in fact no change had occurred. Instead, the company merely returned to
    its previous production level and mix. In addition, real increases or
    reductions in relative hazardous waste generation would be mixed with,
    and thus obscured by, incidental changes such as these.

    It could be argued that these incidental changes would average out on an
    industry-wide basis and that the resulting information thus would give
    relatively valid information on waste minimization for an industry. In
    some industries this may in fact be true, but in other industries it is
    likely that the production mix does vary from year to year, depending
    on seasonal or other market factors. The problem is also more pro-
    nounced in certain industries where waste streams are not often sepa-
    rated according to the specific product produced.


    Page 77            GAO/PEMDBO-3        EPA’s Hazardous   Waste Data Need Further   Improvement
                    Chapter 4
                    Aasessmertc, of Measurement   Instruments




                    We also found problems with the measures of the change in toxicity
                    over time. The measure is a qualitative one and therefore does not pro-
                    vide any quantitative information on changes in toxicity (as measured
                    by concentration of constituents). It provides information only on
                    whether the concentration of hazardous constituents in general has
                    increased or decreased and whether less hazardous constituents were
                    substituted. It provides no information on how much the concentrations
                    changed. The resulting data could only show the percent of cases in
                    which waste minimization resulted in increased concentration. Cor-
                    recting these problems would require measures that address the precise
                    production mix and the waste associated with each product, as well as
                    detailed data on the concentration of hazardous constituents. These
                    detailed data could be collected from a sample, which would avoid the
                    larger burden of collecting the information via a census. However, as
                    noted previously, aggregating these data to create a meaningful compar-
                    ison across production processes and industries would remain difficult.


                    The measurement problems just discussed will produce some error in
Implications for    specific areas of the capacity assessments conducted by EPA and states.
Capacity Analysis   However, without the actual data resulting from the use of the measures
                    to supply information about actual (versus prospective) reliability and
                    validity, we cannot estimate the magnitude of this error.

                    As discussed in chapter 4, the capacity analysis for the land disposal
                    restrictions involves identifying the volumes of wastes covered by a spe-
                    cific restriction that are currently land-disposed. The identification of
                    covered wastes is based on the RCRAwaste code reported by respon-
                    dents. Since the use of these codes results in misclassification, there will
                    be error in determining the volume of waste that represents a given code
                    and is thus subject to specific restrictions.

                    All capacity analyses involve the process of determining the volumes of
                    wastes amenable to specific types of treatment. The qualitative measure
                    used to characterize the treatability of wastes is likely to result in some
                    misclassification that will affect the volumes of waste assigned to spe-
                    cific types of management technologies. This in turn will affect all three
                    types of capacity analyses since all data collection instruments use this
                    qualitative measure. However, the problem will be minimized in those
                    instances where the alternative quantitative or continuous variables are
                    used.




                    Page 78             GAO/PEMDBO-3       EPA’s Hazardous   Waste Data Need Further   Improvement
  ,




                      Chapter 4
                      A s s e s s m e n t of M e a s u r e m e n t Instruments




                      M isclassification errors resulting from th e m ixture o f a ttributes in th e
                      qualitative m e a s u r e o f treatment te c h n o l o g i e sl e a d to p o te n tial inaccura-
                      cies in th e d e te r m i n a tio n o f to tal capacity o f specific treatment te c h n o l o -
                      gies. B e c a u s ethis p r o b l e m exists in all variations o f th e m e a s u r e ,this
                      p o te n tial will b e p r e s e n t in all th e capacity a n a l y s e s E P Aperforms.

                      T h e to tal capacity o f s o m e types o f incineration (such a s c e m e n t kilns or
                      industrial boilers) m a y b e overestimated b e c a u s eE P Ad i d n o t c o n trol for
                      th e h e a t v a l u e o f th e w a s te incinerated or th e variation th e d e m a n d for
                      p r o d u c ts p r o d u c e d with th e a i d o f s o m e types o f incineration. This
                      p o te n tial overestimation p r o b l e m will b e p r e s e n t in all types o f capacity
                      analysis.

                      A n a s s e s s m e n to f th e i m p a c t o f w a s te m inimization practices is n e c e s -
                      sary for integrated capacity a n a l y s e s a n d S A R Acapacity a s s u r a n c e
                      b e c a u s eth e d e g r e eo f w a s te m inimization a c h i e v e d will a ffect th e c a p a c -
                      ity r e q u i r e d in th e fu ture. H o w e v e r , since n o n e o f th e m e a s u r e so f w a s te
                      m inimization c o n trols for p r o d u c t m ix or includes q u a n tita tive m e a s u r e s
                      o f toxicity, all capacity a n a l y s e s will h a v e m isleading estimates o f th e
                      extent o f w a s te m inimization.


                      T h e m e a s u r e s i n c l u d e d in E P A 'Srevised d a ta collection instruments a r e
S u n p n a rya n d   m u c h improved. T h e p r o b l e m s o f u n d e r c o u n tin g a n d overcounting th e
C o klusions          to tal v o l u m e o f w a s te s g e n e r a te d h a v e b e e n corrected. T h e previous
                      p r o b l e m s with th e m e a s u r e so f w a s te m a n a g e m e n ttechnologies,with
                      th e e x c e p tio n o f th o s e involving treatment technologies,h a v e also b e e n
                      fully corrected. In a d d i tio n , significant i m p r o v e m e n ts h a v e b e e n m a d e in
                      th e m e a s u r e s o f w a s te a n d treatment type, a l t h o u g h th e s e improve-
                      m e n ts h a v e n o t c o m p l e tely eliminated th e original problems.

                      T h e r e still r e m a i n s o m e m e a s u r e m e n tp r o b l e m s th a t m a y p r o d u c e signif-
                      icant errors in d e te r m i n i n g th e v o l u m e o f e a c h type o f w a s te g e n e r a te d ,
                      th e type o f treatment te c h n o l o g i e su s e d , th e to tal capacity o f m a n a g e -
                      m e n t te c h n o l o g i e s(such a s s o m e types o f incineration), a n d th e d e g r e e
                      o f w a s te m inimization. T h e R C R Aw a s te c o d e s ,w h i c h a r e u s e d o n all d a ta
                      collection instruments, will c o n tin u e to p r o d u c e inaccurate c o u n ts o f th e
                      v o l u m e s o f regulated w a s te streams d u e to m isclassification. T h e quali-
                      ta tive m e a s u r e o f w a s te type E P Ad e v e l o p e dto s u p p l e m e n t th e R C R A
                      w a s te c o d e s ,a l t h o u g h a n i m p r o v e m e n t over p a s t m e th o d s , h a s n o t fully
                      solved th e p r o b l e m . T h e m ixture o f i n d e p e n d e n ta ttributes creates a
                      c o m p l e x a n d r e d u n d a n t m e a s u r e th a t c a n significantly r e d u c e reliability
                      b e c a u s er e s p o n d e n ts a r e likely to interpret it differently. Qualitative


                      Page 79                       G A O /P E M D - B B - 3 E P A ’e H a z a r d o u s W a s t e Data N e e d Further Im p r o v e m e n t
                  Chapter 4
                  Asee~rmwt   of Measurement   In&umenta




                  measures are also much less accurate than available quantitative meas-
                  ures. On some instruments, EPA has included the preferable quantitative
                  measures; these represent an important improvement over the qualita-
                  tive measures. EPA should continue to expand the use of quantitative or
                  continuous variables for measuring waste characteristics.

                  Although EPA has developed new, much improved measures of the types
                  of management technologies, misclassification remains likely due to the
                  mixing of attributes. The development of a true general classification
                  system with mutually exclusive, exhaustive, and hierarchical categories
                  would fully correct the remaining problem.

                  In general, waste management capacity is now measured well. However,
                  the measure of the capacity of cement kilns and similar forms of inciner-
                  ation was not sufficiently specific to assure that respondents will inter-
                  pret the question in the same way. If respondents assume either
                  unlimited demand for their product or waste with a low heating value,
                  the capacity of this form of incineration could be significantly overesti-
                  mated. Some overestimation of capacity is also likely in some versions of
                  this reporting instrument because some respondents may not have con-
                  sidered permit restrictions in reporting maximum capacity.

                  Finally, the measures of the extent of waste minimization developed by
                  EPA will not produce valid data on changes in waste generation per unit
                  of production. The measures do not account for the production of differ-
                  ent products from one year to the next that may generate unequal
                  amounts of hazardous waste. The results will be misleading because
                  incidental changes in production will be mixed with, and thus will
                  obscure, actual waste minimization. Moreover, the qualitative measure
                  of toxicity can only show the percentage of cases in which waste mini-
                  mization led to increased concentration, not the amount of change in
                  toxicity over time. In order to correct these problems, EPA should mea-
                  sure the volume of waste produced by each type of product and the con-
                  centration of each toxic constituent. These detailed data could be
                  collected from a sample, which method would be preferable to the larger
                  task of collecting the information via a census.


