NATIONAL SURVEY OF HAZARDOUS WASTE
GENERATORS AND TREATMENT, STORAGE
AND DISPOSAL FACILITIES REGULATED
UNDER RCRA IN 1981
This publication was prepared by Westat, Inc
for the Office of Solid Waste
under contract no. 68-01-6861
U.S. ENVIRONMENTAL PROTECTION AGENCY
1984
-------
-------
ACKNOWLEDGEMENTS
The authors wish to acknowledge the assistance of
several individuals and organizations who have made significant
contributions to this difficult two-year study of hazardous
waste generation and management. This includes the review of
drafts and guidance during the study provided by Michael Burns
George Garland, Barry Stoll, and Marlene Suit of the Financial'
Requirements and Assessment Branch of the Office of Solid Waste's
Characterization and Assessment Division. Substantial
contributions were made during questionnaire design by the Waste
Management and Economics Division including members of the
Branch10 AHH^313 ?ranch' Land Disposal Branch, and Waste Treatment
of EP^s Of^ce^Po^cy: P^n^nnd^i^t^nf^ NaP°litan°
We would like to recognize the extensive production of
Land, Greg Faber and Tom Tebo. Our thanks to the Chemical
Manufacturers Association for their cooperation in sharing
results of our respective hazardous waste surveys.
^ <-v, i, aut>]ors would like to express particular gratitude
to the Westat production team under the direction of Arlene
Shykind, including the editors, particularly Carol Drew tvoists
and artists for their many long hours and weekends spent ^P18ts
producing the numerous drafts of this extensive report.
<->, Finally the authors would like to acknowledge the
other members of the study team who played an important part in
conducting this study and preparing and processing quality date
The principal study members are listed below, although there
periodsnof°time? ^ "^ lmp°rtant Contributions over shorLr
Study Team
Michael Burns
Stephen Dietz
Barbara Kreling
Carmen Vincent
Ralph DiGaetano
Judith Strenio
Daniel Tuttle
Thomas Jones
Daniel Ruffner
Dalia Kahane
Chongsoo Kim
Robin McEntire
Linda Cranston
Gail Vossler
Betty Hawes
Reina Sprankle
Pat Picket Tanco
EPA Project Officer
Westat Project Director
Survey Director
Data Preparation Manager
Statistical Design
Statistical Design
Analyst
Data Processing Manager
Lead Programmer
Data Processing
Data Processing
Data Processing
Senior Systems Analyst
Data Preparation
Data Preparation
Data Preparation
Hotline Supervisor
-------
-------
TABLE OF CONTENTS
Section
EXECUTIVE SUMMARY
PART I
INTRODUCTION ..........
n
y
1.1 Background on the RCRA Hazardous Waste
Regulatory Program ............................ 10
1 . 2 The Need for a National Survey ....... .*.*."."."."." ." 12
1 . 3 Scope and Foe as of the Survey ....... ......... 15
SURVEY METHODOLOGY .......................... 21
2.1 Questionnaire Design ..................... 21
2.2 Pre-Survey Screening: Development of the .....
Sample Frame . . ....................... 24
2.3 Sample Design ................... .......... 26
2 . 4 Data Col lee t ion .......... ""!**"**"""* ." ." ." .".".*.""" 31
2 . 5 Quality Control ................ .*.".".".".".*.*.*.*.*."*"* 32
2.5.1 Response Rate Quality Control .......... 32
2.5.2 Nonsampling Error Quality Control ...... 34
2.5.3 Editing of Large Quantity Cases in the*
Component Data Files ................... 35
2.6 Imputation Procedures Used for Quantity
Estimates .......... ^r
.......................... JO
STATISTICAL RELIABILITY AND DATA ACCURACY .......... 39
3.1 The Concepts of Sampling Error and Nonsampling
Error ........................................ 3g
3 . 2 Response Rates and Sample Sizes. ....*."." ." ." '. '. ." .* .* 40
3.3 Statistical Reliability and Sampling Error.".".*. 42
3.4 Nonsampling Error ................... * " 53
PART II
INTRODUCTION TO PART II: NUMBERS OF HAZARDOUS
WASTE GENERATORS AND MANAGEMENT FACILITIES ......... 55
NUMBER OF HAZARDOUS WASTE GENERATORS ............... 57
4.1 Regional Distribution of Generators ........... 61
4.2 Number of Generators by Industry Type... ...... 65
4.3 Number of Generators by Waste Group Generated." 66
4.4 Number of Generators Shipping Hazardous Waste
Off Site ............................... 69
4.5 Number of Generators Recycling Hazardous ......
Waste. ... .....................
v
-------
TABLE OF CONTENTS (continued)
Section Page
5 DEFINITION AND NUMBER OF MANAGEMENT FACILITIES 75
5.1 Regional Distribution of Hazardous Waste
Management Facilities 78
5.2 Number of Management Facilities by Industry
Type 82
5.3 Number of Commercial Management Facilities.... 83
5.4 Number of Facilities Treating, Storing, and/or
Disposing of Hazardous Waste 87
5.4.1 Number of Facilities Treating
Hazardous Waste, by Treatment
Process Type 92
5.4.2 Number of Facilities Storing
Hazardous Waste, by Storage
Process Type 97
5.4.3 Number of Facilities Disposing of
Hazardous Waste, by Disposal
Process Type 100
5.5 Miscellaneous Facility Characteristics 103
5.5.1 Age of Waste Management Facilities 103
5.5.2 Ownership Status of TSD Facilities 106
5.5.3 Operator Status of TSD Facilities 107
PART III
INTRODUCTION TO PART III: QUANTITIES OF HAZARDOUS
WASTE GENERATED AND MANAGED 109
6 QUANTITIES OF HAZARDOUS WASTE GENERATED 121
6.1 1981 Hazardous Waste Generation Estimates 123
6.2 Size Distribution of Generators 138
6.3 Quantities Generated by Industry Type 139
6.4 Quantities of Hazardous Waste Generated by
Type of Waste Group 143
6.5 Disposition of Quantities Generated: Managed
On Site Versus Off Site 144
6.6 Recycling of Quantities Generated by Location
of Recycling Facilities 151
6.7 Future Hazardous Waste Generation Regulated
Under RCRA 153
7 QUANTITIES OF HAZARDOUS WASTE MANAGED 159
7.1 Quantity of Waste Managed as Hazardous by TSD
Facilities in 1981 162
VI
-------
TABLE OF CONTENTS (continued
Section
Paqe
7.2 Size Distribution of TSD Facilities 166
7.3 Comparison of Quantities of Hazardous Waste
Generated and Managed 166
7.4 Quantities Managed by Industry Type .* ." 175
7.5 Quantities Managed by Type of Waste Group..".".*." 176
7.6 Quantities Managed by Commercial Management
Facilities \ ^33
7.7 Quantities of Hazardous Waste Treated, Stored
and Disposed During 1981 185
7.7.1 Quantities of Hazardous Waste Treated
in 1981, by Treatment Process Type 189
7.7.2 Quantities of Hazardous Waste Stored
in 1981, by Storage Process Type 197
7.7.3 Quantities of Hazardous Waste Disposed
in 1981, by Disposal Process Type 204
PART IV
CAPACITY OF TREATMENT, STORAGE AND DISPOSAL
FACILITIES
211
8.1 U.S. Capacity Utilization: Commercial and
On-Site Management Facilities 213
8.2 Regional Utilization of Existing Capacity..... 218
8 . 3 Unused Capacity by Region 220
8.4 Summary and Conclusions About Available
Capacity 224
PART V
NEXT STEPS: FUTURE HAZARDOUS WASTE STUDIES 225
9.1 Mail Survey Followup Activities 226
9.2 Updates and Expansions to Mail Survey Data
Base 227
9.3 Continuing Analysis of Mail Survey Data....... 227
9.4 Small Quantity Generators Study .".'.'.".*."." 228
9.5 Survey of Used Oil and Waste-Derived Fuel
Material 22o
9.6 Industry Studies * 229
9 . 7 RCRA Biennial Report ..^........... 230
VII
-------
Section
APPENDIX A:
APPENDIX B
APPENDIX C
APPENDIX D
TABLE OF CONTENTS (continued)
Page
SAMPLE DESIGN AND STATISTICAL RELIABILITY A-l
FIELD REPORT
SELECTED REGULATIONS IMPLEMENTED UNDER THE
RESOURCE CONSERVATION AND RECOVERY ACT OF
1976, AS AMENDED
B-l
C-l
SUMMARY OF MAJOR DATA ELEMENTS CONTAINED IN
COMPUTER DATA BASE FROM THE NATIONAL SURVEY
OF RCRA-REGULATED GENERATORS AND MANAGEMENT
FACILITIES
D-l
Vlll
-------
LIST OF TABLES
Table Page
1 Special Edit Coverage of the Largest Facilities
in Each Component Data File 37
2 Number of Eligible Responses and Response Rates to
the RCRA Survey 4 ±
3 Statistical Reliability of Estimates from the
Generator and TSD General Questionnaire 48
4 Number of Respondents and Statistical Reliability
of Estimated Percentages, by Process Type 51
5 "K" Factors to Be Used in Obtaining 95 Percent
Confidence Intervals for Population Subsets 52
6 Characteristics of Nonregulated Notifiers that
Responded to the Survey 59
7 Comparison of Notifier and Generator Distributions
in EPA Regions 62
8 Percent of 1981 Hazardous Waste TSD Facilities in
EPA Regions - By Rank Order 80
9 Percentage of Treatment Facilities and of All TSD
Facilities Employing Each Treatment Technology in
1981 96
10 Percentage of Storage Facilities Employing Each
Storage Technology in 1981 102
11 Percentage of Disposal Facilities and of all TSD
Facilities Employing Each Disposal Technology in
!981 :: 105
12 Age Distribution of TSD Facilities 106
13 Ownership Status of TSD Facilities 107
14 Operator Status of TSD Facilities 107
15 Quantity Distribution for Commercial and Other
Types of TSD Facilities 184
IX
-------
LIST OF TABLES (continued)
Table
16 Total Quantities Managed, Average Quantities
Managed per Facility and Number of Facilities
Treating, Storing and Disposing of Hazardous
Waste in 1981 187
17 Total Quantities Treated, Average Quantities
Treated and Number of Facilities Treating
Hazardous Waste by Each Treatment Process Type.... 192
18 Size Distributions in the Treatment Technologies:
Proportions of the Population Accounting for 33
Percent of the Quantities of Waste Treated in
Each Technology 193
19 Total Quantities, Average Quantities Stored and
Number of Facilities Storing Hazardous Waste
by Storage Process Type. 200
20 Size Distributions in the Storage Technologies:
Proportions of the Population Accounting for 33
Percent, 50 Percent and 99 Percent of the
Quantities of Waste Stored in Each Technology 203
21 Total Quantities Disposed, Average Quantities
Disposed and Number of Facilities Disposing of
Hazardous Waste by Each Disposal Process Type 207
22 Size Distributions in the Disposal Technologies:
Proportions of the Population Accounting for
33 Percent, 50 Percent and 99 Percent of the
Quantities Waste of Disposed in Each Technology... 208
23 Summary of 1981 U.S. "Commercial" and On-Site
Hazardous Waste Management Capacity Utilization... 214
24 Distributions of Unused Capacity Over Facilities
for Treatment, Storage and Disposal of Hazardous
Waste 217
x
-------
LIST OF TABLES (continued)
Table Page
A-l Summary of Statistical Aspects of the TSD
and Generator Surveys A-6
B-l Generator Questionnaire Return Status B-13
B-2 Eligibility Rate Among Respondents From the
Notifier File and the Part A/Verification File.... B-18
B-3 Return Status for TSD and Component Questionnaires. B-22
B-4 Eligibility Rates for the Treatment, Storage, and
Disposal Questionnaire and the Technology
Questionnaires B-23
XI
-------
-------
LIST OF FIGURES
Figure Page
1 Purpose of Survey 16
2 Survey Scope 19
3 Distribution of Management Facility Size 45
4 Comparison of Cumulative Distributions of Quantity
of Hazardous Waste Managed in 1981 and the Number
of Management Facilities 45
5 Portion of Notifiers That Generated RCRA-Regulated
Quantities of Hazardous Waste in 1981 60
6 Regional Distribution of Hazardous Waste Generators
in 1981 63
7 Number of Generators by Industry Type 65
8 Number of Establishments Generating Each Major
Waste Group 67
9 Number of Generators Shipping Hazardous Waste Off
Site in 1981 70
10 Number of Generators Recycling Hazardous Waste 72
11 Regional Distribution of Hazardous Waste Management
Facilities in 1981 79
12 Number of Management Facilities by Industry Type... 82
13 Number of Commercial Versus Other TSD Facilities... 86
14 Number of Facilities with Treatment, Storage, and/
or Disposal in 1981 89
15 Number of Facilities Treating Hazardous Waste in
1981, by Treatment Process Type 95
16 Number of Facilities Storing Hazardous Waste in
1981, by Storage Process Type 101
17 Number of Facilities Disposing of Hazardous Waste
in 1981, by Disposal Process Type 104
18 Overlap Among Populations of Generators and TSD
Facilities Regulated Under Subtitle C of RCRA 129
Xlll
-------
LIST OF FIGURES (continued)
Figure Page
19 Overlap Among Generator and TSD Samples and
Among Eligible Respondents 131
20 Estimates of the Quantity of Hazardous Waste
Generated in 1981 135
21 Quantities of Hazardous Waste Generated in 1981
by Industry Type 141
22 Waste Group Percentage Comparisons Between the
1981 Generator and TSD General Surveys 145
23 Disposition of Hazardous Waste Generated:
Quantity Managed On Site vs. Shipped Off Site for
Management 146
24 Hazardous Wastes Recycled in 1981 152
25 Quantity of Waste Managed as Hazardous Wastes by
TSD Facilities in 1981 164
26 Comparisons of Cumulative Distributions of
Quantity of Hazardous Waste Managed in 1981 and
the Number of TSD Facilities 167
27 Quantity Generated Versus Managed 169
28 Description of Quantities Generated by Generators
Without TSD Facilities and Quantities Received
from Off Site by TSD Facilities 172
29 Quantities of Hazardous Waste Managed in 1981 by
Industry Type 177
30 Quantities of Hazardous Waste Handled by Management
Facilities in 1981 by Type of Waste Group 180
31 Quantities of Hazardous Waste Treated, Stored and
Disposed in 1981 186
32 Quantities of Hazardous Waste Treated in 1981, by
Treatment Process Type 190
33 Quantities of Hazardous Waste Stored in 1981, by
Storage Process Type 198
xiv
-------
LIST OF FIGURES (continued)
Page
34 Quantities of Hazardous Waste Disposed in 1981,
by Disposal Process Type .......................... 205
35 1981 Average Facility TSD Capacity Utilization
Rates, by Region .................................. 219
36 1981 Regional Unused Treatment Capacity ........... 221
37 1981 Regional Unused Storage Capacity ............. 222
38 1981 Regional Unused Disposal Capacity ............ 223
59 Future Studies .................................... 232
B-l TSD Questionnaire Returns by End of Each Month
From September 17, 1983 Mailout. RCRA Required
Return Date: November 15, 2,599 Packages Mailed.. B-8
B-2 Notifier Generator Returns by End of Each Month
From September 17, 1983 Mailout. RCRA Required
Return Date: November 15, 10,667 Packages Mailed. B-9
B-3 Return Status Codes and Definitions ............... B-ll
B-4 Titles of the Nine Questionnaires that were Mailed
with the Treatment, Storage, and Disposal
Questionnaire .................................. B-21
xv
-------
-------
EPA
HWDMS
MMT
NEC
NPDES
NSK
OSW
Part A
POTW
PPS
RIA
RCRA
SIC
TSD
ABBREVIATIONS
Environmental Protection Agency
Hazardous Waste Data Management System
Million metric tonnes
Not elsewhere classified
National Pollutant Discharge Elimination System
Not specified by kind
Office of Solid Waste (EPA)
Part A of EPA's Consolidated Hazardous Waste Treatment,
Storage, and Disposal Facility Permit Application
Publicly owned treatment works
Probability of selection proportionate to size
Regulatory Impact Analysis (required under Executive
Order 12291)
Resource Conservation and Recovery Act of 1976, as
amended
Standard Industrial Classification (Code)
Treatment, storage, and disposal (facility)
xvi i
-------
-------
CONVERSION FACTORS
USED FOR HAZARDOUS WASTE
243.9025 Gallons per English (short) ton
268.8519 Gallons per metric tonne
0.00371952 Metric tonnes per gallon
0-9072 Metric tonnes per English (short) ton
1.102293 English (short) tons per metric tonne
0.0041 English (short) tons per gallon
xix
-------
-------
EXECUTIVE SUMMARY
-------
-------
EXECUTIVE SUMMARY
This report summarizes the findings of an extensive
national survey of hazardous waste generators and treatment,
storage, and disposal facilities regulated under Subtitle C of
the Resource Conservation and Recovery Act (RCRA) of 1976, as
amended. The survey, conducted by mail during the fall of 1982
and the spring of 1983, was administered by Westat under the
sponsorship and direction of the United States Environmental
Protection Agency's (EPA) Office of Solid Waste (OSW).
Preliminary findings of the survey were released by
EPA on August 30, 1983, in a package entitled: "Highlights of
Preliminary Findings, National Survey of Hazardous Waste Gener-
ators and Treatment, Storage, and Disposal Facilities Regulated
Under RCRA During 1981." Since that time, Westat, in conjunction
with OSW, performed additional editing of the obtained data,
leading to the development of Version II of the survey data
base. Additionally, the findings presented in this report are
based upon a revised approach for estimating total quantities of
hazardous waste generated. The information presented in this
report, therefore, supersedes information previously released
from the survey data base.
The information developed through the survey provides
EPA and other participants in the hazardous waste decisionmaking
arena with the first complete description of hazardous waste
generation and management activities regulated under Federal law
since enactment of RCRA by Congress in 1976. The survey produced
a statistically weighted data base containing more than 6,000
statistical data elements describing hazardous waste generation
and management activities in 1981, the study year. The findings
presented in this report are drawn from only a portion of these
-------
data, but are intended to provide a broad overview of the nature
and scope of the populations of regulated hazardous waste gener-
ators and treatment, storage, and disposal facilities and of the
hazardous wastes they generate and manage.