                  We recommend that the Administrator of EPA direct the Assistant
Recommendations   Administrator for Solid Waste and Emergency Response to assure the
         ”
                  use of the most appropriate measures of the relevant attributes of haz-
                  ardous waste generation and management. Specifically, quantitative
                  measures should be used to measure waste characteristics (such as those


                  Page 80             GAO/PEMD-90-S    EPA’s Hazardous   Waste Data Need Further   Improvement
                m

    Chapter 4
    Awearment   02 lbaurement   In6trumenta




    needed for assessing management capacity or waste minimization), and
    in addition, a true general classification system should be developed for
,
    treatment technologies.




    Page 81              GAO/PEMD-90-3    EPA’s Hazardous   Waste Data Need Farther   Improvement
CIJapter 5

Assessmentof Data Collection Methods
and Procedures

-
                          This chapter addresses evaluation question six: Are EPA'Srevised data
                          collection methods and procedures likely to result in valid national
                          information? We determined whether the data collection problems iden-
                          tified in the original information system have been corrected in the new
                          and revised data collection mechanisms and whether any new data col-
                          lection problems exist.



D&a Collection in the
Inierim Information
S&tern

The 1987 RCRA Reporting   The RCRA reporting system is a census of large-quantity hazardous waste
Cyicle                    handlers. For a census to yield valid data, it is essential that the regu-
                          lated population be accurately identified and that every member of the
                          reporting population respond in the same manner. The types of data col-
                          lection problems encountered in the past have included (1) inaccurate
                          identification of the regulated population, (2) inconsistent information,
                          and (3) quality control.

                          The major steps EPAtook to correct these problems for the 1987 cycle
                          included developing the revised 1987 reporting instrument (and its data
                          definitions), which was used voluntarily. The complete revised instru-
                          ment was used for the 1987 cycle in 16 states, including the 5 remaining
                          unauthorized states and the 10 states that used the instrument volunta-
                          rily.’ The remaining 34 states used the instruments they had used in the
                          past, and were required to obtain the information that was included in
                          the 1985 RCRA reporting instrument. EPA initially required states to for-
                          ward to EPA disaggregated data (needed for data analysis and quality
                          control) rather than the summaries that were required in the past.2
                          Finally, EPA provided assistance to states to process the data collected
                          into the format required by EPA.We evaluated the likelihood that these
                          actions would correct past data collection problems and the likelihood
                          that they would create additional problems in the areas identified.



                          ‘The District of Columbia also used the entire instrument, and EPA administered the complete instru-
                          ment in one authorized state and four territories.

                          2Ultimately, the submission of disaggregated data was made voluntary. Thus, states may submit
                          aggregated data in a form that is consistent with the summaries required in the past.



                          Page 82               GAO/PEMD-90-3     EPA’s Hazardous    Waste Data Need Further    Improvement
                                  Chapter 5
                                  Assessment of Data Colktfon      Methods
                                  and Procedures




Identifjcation of the Regulated   As discussed in chapter 2, previous cycles of the RCRA reporting system
Population                        did not adequately update the notification forms used by waste handlers
                                  to notify EPA of their activities. EPA thus was unable to develop accurate
                                  lists of active hazardous waste handlers, As discussed in chapter 4, the
                                  new information on regulatory status included in the instrument for the
                                  1987 RCRA reporting cycle fully corrects this problem. If accurate data
                                  on regulatory status are collected in all states, the regulated population
                                  would be adequately identified for the first time. However, the conse-
                                  quences of the voluntary use of the revised data collection instrument,
                                  which we will discuss next, are likely to interfere with the accomplish-
                                  ment of this and other goals.

Infonrjation Consistency          Thirty-five states (70 percent) elected not to use the complete revised
                                  data collection instrument in the 1987 reporting cycle.3 Although EPA
      I                           refers to the 1987 cycle as a pilot test of the revised instrument, the
                                  1987 cycle is one of the principal sources of data that will be used to
                                  support important national initiatives, including SARAcapacity assur-
                                  ance. It therefore is essential that it produce valid national data. States
                                  electing not to use the new reporting instrument used those they had
                                  used in previous reporting cycles. Some of these states used the instru-
                                  ment EPA developed for the 1985 reporting cycle, and some used their
                                  own instruments.

                                  EPA'Srequirement that authorized states not using the revised instru-
                                  ment provide the information that was sought on EPA’S 1985 data collec-
                                  tion instrument cannot produce valid national information because that
                                  instrument did not collect certain important information. Thus, major
                                  gaps will exist for waste characteristics, waste management, and availa-
                                  ble capacity in all states that did not use the revised instrument.

                                  Because many states used the 1985 EPA data collection instrument for
                                  the 1987 cycle, the same measurement problems as were discussed pre-
                                  viously are likely to recur. Since other sources of national data exist for
                                  management and capacity issues, these measurement problems will
                                  affect the data on waste volumes and types most significantly. As dis-
                                  cussed in chapter 2, these problems involved misclassification of RCRA
                                  waste codes, reporting wastes as mixed waste, incomplete identification
                                  of RCRA waste codes that apply to a given quantity of waste, the absence
                                  of waste characteristics data, and under- and overcounting of the total

                                  3Eighteen of the 36 elected to use the waste minimization portion of the form; EPA administered the
                                  waste minimization portion of the instrument in all states that did not elect to use it because the
                                  information was mandated by HSWA; and EPA administered the entire revised instrument in one of
                                  the 36 nonparticipating states.



                                  Page 83               GAO/PEMD-90-3     EPA’s Hazardous    Waste Data Need Further    Improvement
Chapter P
Assessmew:. of Data Collection   Methods
and ProcwUes




volumes of waste generated and managed. As reported in chapter 2, the
1986 EPA instrument and state instruments did not contain items that
accurately identified regulatory status. Therefore, the information
obtained using these instruments in the 1987 cycle will have the same
limitations.

In response to our inquiries on this subject, EPA officials explained that
the states will be expected to perform the necessary quality control
work to assure that the data gathered are valid. However, since prob-
lems such as overcounting are inherent in the measures used on EPA'S
1985 measurement instrument, quality control work cannot correct
them. EPA officials explained, however, that their use of the term quality
control includes using additional data or recontacting facilities to com-
pensate for measurement problems. The instructions for state submis-
sions, however, acknowledge that securing valid data will not always be
possible.

Finally, there are no federal regulations that require states to collect
specific federal data elements. Therefore, EPA'S requirement that states
provide these data elements is questionable.

EPA officials indicated that state submission of disaggregated data rather
than summaries of the data collected would correct the major problems
experienced in previous cycles. Disaggregating data is an effective
method of (1) separating formerly aggregated state regulated wastes
and RCRA wastes in cases where explicit codes are used to identify each
in the raw data, and (2) separating formerly aggregated RCRA defined
wastes. It is also needed for EPA analyses. Although an important
improvement, disaggregated data cannot solve all the major problems
experienced in the past. The measurement problems identified earlier
cannot be resolved by using disaggregated data. For example, since
respondents often did not include all applicable waste codes, disaggre-
gated data will only partially solve the problem. In addition, EPA officials
explained that some states continued in the 1987 cycle to give inconsis-
tent guidance to handlers on whether to report hazardous wastes
treated in units that are exempt from RCRA regulations. These include (1)
waste waters that are treated under National Pollutant Discharge Elimi-
nation System permits and discharged into surface water or to publicly
owned treatment works, and (2) the waste handled by a limited number
of recycling operations. The inconsistent reporting of these large
volumes of exempt waste water created major discrepancies in the 1985
cycle. These wastes cannot be distinguished on the basis of RCRA waste
codes because the distinction is based on how the waste was treated.


Page 84              GAO/PEMD-90-3     EPA’s Hazardous   Wa&e Data Need Further   Improvement
                      chapter 6
                      Assessment of ILta   Collection   Methods
                      and Procedures




                      EPA’S original requirement that states submit disaggregated data is also
                      questionable. EPA has limited ability to require states to submit disaggre-
                      gated data because federal regulations continue to explicitly require
                      states to submit summaries. EPA has now relaxed the requirment to sub-
                      mit disaggregated data for the 1987 cycle because of limited resources
                      and a request from the states that aggregated data be allowed. This may
                      have an adverse impact on the quality of the data, for the reasons dis-
                      cussed previously, in states that use EPA’S 1985 forms and do not submit
                      disaggregated data.

                      Most importantly, much of the new information included in the revised
                      reporting instrument is not fully supported by federal or state record-
                      keeping and reporting regulations. Although full authority is granted by
                      RCRA sections 3002, 3004, and 3007 to collect this information, this
                      authority has not been fully implemented in federal regulations. That is,
                      federal regulations do not require handlers to keep records of many of
                      the requested data elements. Without additional regulations, handlers
                      may not be able or willing to provide the requested data. This problem is
                      likely to affect information on quantitative measures of waste charac-
                      teristics, wastes treated in exempt units, nonhazardous constituents of
                      hazardous waste, and nonhazardous wastes treated in RCRA permitted
                      units. EPA officials noted that the waste minimization and capacity items
                      are also not specifically covered by federal regulations.

                      Finally, small quantity generators-generally      defined as those that pro-
                      duce less than 1,000 kilograms of waste in a given month-are not
                      required to complete the revised data collection instrument.4 Although
                      they produce a small percentage of the total waste, they are by far the
                      largest class of generators. EPA currently estimates that there are
                      100,000 small quantity generators. Although it may be overly burden-
                      some to include all small quantity generators in a census, some routine
                      information that would characterize their activities is necessary. Some
                      states, however, do require small quantity generators to report, which
                      will create further problems. In fact, because the 1985 EPA instrument
                      does not separate amounts generated by month and does not include
                      accurate data on regulatory status, it is likely that it will not be possible
                      to accurately separate small quantity generators and thus produce con-
                      sistent national counts.