The survey's most important finding is that an esti-
mated 71 billion gallons (264 million metric tonnes) of hazardous
wastes were generated during 1981, more than six times previous
estimates of annual hazardous waste generation. It is important
to highlight the fact that this 71 billion gallon estimate differs
considerably from the preliminary survey findings released by
EPA on August 30, 1983. When the survey results were initially
tabulated, the quantity of hazardous waste generated during 1981
was preliminary estimated to have been 40 billion gallons (approx-
imately 150 million metric tonnes). Extreme statistical uncer-
tainty, however, was associated with that preliminary estimate.
At that time, EPA believed (and indicated in its public presenta-
tion of the preliminary findings) that the 40 billion gallon
estimate had a greater likelihood to understate rather than
overstate, the actual quantity of hazardous waste generated in
1981. Subsequent to the release of the preliminary findings,
further analysis of the data obtained through the independent
surveys of generators and the TSD facilities revealed that the
40 billion gallon estimate did indeed substantially understate
the actual quantity of hazardous waste generated. Accordingly,
an alternate approach was developed, using additional data
obtained through the survey, to estimate the actual quantity
generated, and resulted in the survey's final estimate of 71
billion gallons of hazardous waste generated in 1981.
Large portions of this quantity are mixtures of hazardous
and nonhazardous wastes (e.g., hazardous wastes mixed with indus-
trial process waters), although the survey data do not estimate
the actual amounts of such mixtures, nor the concentrations of
-------
specific constituents included in such mixtures. The mixtures
included by the survey are, however, hazardous wastes as defined
under RCRA, and are therefore properly included in the estimate
of hazardous waste generation.
It is important to note, however, that this survey was
not designed to estimate the quantity of all hazardous wastes
generated during 1981. The 71 billion gallon estimate includes
only those hazardous wastes generated in 1981 that were to be
managed in treatment, storage, and disposal processes regulated
under RCRA. Additional quantities of hazardous wastes were also
generated during 1981 that were treated, stored, or disposed of
in processes exempt from regulation under RCRA (e.g., hazardous
wastes treated exclusively in wastewater treatment tanks covered
under NPDES permits, which are excluded from regulation under
RCRA). Furthermore, additional quantities of wastes were gener-
ated in 1981 that have been specifically excluded by statute
and/or regulation from classification as hazardous waste, even
though they may exhibit characteristics of hazardous wastes
[e.g., wastes generated in conjunction with ore and minerals
extraction and beneficiation; wastes legitimately disposed of
through sewers to Publicly Owned Treatment Works (POTW's)j. The
survey was not designed to estimate these quantities, for reasons
explained in the sections that present the quantity estimates.
Prominent among the survey's other findings are the
estimated numbers of generators and treatment, storage, and
disposal (TSD) facilities regulated under RCRA during 1981. The
survey estimates that 14,098 installations generated RCRA-
regulated quantities of hazardous waste during 1981, compared to
more than 55,000 installations that have submitted generator
Notification forms to EPA pursuant to Section 3010(a) of RCRA.
Similarly, the survey estimates that 4,818 facilities treated,
stored (for more than 90 days), or disposed of hazardous wastes
-------
in processes regulated under RCRA during 1981, compared to the
more than 8rOOO facilities that currently have Part A permit
applications on file at EPA. The survey, however, purposefully
excluded specific classes of TSD facilities from its estimate,
as detailed in the report sections presenting the TSD facility
population estimates.
Analyses of data obtained by the survey indicate that
the distributions of the populations of hazardous waste generators
and TSD facilities are each highly skewed along their relative
size spectra. Specifically, each population includes a small
number of sites that account for a very large portion of the
quantities of hazardous wastes generated or managed. These
large sites dominate the quantity estimates provided in the
report, and their presence within the populations complicates
efforts to develop statistically reliable estimates of the quan-
tities of hazardous waste generated and managed annually. The
impact of these large sites on the reliability of the obtained
data is detailed in the report sections presenting the quantity
estimates.
The survey found that while most generators (84%)
shipped some or all of their hazardous wastes off site for treat-
ment, storage, and disposal during 1981, the overwhelming majority
of the quantity of hazardous waste (96%) was managed on site.
The report attributes this phenomenon to the fact that the larger
generators tend to manage their hazardous wastes on site, while
the more numerous smaller generators, for various reasons, ship
their wastes to commercial facilities for treatment, storage,
and disposal.
-------
The survey found that the number of generators that
recycle hazardous waste is increasing with time, with more than
50 percent of the 14,098 generators indicating that they intended
to recycle hazardous waste in years following 1981 (compared
with just over 40 percent recycling in 1980 and 43 percent re-
cycling during 1981). The survey also found, however, that only
a small portion of the hazardous waste generated in 1981 was
actually used, reused, recycled, or reclaimed, indicating that
in most cases, generators recycle only small portions of their
hazardous waste streams. It should be pointed out, however,
that many other generators are exempt from RCRA because they
recycle 100 percent of certain hazardous waste streams. The
survey purposefully excluded such generators and waste streams
from its estimates. Furthermore, the comparison of quantities
recycled with total quantities generated is skewed by the inclu-
sion of large quantities of hazardous and nonhazardous waste
mixtures in the data base. Frequently, only small portions of
such mixtures (e.g., the particular hazardous constituents) are
actually recyclable.
Hazardous waste generators and management facilities
were found to be concentrated in manufacturing industries.
Eighty-five percent of the 14,098 generators and 72 percent of
the 4,818 TSD facilities are estimated to be associated with
industrial manufacturing operations (Standard Industrial Classi-
fication Codes 2000 through 3999). Manufacturing industries
accounted for an even greater portion of the total quantity of
hazardous waste generated in 1981 (92%). The chemical industry
alone (SIC code 28) is estimated to have accounted for 68 percent
of the amount generated in 1981.
The survey found that treatment represented the most
prevalent hazardous waste management technique during 1981. it
is important to note, however, that treatment quantities appear
-------
to be most heavily affected by mixtures of hazardous and nonhaz-
ardous wastes (in particular, mixtures of hazardous wastes with
industrial process waters). Comparison of such quantities with
quantities of hazardous wastes landfilled, for example, may
therefore be misleading, since the quantities entering landfills
tend to contain substantially higher concentrations of hazardous
constituents. Furthermore, a greater degree of uncertainty
surrounds the estimates of treatment quantities, as explained in
detail in the sections presenting these estimates.
The survey estimates that approximately 14.7 billion
gallons of hazardous wastes were disposed of during 1981, with
underground injection observed as the most dominant disposal
mechanism, accounting for nearly 60 percent (approximately 8.6
billion gallons) of the total quantity disposed. Large portions
of this quantity were mixtures of hazardous and nonhazardous
wastes, however, presenting similar comparison problems to those
described above for the treatment quantities. Landfills were
found to outnumber all other disposal sites (199 landfill sites,
versus only 88 injection well sites), and accounted for 807
million gallons of the 14.7 billion gallons of hazardous waste
disposed. Nearly 500 million gallons of hazardous wastes are
estimated to have been incinerated during 1981, representing
just over half the quantity that was landfilled.
The survey estimates that there were 326 commercial
TSD facilities operating during 1981, accounting for 1.3 billion
gallons of hazardous waste management. While small in compari-
son to the total quantity generated, shipments by generators to
!This 0.5 billion gallons is not part of the 14.7 billion gallons
disposed since incineration is classified by RCRA as a treatment
process.
-------
commercial TSD facilities represented 82 percent of the total
quantity of hazardous waste shipped off site in 1981. The
survey's definition of a commercial facility is one that is
privately owned and operated, where more than 50 percent of the
hazardous waste managed during the year was received from firms
under different ownership. Expanding this definition to include
any facility that manages hazardous waste for a fee, the defini-
tion of "commercial" employed by OSW, the survey estimates that
509 such facilities were operational during 1981.
Finally, estimates from the survey indicate that the
total amount of annual unused capacity across the United States
was more than sufficient to accommodate the treatment, storage
and disposal requirements for quantities of hazardous waste
generated at 1981 levels. TSD facilities reported utilizing
only 36 percent of total disposal capacity during 1981, and
reported utilizing less than one-quarter (23%) of their hazard-
ous waste treatment capacity that year. Large, on-site manage-
ment facilities, however, dominated the national capacity picture
and may obscure potential capacity shortages among the more
numerous smaller facilities. Unused capacities were also distri-
buted unevenly across EPA regions, with Region IV accounting for
nearly 60 percent of unused disposal capacity and Region V
accounting for just under half of the nation's unused treatment
capacity. Furthermore, commercial facilities, which represent
"available" capacity, account for only small portions of the
total unused treatment and disposal capacities.
This report represents the completion of a three year
effort by OSW to improve its regulatory and general decisionmaking
information base. The survey data base will provide valuable
information to decisionmakers concerning hazardous waste manage-
ment practices regulated under RCRA over the coming years In
-------
addition, OSW is currently engaged in a number of studies to
improve its information base in specific issue areas, including
surveys of potential small quantity generators and handlers that
burn wastes in boilers and as fuels, and an ongoing series of
industry studies. OSW is also preparing to receive state by
state summaries of the 1983 Biennial Reports that were submitted
by hazardous waste generators and TSD facilities in the spring
of 1984. These reports and their summaries, will provide OSW
with a continuing flow of important, valuable data upon which
its future regulatory decisions may be based.