Quality Control   v   EPAinitially planned to develop an automated data management system
                      and a national repository for the data collected from the RCRA reporting

                      4They are, however, required to confirm their status as small quantity generators.



                      Page 86               GAO/PEMD-90-3     EPA’s Hazardous    Waste Data Need Further   Improvement
Chapter 5
Assessnwnt of Data Collection   Methods
and Prow (urea




system. These activities would be a major accomplishment because they
help transform the data collected into a consistent usable form. How-
ever, due to limited resources, EPA will not provide all of the software
planned for the system. In addition, these technical assistance activities
are applicable primarily to states that voluntarily adopted the revised
reporting instrument. Furthermore, since not all states will supply dis-
aggregated data, the national repository will not contain the complete
disaggregated data, as was called for in the original plan. Finally, EPA is
not requiring any substantive verification, such as on-site validation.
Adequate verification should include on-site validation, in which the
completed data collection instrument would be checked against the han-
dlers’ records, for a small sample of those reporting. The results of this
effort would then be used to estimate reporting error and noncompliance
in the reporting population.

EPA is conducting a review of the mechanisms and instruments in states
that do not use the EPA instrument. One of the purposes of this effort is
to determine the extent to which existing data in the states can be trans-
formed into the data elements on EPA'Snew instrument. However, while
this review will provide valuable information for the future, it cannot
solve the problems of the 1987 cycle.

EPA'Splan for developing accurate information     depends largely on the
extent of the quality control work that is performed by the states to
detect and correct reporting inconsistencies, as well as the measurement
problems that stem primarily from the measures in the 1985 instrument
and the lack of supporting federal reporting and recordkeeping require-
ments. EPA officials explained that the states are expected to perform
the necessary quality control work to provide accurate information.
However, EPA officials agreed that the agency has not provided the nec-
essary funding or incentives to permit and encourage states to perform
this work. EPA documents indicate that the agency pays for approxi-
mately 25 percent of the cost of the RCRAreporting system, while the
states pay appproximately 75 percent to meet their own needs. How-
ever, according to EPA guidelines, grants to states are to provide 75 per-
cent of the funding for federally required activities. Since the RCRA
reporting system effort is mandated by federal regulations, we see no
reason why EPA should not provide 75 percent of the funding. Agree-
ments with states now include specific mention of the importance of the
RCRAreporting system. This is a step toward increasing the incentive for
states to collect accurate data. However, without adequate funding, it is
not likely to be an effective one.



Page 86             GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
                                              0,.
                                chapter   6
                                Assessment of wta Collection       Methods
                                and Procedures




The qational Surveys            As discussed previously, EPAdeveloped and implemented two national
                                surveys: the survey of generators and the survey of management facili-
                                ties. As we did in the case of the RCRA reporting system, we evaluated
                                the two national surveys developed by EPAin the three technical areas in
                                which problems were encountered in past data collection efforts: (1)
                                identification (and sampling) of the regulated population, (2) consis-
                                tency of information, and (3) quality control.

Identifjcation and Samplingof   In addition to the identification of the regulated population (discussed
the RekulatedPopulation         earlier) required for a census, surveys must draw a sample from the
      ,                         population that provides the necessary precision for state or national
                                estimates of population characteristics. The initial identification of the
      !                         regulated population for both national surveys was based on informa-
                                tion submitted on notification forms, which included handlers who were
                                not active and did not include those who had not filed notification
                                forms. However, steps were taken to m itigate this problem in developing
                                the sampling frames. The survey of management facilities was further
                                based on those handlers who had obtained permits.6 In addition, a
                                “screener survey” preceded the survey of management facilities that
                                was, in part, intended to accurately identify the population of manage-
                                ment facilities and the management activities performed by each. Both
                                surveys also included regulatory status questions. The sampling frame
                                for the generator survey included all of the approximately 41,000 han-
                                dlers who reported generating large quantities of hazardous wastes in
                                1985. (This group included all management facilities with the exception
                                of those that only store hazardous waste.) However, although small
                                quantity generators were thus theoretically excluded, many small quan-
                                tity generators who could not be distinguished from large quantity gen-
                                erators were actually included in the sampling frame. Of course, this
                                group was not required to complete the survey instrument. Further,
                                since the sampling frames for the two surveys do not include generators
                                who failed to notify or all those not reporting in 1985, small quantity
                                generators, some recycling facilities, or those generators who began or
                                resumed generating hazardous waste in 1986, the sampling frame for
                                the surveys is likely to be slightly smaller than the total regulated popu-
                                lation. This will result in some undercounting of the total amount of haz-
                                ardous waste generated and total management capacity. W e do not

                                “The sampling frame included all facilities with RCRA permits, This group includes the majority of
                                recycling facilities. A limited number of specific types of recyclable materials and facilities that only
                                recycle on-site without storing wastes prior to recycling are not covered by these requirements. In
                                addition, some off-site recycling facilities do not have permits becausethey contend that they do not
                                store wastes prior to recycling. These facilities were not included in EPA’s sampling frame and may
                                represent a significant amount of available capacity in some states.



                                Page 87                GAO/PEMD-90-3      EPA’s Hazardous Waste Data Need Further          Improvement
Aesesracsnt of Data Collection   Methods
and Prw &wee




consider this a significant data collection deficiency because for techni-
cal reasons perfect sampling frames are rare. However, in publishing the
results of these surveys, EPA should include an assessment of the sam-
pling frame and note that the statistics do not include some members of
the regulated population.

Both national surveys used stratified sampling-that       is, probability
sampling in which different proportions of different groups of the popu-
lation are sampled to assure adequate coverage and precision at the low-
est cost. The stratified sample for the survey of management facilities
included two major strata: (1) facilities that treat, dispose of, or recycle
hazardous wastes regulated under RCRA and (2) facilities that only store
hazardous wastes. The sample included 100 percent of the members of
the first strata-that    is, all of the approximately 2,500 facilities that
treat, dispose of, or recycle hazardous wastes. The second strata, com-
posed of the 682 facilities that only store hazardous wastes, was further
stratified according to the overall volume of waste stored. All seven of
the largest storage facilities were included. Approximately 110 of the
remaining 675 facilities were sampled, with approximately a 16 percent
chance of inclusion. Adding the resulting 117 storage facilities yielded a
total sample of approximately 2,617.

The sample for the generator survey was stratified by size of generator
and state. The 1,000 largest generators (that is, those that produce the
vast majority of all hazardous waste) and all management facilities were
sampled at the rate of 100 percent. The remaining generators were fur-
ther stratified by size, and the larger of these were over-sampled. Strati-
fying the sample by state meant that 50 separate samples of
approximately 200 generators each were included. (All generators were
included in small states with fewer than 263 generators.) The total sam-
ple was approximately 10,000, which represented slightly less than 25
percent of the sampling frame.

Given the many uses to which these surveys will be put (due, in part, to
the absence of other valid data), we conclude that the size of the sam-
ples is not excessive. For example, the sample sizes for both surveys
were increased from the original sampling plan in order to provide data
to assist the states in developing the capacity assurances required by
SARA.In general, it takes a much larger sample to provide estimates for
50 states that all have the same precision as one national estimate.




Page 88               GAO/PEMD-90-3        EPA’s Hazardous   Waste Data Need Further   Improvement
         0




                                Chapter 5
                                A s s e s s m e n t of Data Collection M e t h o d s
                                and Procedurw




                                H o w e v e r , w e believe it w o u l d h a v e b e e n m o r e e fficient for E P A ,a n d less
                                b u r d e n s o m efor th e regulated c o m m u n i ty, to h a v e d e v e l o p e dtwo s a m -
                                ples for e a c h survey-a larger s a m p l e for d e v e l o p i n g th o s e estimates
                                n e e d e dfor e a c h state a n d a smaller s u b s a m p l e for d e v e l o p i n g th o s e esti-
                                m a te s n e e d e donly a t th e n a tio n a l level. T h e larger s a m p l e s w o u l d h a v e
                                b e e n i d e n tical to th o s e actually u s e d for th e two surveys discussedpre-
                                viously a n d w o u l d h a v e b e e n u s e d to p r o v i d e only th e information nec-
                                essary to assist states in p r e p a r i n g th e S A R Acapacity assurances.T h e
                                smaller s u b s a m p l e sw o u l d h a v e b e e n u s e d to p r o v i d e equally precise
                                n a tio n a l estimates for th e m o r e d e tailed information n e e d e db y E P Afor
                                d e v e l o p i n g regulations. T h e smaller s a m p l e s w o u l d h a v e b e e n similar to
                                th o s e E P Aoriginally p l a n n e d . For e x a m p l e , b e fo r e increasing th e s a m p l e
                                sizes to p r o v i d e precise estimates for e a c h state, m o r e strata w e r e s a m -
                                p l e d a t less th a n 1 0 0 p e r c e n t, a n d s e p a r a te s a m p l e s w e r e n o t d e v e l o p e d
                                for e a c h state. Overall, 6 ,0 0 0 g e n e r a tors a n d 8 0 0 m a n a g e m e n tfacilities
                                w e r e a d d e d to th e two s a m p l e s discussedpreviously to d e v e l o p precise
                                state estimates. H o w e v e r , it is likely th a t all o f th e largest g e n e r a tors
                                a n d c o m m e r c i a l m a n a g e m e n tfacilities w o u l d h a v e to b e i n c l u d e d in b o th
                                th e larger a n d th e smaller samples, a s w a s th e c a s e in E P A 'Soriginal p l a n .
                                T h e u s e o f two s a m p l e s w o u l d h a v e b e e n especially e fficient for collect-
                                i n g d a ta to characterize th e a p p r o x i m a tely 1 0 0 ,0 0 0 small q u a n tity g e n -
                                erators, b e c a u s ewhile th e s e d a ta w o u l d n o t h a v e b e e n n e e d e dfor e a c h
                                state, th e y c o u l d h a v e b e e n i n c l u d e d in th e smaller n a tio n a l s a m p l e . T h e
                                u s e o f two s a m p l e s appropriately stratified b y size a n d type o f activity
                                w o u l d reflect th e different levels o f d e tail o f th e r e q u i r e d information
                                th a t w e discussedin c h a p ter 3 . E P A 'Sdecision to u s e only o n e s a m p l e is,
                                in this case, u n d e r s ta n d a b l e in view o f th e tim e constraints u n d e r w h i c h
                                th e surveys w e r e d e v e l o p e d .