-------
1. INTRODUCTION
This report summarizes the findings of an extensive
national survey of hazardous waste generators and treatment,
storage and disposal facilities regulated under Subtitle C of the
Resource Conservation and Recovery Act of 1976, as amended. The
survey, conducted by mail during the fall of 1982 and the spring
of 1983, was sponsored and directed by the United States Environ-
mental Protection Agency's Office of Solid Waste (OSW). The
survey was administered by Westat Research, Incorporated, a
national survey research firm based in Rockville, Maryland.
Westat provided technical assistance in designing the survey,
implementing it, and analyzing its results under EPA contract
68-01-6621.
The survey was national in scope and involved the
development of 10 questionnaires containing more than 6,000
individual statistical data elements. The data obtained from the
11,714 respondents answering one or more of these questionnaires
are capable of providing statistical estimates of the magnitude,
scope, and nature of hazardous waste generation and management
activities regulated under Federal law during 1981. This report,
developed by Westat for submission to the Office of Solid Waste/
details the development of the survey, the procedures followed in
its conduct, and provides analyses and a summary of information
compiled from major portions of the obtained data.
The body of this report is organized into five major
parts composed of nine sections in all. Readers desiring a brief
overview of this report highlighting the major findings should
refer to the Executive Summary immediately preceding this section.
Part I (Sections 1-3) contains the introduction to the report and
-------
the methodology. Part II (Sections 4 and 5) presents the findings
on the numbers of hazardous waste generators and management
facilities described in various ways by the survey. Because the
statistical reliability of these data was generally very good due
to the survey design, they are presented separately from the
quantity data. Quantities generated and managed are presented
within their own less precise context in Part III (Sections 6 and
7). Part IV (Section 8) contains an analysis of national and
regional capacity for managing hazardous waste. Part V (Section
9) describes future EPA studies in the hazardous waste area. More
detailed information on the methodological approach, the survey
data base, and key regulatory requirements is found in the appendices
1.1 Background on the RCRA Hazardous Waste Regulatory
Program
With its enactment of the Resource Conservation and
Recovery Act of 1976 (RCRA) and in its subsequent amendments
thereto in 1978 and 1980, Congress required the Environmental
Protection Agency (EPA) to promulgate a regulatory program ensur-
ing adequate protections to human health and the environment in
the generation, transportation, and management of hazardous
wastes. RCRA was enacted by Congress to address the growing
national problem of waste generation and disposal, compounded by
rapidly developing industrial production and pollution abatement
technologies and the use of increasingly complex chemicals and
materials in the production of goods and services demanded by our
society.
Section 3010(a) of RCRA required all hazardous waste
handlers to notify EPA of their hazardous waste management acti-
vities by August 18, 1980 (90 days after EPA promulgated regula-
tions identifying and listing hazardous wastes). Among other
10
-------
PART I (Sections 1-3)
INTRODUCTION
AND
METHODOLOGY
-------
-------
things, Subtitle C of RCRA required EPA to:
• Promulgate regulations identifying the characteristics
of hazardous waste and listing particular wastes
to be regulated as hazardous wastes under RCRA
(§3001) ;
• Establish a manifest system for "cradle-to-grave11
tracking of hazardous waste shipments (§3002);
• Establish standards governing the generation (§3002)
and transportation (§3003) of hazardous wastes;
• Promulgate regulations to ensure proper treatment
storage, and disposal of hazardous wastes, including
the promulgation of standards governing the location,
design, construction and operating procedures of
?SqnT d°US "afte tf^ent, storage, and disposal
(TSD) facilities (§3004);
Ui?terim status" (temporary permits) to all
Q fq«n *e!Ltha!r,Tre "in existence" on November
19, 1980 and that (1) complied with the Notification
requirements of Section 3010(a), and (2) submitted
Part A Hazardous Waste Permit Applications to EPA
by November 19, 1980 (§3005(e));
Issue final permits to new and existing TSD facilities
as a mechanism for applying the facility standards
developed under §3004 to individual facilities
(§3005 ) ; and,
Promulgate guidelines to assist States in the
development of State hazardous waste programs, and
to grant interim and final authorization for
qualified State programs to administer the RCRA
hazardous waste regulatory program in lieu of the
Federal program, including the issuance and
enforcement of permits for the storage, treatment,
and disposal of hazardous wastes.
Between February 1980 and July 1982, OSW promulgated
the major components of the hazardous waste regulatory program
called for by Congress in its enactment of RCRA. Hazardous wastes
to be regulated under RCRA were identified by characteristic and
listed specifically; a "cradle- to-grave" manifest system was
11
-------
implemented; refinements to the definitions of regulated hazardous
waste generators and treatment, storage, and disposal facilities
were established, including the definitions of certain exempt or
partially exempt handlers such as "small quantity" generators,
ninety-day on-site "accumulators," and certain treatment facilities
already regulated under the National Pollutant Discharge Elimination
System (NPDES); and, finally, TSD facility permitting and operating
standards were issued in stages to govern most processes used to
treat, store, or dispose of hazardous wastes.
Based upon EPA's lists and definitions of hazardous
wastes and its definitions of regulated handlers, the Agency
received nearly 70,000 Notification forms in response to §3010(a)
of RCRA from firms indicating they were handling or might in the
future handle hazardous wastes regulated under RCRA. Of these,
nearly 60,000 indicated that they were or might in the future be
hazardous waste generators. The remaining Notification forms
indicated hazardous waste transportation, treatment, storage, or
disposal, and underground injection that were not located at the
same site as generators. Furthermore, nearly 15,000 of these
firms also submitted Part A permit applications to EPA, indicating
that they would be treating, storing, or disposing of hazardous
wastes regulated under RCRA.
1.2 The Need for a National Survey
Initial analyses of information submitted on Notification
forms and Part A applications provided EPA with its first picture
of the range of industrial and other activities, as well as of
the types of firms and organizations that would be regulated
under Subtitle C of RCRA. Notifications of hazardous waste
generation and applications for its subsequent treatment, storage,
and disposal were received from nearly every sector of American
12
-------
industry, as well as from an equally broad spectrum of firms and
organizations engaged in service-oriented (as opposed to production-
oriented) activities, including municipal, State, and Federal
government agencies. Further analyses of these initial sources
of data provided EPA with indications of the numbers and types of
processes used to treat, store, or dispose of hazardous wastes;
the general geographic distribution of hazardous waste generators
and TSD facilities; and rough indications of the quantities and
types of hazardous wastes that would be managed by these firms.
As EPA continued in its development of the various
aspects of the RCRA regulatory program, however, the need increased
substantially for more detailed information describing the
characteristics, scope, and magnitudes of the populations of
hazardous waste generators and TSD facilities regulated under
RCRA. In particular, Executive Order 12291 required EPA to conduct
extensive regulatory impact analyses (RIA's) of its existing and
pending regulations governing hazardous waste management practices.
The RIA's combine human health and environmental risk assessments
with benefit/cost analyses to assess the probable impacts of
existing or pending regulations. Executive Order 12291 requires
RIA's to be performed for all "major rules": regulations and
standards that are expected, among other things, to have significant
($100 million or more) impacts on the economy. Most of the regula-
tions that comprise EPA's RCRA hazardous waste program meet the
definition of major rules and are therefore required to be analyzed
extensively.
In addition to the requirements of Executive Order
12291, EPA required more information about the specific processes
used to manage hazardous wastes in order to complete its develop-
ment and evaluation of the facility standards required under RCRA
to ensure adequate protections for human health and the environment
in the treatment, storage, and disposal of hazardous waste.
13
-------
Furthermore, as EPA began managing the regulatory program it
created, its need for management information increased accordingly,
including the need to determine more precisely the nature and
numbers of generators and facilities actually regulated under
RCRA and the quantities of hazardous wastes annually generated
and managed in the United States. Finally, EPA received
increasing numbers of requests from Congress, industry groups,
research and development firms, environmental organizations, and
the public at large for reliable information about hazardous
waste management activities regulated under RCRA, information
that was largely unavailable from existing data sources.
Accordingly, OSW began in the summer of 1981 to develop
an extensive national statistical survey of RCRA-regulated hazardous
waste generators and treatment, storage, and disposal facilities.
OSWs purpose in conducting the survey was essentially three-fold:
• The primary purpose of the survey was to character-
ize the populations of hazardous waste generators
and TSD facilities regulated under Subtitle C of
RCRA that were actively generating regulated
quantities of hazardous waste or treating, storing,
or disposing of hazardous waste in processes
regulated under RCRA during 1981. The population
of hazardous waste handlers indicated by submissions
of Notification forms and Part A permit applications
included substantial numbers of firms that were
not actually managing hazardous waste at that
point in time, and many that were not actually
subject to or affected by the RCRA regulations.
Thus, the survey focussed on identifying the size
of the population of active handlers in a given
calendar year, 1981.
• The second purpose of the survey was to develop a
national data base on hazardous waste management
practices for use by OSW and others in the continuing
development and evaluation of the RCRA regulatory
program and in assessing its impact on the regulated
community, pursuant to Executive Order 12291.
This included gathering cost data pertaining to
hazardous waste operations.
14
-------
• The final purpose of the survey was to estimate
the magnitude and scope of hazardous waste
generation and its treatment, storage, and disposal
in the United States. Previous estimates of the
quantity of hazardous waste generated annually
had varied substantially and were often based
upon information from limited segments of the
RCRA-regulated population or upon secondary data
As a secondary objective, this survey was intended
to provide a baseline estimate of the quantities
of hazardous waste generated and treated, stored
and disposed of during 1981 in processes subject'
to the RCRA regulations using primary data
obtained from a statistically valid sample of
facilities regulated under RCRA.
Figure 1 on the accompanying page summarizes the above-
stated purposes for conducting the National Survey of Hazardous
Waste Generators and Treatment, Storage, and Disposal Facilities
Regulated Under RCRA in 1981.
1•3 Scope and Focus of the Survey
The survey was designed to be national in scope, covering
hazardous waste generation and management activities in all 50
States, as well as Puerto Rico, Guam, and the Virgin Islands.
Information obtained through the survey instruments focussed
almost exclusively on 1981 calendar year hazardous waste manage-
ment activities. Calendar year 1981 was selected as the study
year because it was the first complete calendar year following
the implementation of the RCRA regulatory program, and because
it was the most recent year for which complete calendar year
data were available from respondents at the time the survey was
actually conducted.
The focus of the survey was limited to those hazardous
waste handlers and those hazardous waste management activities
regulated under Subtitle C of RCRA during 1981. As a result
15
-------
Figure 1
PURPOSE OF STUDY
Characterize hazardous waste handlers regulated
by RCRA:
Generators
Treatment, storage, and disposal
facilities
• Develop data base to support:
Regulatory development
Regulatory impact analyses
Regulatory review
• Estimate 1981 hazardous waste quantities
Generated
Treated
Stored
Disposed
16
-------
the survey design excluded certain hazardous waste generators
and TSD facilities or processes that have been specifically
exempted from regulation under RCRAf including: small quantity
generators (installations that generate less than 1,000 kilograms
per month of hazardous waste or less than 1 kilogram per month
of acutely hazardous waste); processes used to treat hazardous
wastewaters under National Pollutant Discharge Elimination
System (NPDES) permits where the treatment occurs exclusively in
tanks; storage tanks and container storage areas used exclusively
for the on-site accumulation of hazardous wastes for up to
ninety days prior to their treatment, disposal, or shipment off
site; and publicly owned treatment works (POTW's) as defined by
§502(4) of the Clean Water Act. The survey also excluded generators
and TSD facilities where it was determined that the only hazardous
wastes they generated or managed during 1981 had been delisted
by EPA (even if the delisting occurred after 1981), and genera-
tors of specific waste streams that beneficially used or
recycled all of their wastes on site.
Furthermore, generators and TSD facilities or processes
that did not actively, generate or treat, store, or dispose of
hazardous wastes during the 1981 calendar year were also excluded
from the survey, even though they may well have been subject to
regulation under RCRA during that period of time. As an example,
a landfill into which hazardous wastes were placed during December
of 1980, but into which hazardous wastes were not input at any
time during 1981, would have been excluded from the survey, even
though the landfill was regulated under RCRA due to its use as a
hazardous waste disposal mechanism subsequent to the implementation
of the RCRA regulatory program.
The focus of the survey was designed with these
restrictions in order to produce a "snap shot" of RCRA-regulated
hazardous waste management activities actually occurring in a
17
-------
given calendar year. EPA believed that significant numbers of
firms that submitted Notification forms and Part A permit applica-
tions were not actually generating regulated quantities of
hazardous waste or treating, storing, or disposing of hazardous
wastes in processes regulated under RCRA (see the description of
Westat's telephone verification efforts in the discussion of
pre-survey screening efforts in Section 2.2). Thus, as stated
in Section 1.2, one of the primary purposes of the survey was to
estimate the numbers of regulated generators and TSD facilities
that were actively engaged in regulated hazardous waste manage-
ment activities during 1981.
Finally, the survey did not include any hazardous
waste generators or TSD facilities subject to the RCRA regulatory
program, but who had not complied with the RCRA §3010(a) Notifi-
cation requirements and who therefore handled hazardous wastes
during 1981 in violation of the law. Accordingly, it is beyond
the scope of this report to try to assess the magnitude of illegal
hazardous waste generation and management activities occurring
in 1981 or in any other year.
The various factors affecting the scope and focus of
the survey are summarized in Figure 2 on the accompanying page
and are discussed in greater detail in the substantive sections
of this report and in the survey field report contained in
Appendix B.
18
-------
Figure 2
SURVEY SCOPE
• National Survey of Hazardous Waste Handlers
• 1981 hazardous waste activities only
• Only those who registered with EPA (Notifiers
and Part A permit)
• Only RCRA covered hazardous waste management,
therefore design excludes:
Small generators
90-day accumulators
Wastewater treatment in tanks
Publicly owned treatment works
Delisted wastes
Other exempted handlers and waste
streams
19
-------
-------
2. SURVEY METHODOLOGY
This section of the report provides an overview of the
design and development of the survey, together with a brief
description of the steps and procedures followed in actually
conducting the survey. The information presented in this section
and the next section on statistical reliability, is intended to
provide readers with a basic understanding of the structure of
the survey, how it was conducted, and how the findings presented
in the remainder of the report should be interpreted. A more
detailed account of the survey methodology is presented in the
field report in Appendix B for those requiring more in-depth
information.
2 -1 Questionnaire Design
Beginning in the summer of 1981, OSW, in conjunction
with Westat, conducted a series of information needs assessments
to determine the specific nature of the data to be collected
through the national survey. Information needs were found to
differ considerably across the various aspects of OSWs development
and management of the RCRA hazardous waste program. Three general
categories of information needs did, however, emerge:
(1) Information describing various characteristics of
the populations of hazardous waste generators and
TSD facilities regulated under RCRA;
(2) Information relating to the quantities of RCRA-
regulated hazardous wastes generated and treated,
stored, and disposed; and,
(3)
Information relating to the costs of hazardous
waste management activities.
21
-------
Since the desired characteristics, cost, and quantity
information differed considerably across the various processes
used to treat, store, and dispose of hazardous waste, separate
survey instruments (mail questionnaires) were developed for each
process. In all, eight process-specific management technology
questionnaires were developed for TSD facilities, namely:
• The Underground Injection Well Questionnaire;
• The Landfill Questionnaire;
• The Land Treatment Questionnaire;
• The Surface Impoundment Questionnaire (a single
questionnaire was developed for treatment, stor-
age, and disposal surface impoundments);
• The Waste Pile Questionnaire;
• The Incinerator Questionnaire;
• The Container Questionnaire; and
• The Tank Questionnaire (a single questionnaire was
developed for storage and treatment tanks).
In addition, a Treatment, Storage, and Disposal General
Questionnaire was developed for all TSD facilities. In effect,
the TSD General Questionnaire is a compilation of the questions
that are common to all TSD facilities, regardless of the types of
management technologies they employ. Had OSW conducted a survey
of landfills only, for example, the questionnaire that would have
been developed would have included all of the questions in the
Landfill Questionnaire and all of the questions in the TSD
General Questionnaire. Similarly, if the survey had focussed
exclusively on incineration, the questionnaire that would have
been developed would have included all the questions in the
Incinerator Questionnaire and all of the questions in the TSD
General Questionnaire. Since many of the TSD facilities to be
included in the survey were expected to provide information about
22
-------
more than one process, the TSD General Questionnaire was devel-
oped to eliminate the unnecessary duplication of questions in
each of the process-specific management technology questionnaires.
A Generator Questionnaire was also developed to provide
information relating to hazardous waste generation practices, as
opposed to hazardous waste management (treatment, storage, and
disposal) practices occurring after generation. All of the
questionnaires were designed to obtain data relating to each of
the three information categories defined above: characteristics
data (e.g., whether or not the landfill was lined, whether or not
the facility was located in a floodplain, the type of material
used in constructing the surface impoundment, etc.); quantity
data (e.g., quantity of hazardous waste generated, quantity of
hazardous waste disposed of in landfill, etc.); and cost data
(e.g., process-specific construction costs, facility labor rates,
etc.). Cost questions were generally not included in the
Generator Questionnaire, however, with the exception of general
questions on the costs incurred by generators in sending their
wastes off site for treatment, storage, or disposal.
In all, 6,202 statistical data elements were established
in the 10 questionnaires. All of the data elements were designed
to be fully automated, with closed-end answer category codes
established for almost all questions, and precise question-
specific instructions provided in an attempt to obtain clearly
defined data from the respondents. The survey was designed to
provide a functional data base for continuing analyses by various
parties in the hazardous waste field, including EPA, its contractors,
Congressional committees, and other interested parties. Many of
the major data elements contained in the survey data base are
described and summarized in the substantive sections of this
report. A more extensive description of the survey data is
provided in Appendix D.
23
-------
2.2 Pre-Survey Screening: Development of the Sample Frame
OSW had originally intended to use the files of Part A
permit applications and Notification forms maintained in EPA's
Hazardous Waste Data Management System (HWDMS) as the frame for
drawing statistical samples of generators and TSD facilities
operating specific management technology processes during 1981.
As indicated previously, however, OSW recognized that the
Notification forms and, particularly, the Part A permit
applications were designed to pertain to then current (1980) and
future hazardous waste generators and TSD facilities. Since the
objective was to survey only those sites that were active hazardous
waste handlers during 1981, OSW was concerned that sampling from
the Part A and Notification files would produce inefficient
samples (i.e., the samples drawn from those files would include
large numbers of firms that were intended to be excluded from the
survey, thereby resulting in large numbers of "ineligible"
respondents, smaller obtained data files for analysis, and
reduced reliability in the statistical conclusions reached).