In fo r m a tio n Consistency   T h e p r o b l e m s involving th e lack o f consistent d a ta th a t a ffect th e R C R A
                                reporting system a r e n o t p r e s e n t in th e special surveys. S ince E P Ac o n -
                                d u c ts th e survey o n its o w n without h a v i n g to interface with state d a ta
                                collection systems, every r e s p o n d e n tin th e s a m p l e received exactly th e
                                s a m e d a ta collection instrument. A lth o u g h a r u d i m e n tary research pro-
                                c e d u r e , this is o n e o f th e m a j o r strengths o f th e s e d a ta collection e fforts
                                in c o m p a r i s o n with th e R C R Areporting system.

                                H o w e v e r , th e lack o f fe d e r a l r e c o r d k e e p i n g a n d reporting r e q u i r e m e n ts
                                is likely to a ffect d a ta quality adversely. T h e surveys w e r e c o n d u c te d
                                u n d e r th e a u thority p r o v i d e d b y R C R Asection 3 0 0 7 . A c c o r d i n g to E P A
                                o fficials in th e O ffice o f th e G e n e r a l C o u n s e l ,section 3 0 0 7 a u thorizes
                                E P Ato require h a n d l e r s to p r o v i d e a n y information th a t is available a t



                                Page 89                    G A O /P E M D - 9 0 - 3 E P A ’s H a z a r d o u s W a s t e Data N e e d Further Im p r o v e m e n t
Chapter 6
Assessmmt of Data Collection   Methods
and Pm ~‘~&ree




the facility, but it may not be used to require handlers to develop infor-
mation that is not already available. Therefore, without conducting an
inspection of each facility, EPA cannot determine whether the facility
actually does not have the information or simply neglected to provide it.
The official also stated that such inspections would overload EPA'S
enforcement resources and diminish the credibility of the agency
because it could not enforce its requirements. Yet, it is also the case that
extensive missing data or nonresponses can invalidate an otherwise
flawless data collection effort.

In a sample survey, one of the most important parts of processing the
data collected is the quality control work done to verify the accuracy of
the data obtained. We reviewed the methods and procedures used to
ensure data integrity in EPA'Ssurveys and discussed these with responsi-
ble EPA officials, The EPA effort to ensure complete and consistent
responses for the survey of management facilities was extensive. Each
completed questionnaire was reviewed by technically qualified person-
nel. When problems were detected in the responses, the facilities were
telephoned for additional information. This effort was possible because
all quality control work was sponsored directly by EPA. The agency did
not have to rely on complex efforts to ensure that 50 separate data col-
lection systems performed the necessary quality control work. The abil-
ity to ensure quality control is a second major strength of EPA-sponsored
sample surveys.

As with the RCRAreporting system effort, however, the surveys did not
use on-site validation to assess data quality. When conducted on a small
subsample of respondents such on-site validity studies are an effective
tool for quality control.

Ensuring that the results of a data collection effort are available for use
in achieving agency objectives requires practical arrangements that
ensure proper timing. Thus, in order to ensure that partial results of the
survey of management facilities were available when needed to support
regulations required by HSWA,EPAdeveloped two separate quality con-
trol efforts and two separate data bases. Although cumbersome and not
recommended as a standard practice, the effort ensured that the data
were available to meet agency objectives. This shows that when they are
sponsored directly by EPA,extraordinary efforts to ensure the timeliness
of data are possible.

Similar arrangements, however, were not made to ensure that the
results of the generator survey were made available in a timely manner.


Page 90             GAO/PEMD-90-3        EPA’s Hazardous   Waste Data Need Further   Improvement
                       Chapter 5
                       Aseeesmena ctf Data Collection   Methods
                       andProcedm




                       Consequently, the results of the generator survey will not be ready for
                       use by states in preparing the capacity assurances required by SARA.EPA
                       officials explained that the 1987 RCRAreporting cycle prevented them
                       from expediting the generator survey. However, because there are no
                       other data of comparable quality, this unavailability of the generator
                       survey results will have unfortunate consequences for the quality of
                       state assurances of capacity under SARA.


                       We examined the implications of the just discussed data collection prob-
Implications for       lems for the three major capacity analyses that must be conducted using
Cap&city Assessments   data produced by the 1987 RCRAreporting system and the national
   ,                   surveys.

                       The capacity analyses performed entirely by EPA (the integrated capac-
                       ity analysis and those performed for the land disposal restrictions) rely
                       entirely on the national surveys of generators and management facili-
                       ties. The foregoing assessment of data collection methods shows mini-
                       mal problems stemming from data collection. Some missing data on
                       waste characteristics stemming from lack of federal reporting and
                       recordkeeping requirements may affect the treatability analyses that
                       can be performed on the basis of these surveys. The fact that small
                       quantity generators were excluded may also affect these analyses
                       somewhat.

                       SARAcapacity assurances, however, must rely on some information pro-
                       vided by the RCRAreporting system. The survey of management facili-
                       ties data has now been made available to the states to use in developing
                       the required capacity analysis. Therefore, the information on the
                       amounts and characteristics of wastes must come from the RCRAreport-
                       ing system. However, because the 1987 RCRAdata collection instrument
                       was not used by all states, EPA cannot require states to use the informa-
                       tion provided by the instrument in preparing the capacity assurances. In
                       addition, since some states may not have the results of the 1987 RCRA
                       reporting cycle completed in time, EPA is allowing states to use 1986 or
                       1985 data if the state believes it is of higher quality than the 1987 data.
                       The result is that the capacity data will cover 1986 for all states while
                       the data on waste generation will vary between 1985 and 1987 and will
                       be obtained from data collection instruments with the different sources
                       of error discussed previously. As already has been shown, the data pro-
                       vided in past reporting cycles and instruments was seriously flawed.




                       Page 91              GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
Chapter 5
Assesswwmt of Data Collection   Methods
and Prwedures




EPA has provided states not using the revised   RCRA reporting system
instrument with default factors to apportion total quantities of each
waste code reported to specific treatment types. The default factors are
based on assumptions about the proportion of the reported volume of
waste that will require a specific type of management. The default fac-
tors use several other sources of information and attempt to represent
the typical distribution of treatability groupings. However, EPA acknowl-
edges that these factors do not reflect the treatability groupings of
wastes generated in a particular state. Instead, they reflect typical
national treatability groupings.

Two options for conducting the treatability analysis are provided for
states that use the revised instruments. The first option is more accurate
because it relies on the quantitative measures of waste characteristics
discussed in chapter 4. However, since handlers are not required to have
this data available, states may have to use the less precise qualitative
measures of waste characteristics.

In addition, since the national survey of management facilities covers
1986, states using generation data from other years will have to make
adjustments to the capacity information taken from the survey of man-
agement facilities in order for it to be consistent with the generation
data. The way these changes will be made will differ across states due to
the different years and data collection instruments involved.

The EPA project manager for the SARAcapacity assurance initiative
stated that the foregoing data limitations will make interstate agree-
ments difficult to achieve. Officials representing importing states may
not have sufficient confidence in the information to use it as the basis
for agreements that may appear as blanket acceptance of continuing
imports.

State and EPA experts believe the SARAdata collection plan is as good as
it can be, given the available data and the time frame. However, state
experts indicated that the level of detail in the assurances would not be
sufficient to enable waste management companies to determine
whether, or where, to locate additional capacity. This raises the ques-
tion of the level of detail and accuracy that is sufficient for the purposes
of SARAcapacity assurance. EPA did not address this important question
in the guidelines published in the Federal Register. Our work shows that
SARAcapacity assurances will be seriously limited by data collection
methods that resulted in the use of data from different instruments



Page 92             GAO/PEMDBO-3      EPA’s Haz+wdous Waste Data Need Further   Improvement
                            Chapter 5
                            Assessmenf- of Data Collection   Methods
                            and Procetlnsw




                            from different years across different states. These problems, in conjunc-
                            tion with the remaining data gaps and measurement problems discussed
                            previously, will seriously threaten the validity of the assurances that
                            can be provided. Due in part to these limitations, EPA is planning to
                            require that capacity assurances be submitted periodically.