In response to these concerns, Westat conducted a small
telephone survey of firms that filed Part A applications, and an
even smaller telephone survey of firms that submitted Notification
forms to EPA. The purpose of these surveys was to test the
suitability of these two data files to serve as valid sample
frames from which to select samples of sites that were actively
engaged in regulated activities in 1981. The results of these
surveys forced OSW to conclude that, indeed, the Part A file
would not serve as a useful sample frame for the survey of TSD
facilities actively engaged in regulated activities in 1981, due
primarily to protective filings and erroneous filings by
facilities not then subject to the RCRA regulations.
24
-------
Westat was then instructed to construct a new sample
frame by conducting telephone interviews with all TSD facilities
that submitted Part A applications to determine the actual
processes, if any, used in RCRA-regulated management of hazardous
wastes during 1981. Westat was successful in completing inter-
views with approximately 85 percent of the nearly 9,500 facili-
ties listed in the HWDMS Part A file during July and August of
1982. The results of this survey are summarized in Westat's
"Report on the Telephone Verification Survey of Hazardous Waste
Treatment, Storage, and Disposal Facilities Regulated Under RCRA
During 1981." A hybrid file of telephone verified facilities and
remaining uncontacted facilities on the Part A file was then
established as the sample frame for the TSD facility sample.
This hybrid sample frame is referred to in the remainder of this
report as the "Part A/Telephone Verification" file.
The initial telephone interviews also suggested that an
inefficient sample of the 1981 generators of regulated quantities
of hazardous waste would result from the file of firms that
submitted Notification forms indicating they were or might in the
future be hazardous waste generators. In fact, the telephone
interviews with firms that indicated generation on their Notifi-
cation forms but that did not submit Part A applications identi-
fied an average of only one actual 1981 generator of regulated
quantities of hazardous waste out of every 20 interviews.
Therefore, in order to successfully use the Notification file
as the sample frame for 1981 generators, oversampling was employed
to assure that a sufficiently large number of 1981 generators
would, indeed, be observed. The alternative was to conduct
nearly 60,000 telephone verification interviews prior to conducting
the mail survey, which would have been prohibitively expensive
and time consuming.
25
-------
2.3 Sample Design
As indicated in the introduction, the survey was
intended to be national in scope. Accordingly, the samples of
generators and TSD facilities were designed to produce nationally
reliable statistical results. Essentially, the samples fall into
two categories:
• A sample of TSD facilities, comprising eight
subsamples of facilities corresponding to the
various process type questionnaires, drawn from
the Part A/Telephone Verification file; and
• A sample of generators, drawn partially from the
Part A/Telephone Verification file (to observe
generators that managed their waste on site) but
primarily from the file of firms that submitted
Notification forms (indicating generation) but did
not file Part A applications (to observe genera-
tors that ship their hazardous wastes off site for
treatment, storage, or disposal).
A total of 10,667 generators was selected from the
Notification file to receive the Generator Questionnaire, and an
additional 553 Generator Questionnaires were mailed to firms
selected from the Part A/Telephone Verification file. The total
number of Generator Questionnaires mailed out, therefore, equaled
11,220. The generator samples were each stratified by industry
type. Each frame was divided into three groups defined by
combinations of Standard Industrial Classification (SIC) codes
provided by OSW. (See Appendix A for a detailed description of
^Note that the reason such a large number of generators was
sampled from the Notification file was to compensate for the
anticipated high proportion of nonregulated firms that would be
included.
26
-------
the sample stratifications.) Furthermore, the samples were drawn
so as to provide representation of generators in each EPA region
in numbers proportional to the number of firms found in each
region on the sample frame.
The sampling scheme developed and employed for TSD
facilities, however, was considerably more complex. Separate
independent random samples of facilities were drawn from the
Part A/Telephone Verification file for each of the eight process-
specific management technology questionnaires. Actually, 10
independent sampling lists were created from the Part A/Telephone
Verification file: lists of facilities indicating that the
process of interest (e.g., incineration) was used to manage
hazardous waste during 1981 were created for seven of the eight
process-specific management technology questionnaires; two such
lists were created for the Tank Questionnaire (one list for
storage in tanks, a second list for treatment in tanks); and,
finally, a tenth list was created from which to select TSD
facilities from the Part A/Telephone Verification file that also
indicated that they generated hazardous waste. Since the tele-
phone survey obtained size indicators for the processes confirmed
to be used at facilities in 1981, each list was ranked by the
appropriate size measure, and then random samples were drawn from
each with a systematic selection procedure. It should be noted
that the size-measure rankings were employed to assure propor-
tional representation of hazardous waste handlers of all sizes,
not as a mechanism to oversample large or small handlers.
Facilities that operated more than one type of waste
management process in 1981 were included on as many lists as they
had processes, and were thus eligible to be selected in more than
one sample. In fact, many facilities were selected to receive at
27
-------
least two process-specific management questionnaires and a large
number received even more (the survey design limited the maximum
number of process-specific management technology questionnaires
sent to a single site to four, plus a General Questionnaire).
Finally, any facility selected from the Part A/Telephone Verifi-
cation file to receive one or more process-specific questionnaires
or the Generator Questionnaire also was required to complete the
TSD General Questionnaire. As described in Section 2.1, the TSD
General Questionnaire was designed to combine the questions common
to all the process questionnaires, and thereby reduce the workload
for respondents by eliminating the need to fill out duplicate
answers in two or more process-specific questionnaires.
The resulting sample of TSD General Questionnaires was
therefore selected with varying probability. The weighting
technique used to account for the varying selection probabilities
of sampled sites is described in Appendix A. A total of 2,557
TSD facilities were eventually selected through this multiple
list selection process. An additional 42 facilities were
subsequently added to the sample to compensate for an initial
low response rate in one category, bringing the total number of
facilities sampled, and therefore the total number of TSD General
Questionnaires mailed out to 2,599. Section 3 provides specific
sample size information for each of the process-specific management
technology questionnaires, together with a presentation of the
response rates and an assessment of the statistical reliability
of the data obtained from all of the questionnaires.
The last sample design issue that requires discussion
at this point in the report concerns the selection technique
employed within each sample list, after the lists had been
stratified by SIC code (generators) and process type (TSD facilities),
organized by EPA region (generators and facilities), and ranked
by size (TSD facility process lists). Once each list was organized,
28
-------
the actual sampling approach used was that of "equal probability
of selection." As indicated in Section 1.2, OSW had a number of
purposes in conducting the national survey, and desired to develop
at least three distinct types of data: (1) characteristics
data, such as estimates of actual numbers of regulated and active
generators and TSD facilities, the proportion of landfills with
liners, the proportion of incinerators employing rotary kiln
technology, etc.; (2) cost data, such as estimates of the cost
of installing liners in surface impoundments, the cost of sending
hazardous waste to commercial facilities, the cost per ton to
incinerate hazardous waste as opposed to disposing of it in a
landfill, etc., and (3) quantity data, such as estimates of the
total quantity of hazardous waste generated annually, the quantity
of hazardous waste disposed of in landfills or injection wells,
the quantity of hazardous waste recycled, etc.
The precision of statistical estimates in each of
these three categories is sensitive to the sampling procedure
employed in obtaining the data to be used in generating the
estimate. Estimates of population characteristics are best
served through the sampling technique employed in this survey:
equal probability of selection (i.e., each facility on the
sampling list has an equal likelihood of being selected to be
included in the sample). Estimates in the quantity area and, to
a lesser extent, the cost area are often better served through
the use of selection probabilities that are proportionate to a
relevant size measure (PPS sampling). with PPS selection, the
larger facilities on the sample frame, or list, have a greater
likelihood of being selected in the sample.
An equal probability sampling approach was employed in
the national survey for two specific reasons. First, at the
time the survey was being designed and developed, OSW placed
greater priority on the ability to describe the characteristics
29
-------
of the RCRA-regulated population than it did on the ability to
estimate the quantity of hazardous waste generated annually.
OSWs primary motivation behind this allocation of priorities
was that the regulatory impact analyses required under Executive
Order 12291 are intended to assess the impact of regulations
that require changes in the practices or the structures of
facilities affected by them; that is, changes in the characteristics
of the regulated population. Thus, greater emphasis had to be
placed on obtaining reliable estimates of the characteristics of
the regulated population, particularly since support for the
RIA's was one of the primary purposes of the survey.
The second factor supporting the use of equal probabil-
ity sampling in this survey was one of practicality: the alter-
native, probability of selection proportionate to size (PPS),
requires that the sample frame contain reasonable size indicators
for all, or most, of its members. OSW did not have access to
any size indicators for the generators selected from the
Notification file, since paperwork reduction requirements prevented
a quantity question from being included on the Notification
forms. Thus, PPS sampling was not possible for generators.
Some size indicators, however, were available for the specific
processes employed at TSD facilities, such as the quantity of
hazardous waste incinerated, or the maximum size of the waste
pile at any point during the year. These size indicators, however,
were obtained for fewer than half of the facilities during the
telephone survey. Furthermore, the size indicators obtained
varied from process to process, and therefore did not provide a
consistent size indicator for TSD facilities as a whole. Thus,
the size data available for TSD facilities was considered
inadequate for the use of PPS sampling selection.
The impact of equal probability sampling on the
statistical reliability of the survey data will be discussed in
Section 3.
30
-------
2-4 Data Collection
The data for this study were collected using mail
survey techniques, during a field period that lasted from fall
1982 to spring 1983. A set of 10 questionnaires was developed
for data collection as described in Section 2.3.
The questionnaires were designed to be self-explanatory.
Each question of each questionnaire contained its own instructions
to the respondent. In addition, respondents were provided with
an instruction booklet and three appendices of supplemental
information. Also, technical assistance in the form of "hotline-
telephone services was provided to respondents both by Westat
and by the EPA's RCRA hotline. Approximately 4,000 calls were
handled by the two hotline services combined. Finally, all
quantity questions included in the survey questionnaires provided
respondents with the ability to answer in units of measure of
their choosing, thereby reducing the potential for respondent
error in performing conversion into consistent units of measure.
Westat then converted all quantity answers into consistent units
by computer recedes.
The questionnaire packages (including the questionnaires,
a cover letter, instructions, appendices, and a return envelope)
were mailed to sampled facilities and installations during the
first two weeks of September 1982. The respondents were
instructed to complete and return their questionnaires to the
Office of Solid Waste, Environmental Protection Agency, within
45 days of receipt of the questionnaire package. The cover
letter indicated that time extensions beyond the 45 days' response
time could be requested, and that respondents could request that
information submitted in the questionnaires be treated as
confidential business information.
31
-------
Response rates for the questionnaires were high, which
is to be expected in a data collection effort where response is
mandatory. The response rate for the Generator Questionnaire
was 88.0 percent, and the response rate for the Treatment, Storage
and Disposal General Questionnaire was 90.5 percent. Response
rates for the process-specific questionnaires ranged from a low
of 81.1 percent for the Land Treatment Questionnaire to a high
of 92.0 percent for the Container Questionnaire. Lists of
nonrespondents were sent to the EPA regional offices for review
and followup actions.
Further details regarding data collection techniques,
return status frequencies, response rates and nonresponses
may be found in Appendix B: Field Report.
2.5 Quality Control
Quality control efforts were focussed on two goals.
The first goal was to obtain the highest possible level of
statistical quality by achieving a high response rate. The
second goal was to minimize nonsampling error to the greatest
extent possible.
2.5.1 Response Rate Quality Control
In order to encourage a high response rate, the question-
naires were designed to be self-administered and easy to answer.
In addition, the questionnaire mailout package included a general
instruction booklet and other supplementary information. Technical
assistance, in the form of telephone hotline information services,
was offered to all respondents both by Westat and by the EPA
RCRA hotline. Efforts to make the data collection instruments
-------
easy for respondents to answer were also intended to make it
easy for respondents to answer accurately and consistently, thus
minimizing nonsampling error as well.
Questionnaire packages were sent by certified mail, in
order to provide proof of delivery. A small percentage of the
certified mail cards were not returned, even though the question-
naire package itself had not been returned as undeliverable.
The Post Office was requested to trace those missing certified
packages and report back on the delivery status. Telephone
directory searches and some phone calls were used to research
new addresses for facilities/installations with inadequate
addresses and for questionnaire packages that had been returned
undelivered. Most of the facilities receiving Treatment,
Storage and Disposal General Questionnaires and process specific
technology questionnaires had been verified or updated through a
telephone verification survey that preceded the mailout survey.2
A number of approaches were used to encourage the
recipients of the questionnaire packages to complete and return
them. The cover letter to the questionnaire package informed
the recipients of their legal obligation to respond to the
questionnaires. Nonrespondents who had not returned their
questionnaires within the 45-day period permitted, received
followup letters informing them of the lateness of their response
Generator installations that had not responded within 80 days of
receiving the questionnaire package were sent a second followup
letter. TSD facilities not responding after 80 days were followed
up by telephone.
The telephone verification survey is described in detail in the
Report on the Telephone Verification Survey of Hazardous Waste
Ine?Q«?n»' STS^ and DisP°sal Facilities Regulated Under RCRA
in 1981," produced by Westat in November 1982
33
-------
2.5.2 Nonsampling Error Quality Control
Efforts to minimize nonsampling error focussed on
respondent error (which included inconsistent as well as missing
responses), and processing error (which included coding error,
data entry error, and programming error). Editing for these
types of error occurred at virtually all stages of survey proces-
sing. Coders were trained to edit for missing and inconsistent
responses while coding. Coders' work was 100 percent sight
verified ("proofed") by trained coder verifiers, who were also
assigned to check for respondent errors. After data entry, all
data files were "machine edited" using computer programs that
checked the valid range of each data item and performed logical
consistency checks among data items. Coding supervisors reviewed
the frequency distributions for all data items, looking for
errors that the computer machine edit program had not found.
Analysts reviewed complex logical comparisons within each data
file.
Some respondent errors involving incomplete responses
and inconsistent responses could not be resolved without recon-
tacting the respondent. The process of recontact is called
"data retrieval." For this study, data retrieval generally
required telephone recontact of the respondent. About 30
percent of the Treatment, Storage and Disposal General Question-
naire respondents were recontacted during data retrieval, and
about 10 percent of the Generator Questionnaire respondents were
recontacted. In addition, a rigorous effort was made by survey
analysts to recontact the TSD facilities that managed the
largest quantities of hazardous waste by each of the management
technologies. A separate report on this editing effort is
presented in Section 2.5.3, following.
concept of nonsampling error is further expanded upon in
Section 3.1.
34
-------
2'5'3 Editing of Large Quantity Cases in the Component
Data Files
In addition to the editing procedures described in
Section 2.5.2, the largest quantities in each of the processes
reported in the component questionnaires received special edit
consideration. First, facilities were ranked according to their
weighted quantities for each component process. Professional
staff then checked questionnaires of the largest facilities for
internal consistency. Component information reported in the TSD
General Questionnaire provided the primary consistency check,
but component questionnaires of other processes were also used
for facilities belonging to the sample frame of more than one
process .
Particular attention was given to known problem areas
such as misreporting exempted wastewater streams, misreporting
measurement units, or defining "hazardous" incorrectly.
The number of top facilities examined varied with each
process type, but generally included:4
• At least the top five facilities in each sample;
5° Per°ent °f the total weighted quantity;
• All sampled facilities representing at least five
percent of the total weighted quantity.
Fortunately, a large part of the quantity managed by
each hazardous waste treatment, storage or disposal process was
4
Waste pile quantities could not be examined because the comDonenf
questionnaire did not contain a general quantity question Ontv
38 percent of the (weighted) incinerated quantity was edited
35
-------
represented by a small number of facilities in the sample. This
is most dramatically illustrated in disposal surface impound-
ments where the largest facility in the sample accounted for 93
percent of all surface impoundment disposal quantities. The
process showing the least skewed distribution of quantities was
the injection wells' disposal process. Even so, more than
50 percent of the total quantity was accounted for by only 8
of the 73 respondent injection well facilities. Table 1 presents
the exact quantity coverage of this editing procedure for each
component.
2.6 Imputation Procedures Used for Quantity Estimates
In order to present quantity estimates representing
the entire population of hazardous waste facilities regulated by
RCRA, quantity data presented in this report include imputed
values. If a facility reported "don't know" instead of a quantity,
the mean value of applicable cases was imputed for that facility.
This is also true for facilities where the quantity was "not
ascertained," or where at least one element determining the
quantity was unknown.
The quantity totals are always equal to the known
number of facilities times the mean quantity of applicable
facilities with known values. This insures that the entire
population is represented by the quantity estimate.
36
-------
Table 1. Special Edit Coverage of the Largest Facilities in Each
Component Data File
Process
Treatment
Tanks
Surface impoundments
Incinerators
Storage
Tanks
Containers
Surface impoundments
Waste piles
Disposal
Injection wells
Landfills
Surface impoundments
Land applications
Percentage of Total Weighted Quantity
Represented
Rankings
of Edited
Facilities
1-5
1-8
1 R
3-5
1-6
1-7
*
1-8
1-5
1-5
1-9
Largest Unedited
Facility
< 5%
< A 9.
<•" 1 o
^ 2%
< 2%**
< 4%**
< 4%
*
< 4%
< 3%
< 1%
< 5%
All Edited
Facilities
72%
66%
38%
76%**
52%**
77%
*
51%
68%
90%
72%
°f
entered
ed^d The ""mS "edite<3" and "unedited" in this table refer to the
edit described in Section 2.5.3. All questionnaires returned
-chine-assis d ng
37
-------
-------
3.1
3. STATISTICAL RELIABILITY AND DATA ACCURACY
The Concepts_o^SamŁljLng Error and Nonsampling Error
A variety of estimates are presented in this report
providing measures for characteristics of interest. These
include estimates of the percentage (or number) of sites with a
particular characteristic and total quantities of hazardous waste
in various contexts. It is important to keep in mind that these
are survey estimates and am subject to errors of various kinds.
Such errors can be classified into two general categories:
sampling error and nonsamplang error.
A measurement of sampling error is an assessment of the
precision of estimates obtained from a sample, as opposed to a
census, of a population. A census - where all sites would be
selected to be respondents -- has no sampling error associated
with it. Differences between measures obtained from a census and
the "true" values for the population cannot be attributed to
having taken only a sample or subset of the population as
respondents. An estimate from a sample will usually differ from
the "true" population measure, and the extent to which this is
attributable to the variability inherent in selecting a sample
can be measured. Confidence intervals arid standard errors
(standard deviations of an estimate) are measures of sampling
variability or sampling error. if the sampling error is
relatively small, the sample estimate is likely to be close to
the "true" population measure, assuming that the effect of
nonsampling error on the estimates is minimal.