    I
                            This section presents the results of our evaluation of the quality of the
Dat$ Collection in the      data likely to be produced by the data collection mechanisms used in the
Per+anent                   permanent information system. We assessed whether the issues affect,-
Infopnation System          ing data collection in the past or during the interim phase have been
                            resolved in the permanent information system as it was planned in
                            August 1988. As discussed in chapter 2, the primary difference between
                            the interim and permanent information systems is that the permanent
                            system will not include further national surveys but will continue to
    I                       include the more limited scope surveys contained in the original infor-
                            mation system. Thus, the RCRAreporting system will become EI~A'Sprin-
                            cipal mechanism for collecting regulatory development data. The
                            permanent information system will also be able to draw on the inform&-
                            tion provided by the toxic chemical release inventory reporting system.
                            These data collection mechanisms will have to support state implemen-
                            tation activities, additional capacity assurances, and all EPA regulatory
                            development activities.

                            We found six factors that are likely to affect adversely the quality of
                            the data provided by the permanent information system: (1) continued
                            lack of support for the RCRAreporting system, (2) phaseout of national
                            surveys despite continuing need, (3) lack of integration of permanent
                            data collection instruments, (4) incomplete federal recordkeeping and
                            reporting requirements for handlers, (5) inconsistent requirements for
                            state programs, and (6) the problematic relationship between EPA and
                            the states under RCRA.

                                                                                                                         -..
Continued Lack of Support   In the past, EPA has not provided sufficient funding for the IKRA report-
for tihe RCRA Reporting     ing system. EPA officials told the states that the agency could not guar-
                            antee additional funding for the 1989 cycle, despite its expanded size
Sysqem                      and importance. As discussed previously, EPA officials stated that EPA
                            pays 25 percent of the cost of the reporting system, although EPA gener-
                            ally funds required activities at the rate of 75 percent.




                            Page 93              GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
                     Chapter 5
                     Assessxwnt of Data Collection   Methods
                     and plli ‘wedures




                     In addition, despite the fact that EPA'Splan for data collection relies
                     heavily on incentives for states to collect and verify the data, few incen-
                     tives have been provided. State program experts indicated that one
                     important incentive would be to include specific indicators related to
                     data collection in the mechanism for monitoring state program perform-
                     ance. State experts indicated that since data collection is not among the
                     things specifically monitored (such as the number of inspections per-
                     formed) data collection and quality control suffer. However, if EPA were
                     to increase the funding for the RCRAreporting effort to 75 percent of the
                     estimated cost and include specific indicators of quality control work in
                     monitoring state performance, improved quality control would be likely.


Phaseout of EPA-     In the past, national surveys have been the primary source of useful
Sponsored National   national data that the agency has been able to obtain. The RCRAreport-
                     ing system has never produced complete and valid national data. We
Surveys              believe it is unwise to abandon EPA-sponsored national surveys without
                     having a proven alternative.

                     In addition, there are two factors that lead us to the conclusion that
                     combining intermittent EPA-sponsored national surveys with the RCRA
                     reporting system census will continue to be appropriate. First, the data
                     required for many activities are not required for every handler; only the
                     more limited data necessary for implementation purposes are required
                     for every handler. Appropriately stratified probability samples, such as
                     those discussed earlier, would be sufficient for collecting the data
                     required for capacity assurance and regulatory development. Second,
                     the amount of quality control work necessary to ensure the accuracy of
                     the data from a census is extensive and may exceed state and EPA
                     resources.

                     To minimize the regulatory burden, we believe the RCRAbiennial census
                     should collect only that information that is needed for ongoing routine
                     implementation purposes from every handler. These data are needed
                     primarily by the state programs and EPAregional offices. These data
                     also would enable the development of more refined stratified samples.
                     According to EPA’Splan in effect in August 1988, the permanent RCRA
                     reporting system was to collect data necessary for regulatory develop-
                     ment, As was discussed in chapter 2, this is not consistent with the two-
                     domain concept for structuring the RCRAreporting system because it
                     asks states to collect information that is used primarily by EPA.In addi-
                     tion, the 1987 RCRAreporting instrument does not provide the level of
                     detail that is necessary for regulatory development. That is, the level of


                     Page 94             GAO/PEMD-99-3     EPA’s Hazardous   Waste Data Need Further   Improvement
                          Chapter 6
                          Assessment trf’ Data Collection   Methods
                          and Proceduh 8~




                          detail in the revised instrument is greater than states need for imple-
                          mentation purposes and not great enough for many regulatory develop-
                          ment purposes. The effects of this policy concerning the RCRAreporting
                          system are (1) to shift a portion of the responsibility and the cost of
                          data collection for EPA'Sregulatory development activities to the states
                          and (2) to make EPA even more dependent on a diffuse and hard to man-
                          age system that has repeatedly failed to provide valid national data.

                          These problems would be corrected by combining the two types of sam-
                          ple surveys discussed previously with the census provided by the RCRA
                          reporting system. Three data collection instruments would then be used
                          during each cycle. The first would collect basic implementation data
                          from every handler in the full census conducted either by the states or
                          directly by EPA. The second would collect the additional information
                          required for capacity assurances from a sample designed to assure ade-
                          quate precision for each state. Finally, the third would collect the most
                          detailed and burdensome data required for regulatory development
                          from a smaller subsample sufficient to assure adequate precision only at
                          the national level. The result would be that the least burdensome data
                          would be collected in the census and the most burdensome from the
                          smallest national sample, with the sample portions of the system being
                          conducted entirely by EPA. This alternative would incorporate the best
                          features of EPA'S 1982 proposal as discussed in chapter 2.

                          Based on the foregoing factors, we believe that a streamlined biennial
                          census focusing on the implementation data needed from each handler,
                          in conjunction with simultaneous sample surveys sponsored by EPA,
                          would be the most effective and efficient approach to data collection.
                          Such an approach would reduce the extensiveness of quality control
                          problems and reduce reliance on a diverse and hard to manage reporting
                          system. It would also avoid the complicated and expensive process of
                          developing entirely separate national surveys. The data elements could
                          also be relatively standard from cycle to cycle (with only the sample
                          changing), thus eliminating the problem of potential inconsistency from
                          one questionnaire to the next. Used in conjunction with on-site verifica-
                          tion from a small subsample, such an approach would ensure collection
                          of the valid information needed to implement the program called for by
                          the Congress.


Lack of Integrated Data   The different data collection mechanisms included in the permanent
                          information system have not been designed to work together as an inte-
Collection Mechanisms     grated information system. Specifically, EPA has not designed the toxic


                          Page 95               GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
Chapter 5
Aesessmcnt of Data Collection   Methods
and Prowdures




chemical release inventory reporting system so that it complements the
revised RCRAreporting system. This failure to integrate mechanisms
results in duplication and lost opportunities for maximizing available
information.

The information provided by the toxic chemical release inventory
reporting system will not be consistent with the other data collection
mechanisms we have discussed. The National Governors’ Association
has concluded that a consequence of this lack or integration may be a
duplication of effort that still will not produce results that are compar-
able to, or useful in conjunction with, other data collected by EPA.

First of all, there is a significant overlap between the reporting popula-
tions for the toxic chemical release inventory reporting system and the
RCRApopulation. Avoiding duplication requires careful coordination
between the reporting systems to ensure that members of both popula-
tions are not required to provide the same information twice. EPA has
already developed standard techniques for avoiding duplication that
involve including a question on one instrument that asks if the facility
has completed the other. If the response is “yes,” only the nonduplica-
tive information is required. However, this technique cannot be used
effectively on account of a second problem-that      is, that the instru-
ments used in the two systems measure some of the same attributes in
different ways. These differently measured attributes include the waste
characteristics that determine appropriate management methods, the
management methods actually used, and some of the information on
waste minimization. In this way, information that is essentially duplica-
tive is rendered noncomparable. However, the standardized measures
discussed in chapter 4 would help eliminate this problem.

In addition to duplicating some information, an important opportunity is
missed here for using the results of the two systems to complement each
other. The toxic chemical release inventory reporting system will obtain
information that the RCRAdata collection mechanisms do not that could
be useful for regulatory development. SARArequires the system to obtain
information on the amounts of specific chemical substances and on the
efficiency of the methods used to treat wastes. Risk assessments require
detailed data on the concentration of toxic constituents of waste streams
and, although not essential, information on treatment efficiencies would
be useful. Data on the concentrations of toxic substances in RCRAregu-
lated wastes would also be helpful to enforcement officials. Although
the toxic chemical inventory reporting system provides this information,
it cannot be used to supplement other RCRAinformation because it is not


Page 96             GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
                           Chapter 5
                           Assessment of Data Collection   Methods
                           and Procedurtbn




                           cross-referenced to the total quantity of the RCRA regulated wastes in
                           which these chemicals are found. Consequently, EPA cannot use the data
                           in conjunction with data collected on RCRA regulated wastes. As a
                           National Governors’ Association evaluation of the final regulation
                           states: “Given the current designs of both the . . . [toxic chemical release
                           inventory] and .. . [the revised RCRA reporting system instrument], an
                           important informational advantage . . . is being passed UP.“~


Inco*plete Recordkeeping   According to existing federal regulations, hazardous waste generators
and Ikeporting             are not required to perform any tests to determine either whether a
                           waste is hazardous or what type of waste it is unless information from
Requiirements              other sources (such as product labels or the RCRA waste code descrip-
                           tions) is not sufficient to make this determination. Generators are, how-
                           ever, required to keep records of any tests that are conducted, although
                           these tests are not standardized. Conversely, management facilities must
                           perform extensive tests (if generators do not) in order to determine
                           appropriate management methods. Management facilities are also
                           required to keep records of the results of tests that are performed. How-
                           ever, no standard tests are required. In addition, management facilities
                           are not required to return copies of the test results to generators.

                           The foregoing regulations are responsible for important data collection
                           problems discussed earlier in this report. Handlers who do not conduct
                           the tests necessary to provide the requested quantitative data on waste
                           characteristics included on new and revised data collection instruments
                           cannot be required to provide the information. As previously discussed,
                           the qualitative data EPA sought on the new and revised measurement
                           instruments do not require additional testing. However, these qualita-
                           tive measures are much less accurate than quantitative measures. This
                           problem would be corrected if EPA standardized the tests management
                           facilities are already required to perform, and required management
                           facilities to return copies of the tests to generators. We believe this
                           would be a relatively minor change to current regulatory requirements
                           and would produce a large increase in useful information available to
                           EPA, state programs, the Congress, and the public. Such recordkeeping
                           requirements would also enhance EPA'S ability to collect information on
                           surveys, reduce the complexity of data collection instruments, reduce


                           “National Governors’ Association, “Comparison of Final Toxic Chemical Release Inventory Reporting
                           Form With 1987 Biennial Hazardous Waste Report Questionnaire Package” (Paper provided to the
                           Annual/Biennial Hazardous Waste Report Advisory Council, June 1988), p. 2.



                           Page 97              GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further    Improvement
Chapter 6
Assessment of Data Collection   Methods
and Prowdures




the amount of quality control required, and greatly facilitate on-site
validation.

EPA could use the authority provided by section 3007 of RCRAto collect
samples of wastes and other information from handlers directly. If EPA
used this authority, it would not need to develop additional recordkeep-
ing regulations. However, it would be unreasonable for EPA to rely on
inspections for primary data collection activities since this could reduce
resources available for enforcement. EPA also has the authority to pro-
mulgate additional recordkeeping regulations that would require every
handler to keep all records that EPA might need in the future. However,
this approach seems unreasonably burdensome and inefficient since EPA
might never need much of the information and, in any case, much of the
information would not be required from every handler.

The establishment of two kinds of recordkeeping requirements could
solve the problem. The first kind would pertain to those records that,
like existing regulations, are required to be kept routinely by every han-
dler of a particular class. This type of regulation, as discussed earlier,
should standardize the tests that management facilities are already
required to perform, and require generators to retain copies of these
tests.

The second type of regulation would mandate additional records, data
elements, or tests that EPA may require. However, no individual handier
would be required to perform these tests or keep records of these data
elements unless directed to do so by EPA for a specific time and purpose.
This would enable EPA to collect the needed data from a sample of the
regulated population without requiring every handler to routinely main-
tain records that are only needed from some handlers some of the time.

These regulatory changes would improve both the RCRAreporting sys-
tem and national surveys. In addition, since failure to provide the
requested information would be a clear violation of federal regulations,
EPA'S ability to enforce reporting requirements would be greatly
enhanced. We believe that such a regulation is necessary to ensure
timely and accurate national information. If EPA finds it does not have
authority under current law to develop such a regulation, the agency
should seek legislative refinement.




Page 98              GAO/PEMB99-3     EPA’s Hazardous   Waste Data Need Further   Improvement
                            chapter 5
                            Assessment of Data Collection   Methods
                            and Procedure




Incoasistent Requirements   Federal regulations governing the RCRAreporting system remain incon-
for Sqate Programs          sistent with EPA'Sexpectations for states. In the permanent system, EPA
                            expects the states to submit disaggregated data rather than aggregated
                            summaries of the data. However, existing federal regulations explicitly
                            require states to submit summaries rather than disaggregated data. We
                            believe federal regulations should be modified so that they are consis-
                            tent with EPA'Sexpectations and needs. In addition, existing federal reg-
                            ulations, which provide that state programs shall develop reporting
                            requirements that are equivalent to federal rules, do not require states
                            to collect data identical to those contained in EPA'Sdata collection instru-
                            ment. Yet, without such a requirement, EPA has limited ability to ensure
                            that uniform data elements are actually collected.

                            Based on the foregoing factors, we believe EPA'S 1982 assessment of the
                            KCRAreporting system remains valid; that is, the variety of data collec-
                            tion and processing systems used by the states will probably continue to
                            preclude timely and efficient data analysis. We believe that an effective
                            strategy will require modification of federal regulations so that they are
                            consistent with EPA'Sexpectation that states collect identical data ele-
                            ments and submit the data in a disaggregated form. However, requiring
                            states to collect specific data elements may conflict with the process for
                            authorizing and revising state programs required by RCRA.


The @elationship Between    EPA'S ability to ensure that states collect identical data in a timely man-
EPA and the States Under    ner is limited by RCRAitself. EPA must authorize a state program if its
nfin                        statutes and regulations are equivalent to, consistent with, and not less
nu3A                        stringent than applicable federal standards. Accordingly, RCRAdoes not
                            require state programs to be identical to the federal program, and state
                            data collection efforts are subject to this same general requirement.
                            However, in order for technical data elements to be consistent with one
                            another, they must be identical, Therefore, we conclude that RCRAdoes
                            not provide a clear basis for EPA to require states to collect identical data
                            in a timely manner.

                            EPA might successfully argue that in order to be consistent, data ele-
                            ments must be identical. In fact, EPA has done this to a limited extent.
                            Based on the need for consistency, federal regulations (1) require state
                            programs to use a national manifest form with required federal data ele-
                            ments and (2) prohibit states from requiring additional information.
                            However, as discussed in chapter 2, the manifest form contains numer-
                            ous optional state data that continue to differ across states and, in part,
                            prevent the manifest from being the primary data collection instrument


                            Page 99              GAO/PEMD-90-3
                                                            EPA'sHazardous   Waste Data Need Further   Improvement
chapter 6
Assessment of Data Collection   Methods
and Procedures




EPA originally intended it to be. Furthermore, requiring states to use a
standard data collection instrument would not fully solve the problem.
There is insufficient time between reporting cycles for EPA to issue new
rules to make necessary modifications to the instrument (or supporting
recordkeeping rules) and for states to modify their regulations accord-
ingly. A streamlined state census in conjunction with EPA-sponsored
sample surveys, as discussed earlier, would minimize but probably not
eliminate this problem. Finally, EPA is limited in its ability to compel
authorized states to collect standard data.

EPA can withdraw authorization but this is time-consuming and disrup-
tive. The EPArepresentative of the General Counsel’s office with whom
we discussed this issue stated that EPA would not withdraw a state’s
authorization if the state did not collect identical data elements because
it would be too disruptive for the overall national program. He stated
that valid national data are vital but not worth the disruption that
would be caused by withdrawing state authorizations.

RCRA  does provide EPA with another alternative. EPA could promulgate
federal regulations that would enable the agency to conduct a periodic
census on its own authority in authorized states. However, EPA unsuc-
cessfully tried a similar approach in 1982. The agency received critical
comments when it proposed to conduct a biennial sample survey under
its own authority. Those people who filed comments with EPA believed
that an overall increase in regulatory burden would occur because many
states would continue to maintain duplicate reporting systems under
their own authority.

The result of the current arrangements is that the federal information
system must be pieced together from separate state systems rather than,
as suggested by RCRA,having the states add data to a minimum, consis-
tent federal system to support their own more stringent rules or unique
needs and interests. Available remedies either limit EPA'S ability to
obtain consistent data in a timely manner or make EPA appear to be add-
ing additional information to state systems.

We considered two possible remedies to this impasse, both of which
would require legislative refinements. First, EPA would be more able to
ensure the collection of consistent data without causing major disrup-
tions if it had authority under RCRAto withdraw authorization for a spe-
cific program activity (rather than only for the program as a whole).
Such authority would also be consistent with the way EPA initially
grants authorization; that is, authorization is granted on the basis of


Page 100            GAO/PEMD-90-3     EPA’s Hazardous   Waste Data Need Further   Improvement
              Chapter li
              Assessmerk? I,?+’
                              Data Collection   Methods
              and Procedm lb




              specific program activities that correspond to specific federal regula-
              tions. However, this possible remedy has a major limitation in that it
              would increase the potential for EPA to dictate all program elements to
              the states.

              The second remedy we considered was to place the recordkeeping and
              reporting provisions of RCRA that enable EPA to develop an information
              system in a section of the act separate from those that authorize sub-
              stantive standards applicable to hazardous waste handlers. Accordingly,
              federal recordkeeping and reporting requirements would take effect in
              authorized states upon promulgation in the way that HSWArequirements
              now do.7 This remedy would solve all four of the foregoing problems-
              consistency, duplication, timeliness, and enforcement-since the new
              information system would be independent of state authorization. EPA
              would then be responsible for conducting the national RCRA reporting
              system with its own resources under its own authority in all states. EPA
              could delegate the responsibility for operating the system to a state, but
              the delegation of this authority would be independent of authorization.
              States would retain the authority to add data elements that reflect dif-
              ferent program needs or more stringent regulations but not to modify
              federal data elements. States could also continue to collect data more
              frequently and continue to use their current supplemental data collec-
              tion mechanisms (such as manifest tracking systems), but there would
              be only one basic national RCRA reporting system. In effect, this would
              locate control of the national information system in EPA yet allow states
              the same flexibility to collect additional information needed to support
              their own more stringent regulations or unique needs and interests.
              Some increased reporting burden might occur in states that have exten-
              sive supplemental data collection mechanisms. However, this is consis-
              tent with RCRA, which permits state requirements to be more stringent
              than federal requirements.


              EPA data collection methods and procedures have improved in important
Summary and   ways. The most recent national surveys used accepted stratified sam-
Conclusions   pling techniques, and the same data collection instrument was used for
              each respondent sampled. These are basic and sound methods of data
              collection. However, because of the continued use of different data col-
              lection instruments among the states, the 1987 RCR.A reporting cycle can-
              not produce complete and valid data. Some of the same problems that

              7EPA was able to administer the waste minimization portion of the 1987 instrument in all states
              because it was mandated by the 1984 amendments.



              Page 101              GAO/PEMD-SO-3     EPA’s Hazardous    Waste Data Need Further    Improvement
ChapterIS
Assessment of Data Collection   Methods
and Procedures




characterized previous reporting cycles continue to affect data from
states using old instruments. In addition, the new information requested
on the revised instrument is not fully supported by federal regulations.

Data collection problems will not seriously jeopardize the two types of
capacity analyses performed directly by EPA. The integrated capacity
assessment and the assessments performed to support the land disposal
restrictions rely entirely on the national surveys, which had minimal
data collection problems. However, SARAcapacity assurance will be seri-
ously impaired because of its reliance on the RCRAreporting system and
the fact that data from different years and different data collection
efforts will be used. Necessary data will be entirely missing in states
that did not use the revised RCRAreporting system instrument, and pre-
vious measurement problems will persist. EPA has attempted to compen-
sate for the missing data by providing states with assumptions based on
engineering judgments. However, EPA acknowledges that these do not
reflect actual state conditions. The resultant uneven quality of the data
will seriously weaken confidence in the SARAcapacity assurances.
Although state and EPA experts we interviewed believed the capacity
assessments will be as good as possible given the available data and the
time frame, state experts indicated that the level of detail will not be
sufficient for waste management companies to determine whether or
where to locate additional capacity.

Our overall conclusion is that data collection problems will threaten the
quality of the data produced by the permanent information system. EPA
has not supplied the same 76 percent of the funding for the RCRAreport-
ing system that it provides for other required program activities, nor
has it provided effective incentives to the states, even though EPA
intends the system to be the single routine mechanism for providing reg-
ulatory development data. Rather than conduct further national
surveys, the agency plans to have states collect regulatory development
data needed by EPA in the RCRAreporting system-this despite the fact
that the RCRAreporting system (1) has never produced valid and com-
plete national data, (2) sample surveys in conjunction with the RCRA
reporting system would be more efficient and effective, and (3) the
revised instrument is more detailed than necessary for implementation
purposes (although less detailed than needed for regulatory develop-
ment). We believe it is unwise to abandon EPA-sponsored national
surveys without a proven alterative. In addition, the toxic chemical
release inventory reporting system has not been designed to complement
other data collection mechanisms that provide information on hazardous



Page 102             GAO/PEMD-SO-3        EPA’8 Hazardous   Waste Data Need Further   Improvement
     : ,

 .
                      Chapter 6
                      Assessment 4’ 0ata Collection    Methode
                      and Procedurer




                      waste generation and management. Moreover, existing federal regula-
                      tions are inconsistent with EPAexpectations for state data collection and
                      do not support the collection of new information. Finally, EPA'Sability to
                      require states to collect standard data is limited by the authorization
                      procedures established in RCRA.The result of the current arrangements
                      is that the federal information system must be pieced together from sep-
                      arate state systems rather than, as RCRAsuggests, having the states add
                      data to a minimum, consistent federal system.

                      EPAofficials generally agreed with our findings and stated that the
                      agency has already taken steps that at least partially address some of
                      our findings. These steps include increasing the role of limited scope
                      sample surveys8, planning to begin revision of the federal regulations in
                      the fall of 1989, and creating a more streamlined RCRAreporting instru-
                      ment. Since these actions were taken after we finished our field work,
                      we could not evaluate them for this report.


                      We recommend that the Administrator of EPAdirect the Assistant
Recchmendations       Administrator for Solid Waste and Emergency Response to

                  l ensure that state data collection and quality control efforts receive fully
                    adequate support and include specific indicators related to data collec-
                    tion and verification in the agency’s mechanism for monitoring state
                    performance;
                  . use probability sampling rather than a census of waste handlers when-
                    ever feasible for routine national data collection and quality control to
                    ensure that EPAobtains the information necessary to develop regulations
                    efficiently and without unnecessary data collection burden;
                  . ensure that the toxic chemical release inventory reporting system com-
                    plements other hazardous waste data collection efforts so that the data
                    it provides on toxic chemical concentrations can be used to their maxi-
                    mum potential; and finally,
                  . amend federal recordkeeping and reporting regulations so that states
                    are required to collect and provide standard data elements in a disaggre-
                    gated form and hazardous waste handlers are required to provide suffi-
                    ciently detailed data.




                      sEPA is not now planning further national surveys like the current surveys of generators and man-
                      agement facilities.



                      Page 108             GAO/PEMD-90-S      EPA’s Hazardous   Waste Data Need Further    Improvement
                Chapter 5
                Assessment   of Data Collection   Methods
                andPromdures




                In addition to the improvements EPA can make, we believe a refinement
Matters for     in legislation may also be necessary to improve the quality of EPA’S
Cbngressional   information. Nonuniform data and procedures across the states, which
Cbnsideration   are associated with a joint federal-state data collection effort under
                RCRA,degrade the quality of information about hazardous waste. Under
                current law, responsibility for data collection, as well as other regula-
                tory activities, is shared by federal and state governments. This problem
                could be corrected by separating the recordkeeping and reporting provi-
                sions of the act from other regulatory provisions and making EPA solely
                responsible for collecting the information required for developing and
                implementing the federal program. Uniform national data would then be
                assured, but states would retain the authority to add data elements and
                to use supplemental data collection mechanisms to support state needs.




                Page 104              GAO/PEMNO-3EPA’sHazardous WasteData   Need Further   Improvement
Page 105   GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
Apbendix I                                                                                          ..

Panel of State Experts


                 Kathleen C. Gob, Chairwoman and Executive Officer
                 Connecticut Hazardous Waste Management Service

                 -._   .-
                 Ron Gr:rhm
                 Tennessee Division of Solid Waste


                 Joyce Noel
                 Maine, Department of Environmental Protection


                 Kim Wilhelm
                 California, Department of Health Services


                 Nancy Power, Chief
                 New Jersey, Bureau of Manifest and Information Systems
                 Department of Environmental Protection




             Y




                 Page 106       GAO/PEMD-30-3   EPA’s Hazardous   Waste Data Need Further   Improvement
      ;   ‘L




Appeqdix II

M & r Contributors              This Report


                    Boris Kachura, Assistant Director
Pro$am Evaluation   David F. Scudder, Project Manager
and/ Methodology
Division




                     Page 107       GAO/PEMD-90-3   EPA's I-kuardo~~   Waste Data Need Further   Improvement
Glossary


Authorized State            RCRAencourages EPAto authorize states to operate their own hazardous
                            waste programs in lieu of RCRA.In order to obtain authorization, states
                            must develop laws and regulations that are consistent with, equivalent
                            to, and not less stringent than the provisions of federal law and regula-
                            tions. S;tates are not required to have identical program elements and
                            are expl&ritly permitted to have more stringent requirements as long as
                            they do not interfere with interstate commerce. EPA operates the RCRA
                            program in unauthorized states.


Be& Demonstrated            A term used by EPAto refer to the technology-based performance stan-
Avbilable Technology        dards required by HSWAfor the treatment of hazardous waste restricted
                            from land disposal.


Biennial Report             The term used by EPAto refer to any or all of the three tiers of the
                            reporting system for collecting hazardous waste generation and manage-
                            ment data under sections 3002 and 3004 of RCRA.RCRAdoes not require
                            EPAto publish a report on the biennial census. In this report, we refer to
                            this system as the RCRAreporting system.


Classification System       See Qualitative Variable. Also see General Classification System.


Commercial Status (Waste    Designates treatment, storage, disposal, or recycling facilities that
                            accept wastes on a commercial basis from facilities not under the same
Management Facility)        ownership.


Cross-Media Management      A comprehensive EPA effort to track and manage the effects of pollut-
                            ants across all environmental media (including air, water, and soil) and
Initiative                  EPAprogram areas.


Data Collection Mechanism   We define a data collection mechanism as a component of an informa-
                            tion system designed to collect technical data, which embodies scientific
                            measurement instruments and methods of data collection. The federal
                            data collection mechanisms evaluated in this report also must be autho-
             ”              rized and fully supported by applicable federal and state laws and
                            regulations.




                            Page 108        GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
Desigr Capacity           The amount of hazardous waste a unit of equipment is technically
                          designed to process. This may differ from the amount the unit is allowed
                          to process according to its operating permit.


Gene$al Classification    One of two basic types of classification systems: special and general. A
Systei-n                  special classification system contains discrete values of one attribute
                          that are mutually exclusive and exhaustive so that each individual in a
                          category is equal to the others, no individual can be classified in more
                          than one category, and all individuals can be classified. A measure of
                          religion (such as Protestant, Catholic, Jewish, Other, None) is an exam-
                          ple of a special classification system. These systems are also commonly
                          referred to as discrete or qualitative variables, or nominal scale meas-
                          ures. In contrast, general classification systems are taxonomies. In addi-
                          tion to being mutually exclusive and exhaustive, they are hierarchical
                          and order all attributes of a class of objects. The additional criteria of a
                          general classification system mean that there are subtypes within major
                          types and that no category can be a subtype of more than one major
                          type.


Hazardous Waste Handler   The owner or operator of any site or facility that generates, transports,
                          stores, treats, disposes of, or recycles hazardous waste. Specific federal
                          regulations exempt some sites and facilities from some or all regulation.
                          These exemptions include businesses that generate very small quantities
                          of hazardous waste, facilities that store hazardous waste for fewer than
                          90 days, and recycling facilities.


Hazardous Waste           The use of specific technologies and procedures for waste minimization,
Management                treatment, storage, disposal, or recycling.



Interim Hazardous Waste   The results of EPA improvement efforts implemented between 1987 and
Information System        1990. The principal data collection mechanisms in the interim phase
                          include the partially revised RCRAreporting system and the national
                          surveys of hazardous waste generators and management facilities.


Information System        The organizational subsystem in which the observation, recording, stor-
                          age, retrieval, transmittal, analysis, and presentation of information
                          occurs. The information system includes both formal and informal, as


                          Page 109         GAO/PEMD-99-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                         Glossary




                         well as automated and manual, systems. Our evaluation focused on the
                         generation and management portion of EPA'Slarger formal hazardous
                         waste information system.


Information System       A speL 1F’icpart of a larger information system. Typical components
                         include computerized data bases and data collection mechanisms.
Ckjmponent

Mgnagement Capacity      Refers to the capacity to treat, store, recycle, and dispose hazardous
                         waste. Capacity refers to individual units of equipment, specific man-
                         agement technologies conducted in the equipment, and the linkage of
                         units of equipment into systems that apply more than one management
                         technology in series.


Mbnagement Facility      A facility that treats, stores, disposes of, or recycles hazardous waste.
                         Generators who store hazardous waste for fewer than 90 days before
                         shipping them off-site for further storage, treatment, disposal, or
                         recycling are not considered management facilities.


M/tnagement Technology   A single treatment, storage, disposal, or recycling technique applied to a
                         waste in one or more management units.


Management Unit          A single piece of equipment in which one or more management technolo-
                         gies occur. Incinerators, tanks, distillation towers, waste piles, surface
                         impoundments, and landfills are considered management units.


Management System        Multiple management technologies applied to a waste in one or more
                         management units. In some cases, EPA has designated specific types of
                         management units (such as incinerators) as systems.


Manifest                 A control and tracking document that accompanies all transportation                  of
                         hazardous waste. It identifies the volume and type of wastes being
                         transported from one facility to another, as well as the management
                         technologies to which it should be subjected prior to final disposal.




                         Page 110        GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
  I   :

                                      .._

                           GlOSSary




No M igration Variance     The 1984 amendments or RCRAbanned the land disposal of hazardous
                           wastes unless the wastes are treated to reduce their toxicity or mobility
                           or unless it can be shown that the disposal unit completely prevents
                           m igration of the waste from the unit. In cases where a land disposal unit
                           is shown to prevent m igration, a variance may be given to allow the unit
                           to accept hazardous wastes. Such a variance is referred to as a no m igra-
                           tion variance.


Permanent Hazardous        The results of EPAimprovement activities that will be fully implemented
Waste Information System   after 1990. The data collection mechanisms in the permanent phase
                           include the fully revised RCRAreporting system and the toxic chemical
                           release inventory reporting system.


Perm it Status             Designates the type of authorization EPAhas granted the facility for the
                           management of hazardous waste. Facilities must comply with either
                           interim perm it standards (40 CFR 265) or final perm it standards (40
                           CFR 264).


Qualitative Variable       A discrete variable or nominal scale measure composed of categories
                           that measures the types of a larger class of attributes. The resulting
                           measure is also referred to as a classification system.


Quantitative Variable      A continuous variable that measures the extent to which a single attri-
                           bute is possessed by developing a metric with equal intervals.


RCRA Reporting System      A term used in this report to refer to the national system for collecting
                           data on hazardous waste generation and management authorized by sec-
                           tions 3002 and 3004 of RCFLA. EPArefers to this system as the biennial
                           report.


Residual Waste             The wastes remaining after treatment by a specific treatment technol-
                           ogy or process, such as incinerator ash or sludge from settling tanks.


Treatability Analysis      A type of analysis performed to determ ine the most appropriate treat-
                           ment technology or technologies for a given type of waste. It is based on



                           Page 111         GAO/PEMD-90-3
                                                       EPA'sHasxrdous   Waste Data Need Further   Improvement
                                                                                                    ,



                     Glossary




                     the physical and chemical form of a quantity of waste and the concen-
                     tration of certain hazardous and nonhazardous constituents.


Wdste Minimization   Refers to practices to reduce generation and/or the recycling and treat-
                     ment of wastes that lead to overall reductions in the volumes of hazard-
                     ous waste that ultimately enter the environment.




                     Page 112        GAO/PEMD-90-3   EPA’s Hazardous   Waste Data Need Further   Improvement
Bibliography


               Blalock, Hubert M. Measurement in the Social Sciences: Theories and
               Strategies. Chicago: Aldine Publishing Company, 1974.

               Blokdijk, Andre, and Paul Blokdijk. Planning and Design of Information
               Systems. New York: Academic Press, 1987.

               Ghiselli, Edwin E., John P. Campbell, and Sheldon Zedeck, Measurement
               Theory for the Behavioral Sciences. San Francisco: W. H. Freeman and
               Company, 1981.

               McKelvey, Bill. Organizational Systematics: Taxonomy, Evolution, Clas-
               sification. Berkeley: University of California Press, 1982.

               Martin, James. Strategic Data-Planning Methodologies. Englewood Cliffs,
               New Jersey: Prentice-Hall, Inc., 1982.

               Namboodiri, N. Krishnan, Lewis F. Carter, and Hubert M. Blalock, Jr.
               Applied Multivariate Analysis and Experimental Designs. New York:
               McGraw-Hill Book Company, 1976.

               Office of Management and Budget. “Circular A-l 23. Internal Control
               Systems (Revised).” Washington, D.C.: Executive Office of the President,
               Publications Office, August 1986.

               Office of Management and Budget. “Circular A-130. Management of Fed-
               eral Information Resources.” Washington, D.C.: Executive Office of the
               President, Publications Office, December 1985.

               President’s Council on Management Improvement and President’s Coun-
               cil on Integrity and Efficiency. Model Framework for Management Con-
               trol Over Automated Information Systems. Washington, D.C.: US.
               Government Printing Office, 1988.

               Senn, James A. Analysis and Design of Information Systems. New York:
               McGraw-Hill Book Company, 1984.

               Trute, Barry, Bruce Tefft, and David Scuse. Human Service Information
               Systems: How to Design and Implement Them. Lewiston, New York: The
               Edwin Mellen Press, 1985.

               US. General Accounting Office, Evaluating Internal Controls in Com-
               puter Based Systems. Washington, D.C.: GAO, 1981.



               Page 113        GAO/PEMD-(IO-3   EPA’s Hazardous   Waste Data Need Further   Improvement
                                                                                       I,




           Bibliography




           U.S. General Services Administration. Strategic Information Resources
           Management Planning Handbook. Washington, DC.: U.S. Government
           Printing Office, 1987.

           U.S. General Services Administration. Information Systems Planning
           Handbook: Phase II. Washington, DC.: U.S. Government Printing Office,
           1985.

           Zeller, Richard A., and Edward G. Carmines. Measurement in the Social
           Sciences: The Link Between Theory and Data. London: Cambridge Uni-
           versity Press, 1980.




(978260)    Page 114       GAO/PEMD-PO-8   EPA’s Hazaxdow   Waste Data Need Further   Improvement
.
.   I.   ._   ._   .   ._   .__   -I   ____.__...._   ^.   _..   .._   ._   -   _.-   ..---   ..---   .--   -...   --..   ._-   “..“-   --..   1.1-1.   I.-   I
‘I’t+~~,lltme   20&27!x24   1