Nonsampling error refers to all other sources of error
that might occur in a survey. These include mistakes in entering
39
-------
values on a questionnaire, misinterpretation of questions, •
undetected data entry errors, transcription errors, nonresponse,
etc. A census, as well as a sample, is susceptible to nonsam-
pling errors. In general, nonsampling errors cannot be measured
from the data collected in a survey. Nevertheless, for this
survey, efforts have been made to assess the possible impact of
such errors. In cases where there is a possibility that non-
sampling errors may have an important effect on the estimates,
this is noted in the report.
3.2 Response Rates and Sample Sizes
A major reason to maximize the number of respondents
providing data in a survey is that one can never be sure that
nonrespondents have the same characteristics as respondents. To
the extent that nonrespondents are different from respondents,
survey estimates, based solely on respondents, are to some degree
biased. Response rates obtained for this survey were exception-
ally high for a mail survey, due in part to the mandatory response
authorities under which the survey was conducted. Table 2
presents the number of questionnaires mailed out, the number
received, and the number of eligible respondents for each question-
naire type. It should be noted that the number of TSD General
Questionnaires mailed out cannot be obtained by summing across
the number mailed out for the individual components. This is
true for two reasons. First, a TSD facility may have received
questionnaires for more than one process type. Second, generators
associated with on-site TSD facilities on the sampling frame also
received TSD General Questionnaires. Also, note that the sum of
the number of eligible storage tanks and treatment tanks (233 +
121 = 354) exceeds the number of eligible respondents to the Tank
Questionnaire (283) because some respondents used both process
40
-------
Table 2. Number of Eligible Responses and Response Rates to the RCRA Survey
Responses
Questionnaire
Typ_e
Questionnaires
Mailed Out
' i
Response
#Eligible Rate
Total Received on Data File (Col. 3-Col 2)
3 4 5—
Generators
From Notifier
File
From Part A/
Telephone
Verification
File
TSD's (General
Questionnaire)
[TSD Components]
Injection Well
Landfill
Land Treatment
Surface Impoundment
Waste Pile
Incinerator
Storage Container
Tank
Storage
Treatment
11,220
10,667
553
2,599
9,877
9,361
516
2,348
2,084
1,710
374
1,462
88%
88%
93%
90%
115
202
122
327
243
265
423
847
103
172
99
298
215
239
389
772
73
79
37
145
73
125*
191
283
233
121
90%
85%
81%
91%
88%
90%
92%
91%
*117 of the eligible incinerator respondents represent facilities with active
inc.nerators in 1981. The refining eight eligible incinerator res^nden s
had hazardous waste incinerators that were temporarily shut down cruder
construction during 1981 but were defined as eligible because the incinerator
'"6-"5^6519116" t0 °btain addi"°nal infection conceding
n these categories.
41
-------
types. "Ineligible" respondents are those respondents that did
not generate regulated quantities (generators that produce less
than 1,000 kilograms of hazardous waste or less than 1 kilogram
of acutely hazardous waste during 1981 are examples of "ineligible"
generators) or that did not actively manage hazardous wastes in
processes regulated under RCRA during 1981 (exempted wastewater
treatment tanks under NPDES or RCRA-regulated landfills into
which hazardous wastes were not disposed during 1981 are examples
of "ineligible" facilities or processes).
Column 5 of Table 2 presents the response rates obtained
for each questionnaire component. Please note that compliance
with the survey was mandatory under authority delegated to EPA in
Section 3007 of RCRA. Response rate is defined as the percentage
of responding facilities among all facilities sent a particular
questionnaire (Column 3 divided by Column 2 in Table 2). As
shown in Column 5, the response rates fell in the vicinity of 90
percent for most process types. With response rates this high,
there is generally little ne.ed for concern about nonresponse bias
unless an extremely large handler fails to respond, but the extent
of such occurrences among the nonrespondents to this survey is
unknown. It is known, however, that certain item nonresponses
(failure by a respondent to answer a particular question) involved
very large facilities and may therefore subject certain items to
nonresponse bias.
3.3 Statistical Reliability and Sampling Error
The statistical precision achieved in the survey depended
on the type of questionnaire (individual process type, TSD general,
or generator) and the type of estimate (characteristic percentages
or quantities). The sample and universe sizes varied by question-
naire type. For the Tank, TSD General, and Generator Questionnaires,
42
-------
eligible sites on the sample frame were assigned varying proba-
bilities of selection depending on what process types were
indicated as being on site according to the information on the
sample frame and, for generators, depending upon their Standard
Industrial Classification code (see Appendix A, Section A.I, for
more details on SIC code stratification). Each of these situations
contributed to varying levels of precision for the questionnaire
types.
More importantly, the sample was designed to be more
efficient for estimates of facility characteristics (e.g., the
percentage of landfills that have liners) rather than quantities
(e.g., the quantity of hazardous waste generated). As explained
in Section 2.3, this type of sample design (equal probability of
selection for all components except tanks, generators, and the
TSD General Questionnaire) was chosen partly due to the importance
of determining facility characteristics and partly due to the
absence of reliable facility size information at the time the
sample was drawn. Without adequate size information, it is not
possible to make an effective probability proportional to size
(PPS) sample selection.1 (Section 2.3 discusses the issues of
the sample design in more depth.)
As a result, quantity estimates (or ratios of quantity
estimates) are subject to considerably more estimation error than
are estimates about the number or percentage of sites with
certain characteristics. A high degree of precision was achieved
for these estimates of site characteristics. For example, the
precision achieved at the 95 percent confidence level for
By "facility size" we mean a measure of the quantity of ha7ardnn«s
waste managed or generated at the facility. hazardous
43
-------
estimated percentages from the TSD General Questionnaire was 3.0
percent. This indicates that one can be "95 percent confident"
that the true percentage is within +3.0 percent of the estimate.
As the quantities of hazardous waste generated or
managed vary dramatically among facilities, the precision of the
survey's quantity estimates is much lower. To illustrate the
large variation that occurs in facility size, consider Figure 3.
The percentage of the TSD facility population managing at least a
given quantity of hazardous waste is presented for several different
percentages. For example, an estimated one percent of the TSD
facility population each managed at least 306 million gallons of
hazardous waste, while an estimated 5 percent each managed at
least 24 million gallons. Clearly, the size distribution of
facilities is skewed, due to the existence of some very large
hazardous waste management facilities. The resulting mean value
is 195 times the median. Figure 4 is an alternative way of
looking at the population distribution of hazardous waste TSD
facilities. From Figure 4 we see that if all hazardous waste
facilities were ranked from the largest to the smallest, then a
mere one percent of the facilities would account for more than 70
percent of the hazardous waste managed. In fact, sites representing
the largest 240 (5%) TSD facilities were associated with 67.7
billion gallons of managed hazardous waste in 1981, accounting
for 95 percent of the 71.3 billion gallons estimated to have been
managed by all 4,818 TSD facilities in 1981. To confirm the
quantities reported by the largest facilities, EPA made independent
verification telephone calls to the largest cases and also deter-
mined through in-depth discussions that the wastes and processes
reported were indeed RCRA-regulated.
44
-------
Figure 3
DISTRIBUTION OF MANAGEMENT FACILITY SIZE
(quantity managed)
300 -
o
UJ
Z
Ł 200 —
i~ ._^
^^ "Tr*
r^
O "o
O ,/,
< .2
—
—
< ^
u_
O
> 100 -
h-
z
Q
50 -
25 -
10 -
0 —
306
I
i
I
i
I
!
t
i-
AVERAGE QUANTITY MANAGED:
MEAN: 14.8 Million gallons
MEDIAN: .076 Million gallons
1 H)
24
i ! r-4-, -2 .02
5%
0%
25%
50%
% OF MANAGEMENT FACILITIES AT OR ABOVE INDICATED QUANTITIES
45
-------
Figure 4
COMPARISON OF CUMULATIVE DISTRIBUTIONS OF QUANTITY OF HAZARDOUS WASTE
MANAGED IN 1981 AND THE NUMBER OF MANAGEMENT FACILITIES
O <
< z
h- <
u ^
25
100
90 -
70 -
50
1<
*<
u Ł
20 -
10
10
20
30
40
50
~T
60
~T
70
80
~T
90
100
CUMULATIVE PERCENTAGE OF HAZARDOUS WASTE MANAGEMENT FACILITIES IN THE U.S.
(FROM LARGEST TO SMALLEST FACILITY)
46
-------
Similarly skewed distributions were found for hazardous
waste generators according to responses to the Generator Question-
naire that were also verified by telephone for large facilities.
Furthermore, since most hazardous waste is generated by facilities
that manage their waste on site, the skewed distribution of
generators could also be observed by examining the quantities
generated at management facilities, according to responses to the
TSD General Questionnaire. In fact, this method of estimating
generation proved to be statistically more reliable than estimates
of the same population from the Generator Questionnaire. As a
result, generation data from the TSD General Questionnaire were
combined with data on generators without management facilities
from the Generator Survey to form the generation estimate that is
used in this report.2 Despite this improvement, the skewed
distribution still led to an imprecise estimate.
Table 3 provides a comparison of the low degree of
precision achieved for three quantity estimates: 1) the survey's
preliminary estimate of the total quantity of hazardous waste
generated as derived from the Generator Questionnaire; 2) the
total quantity of waste that was managed as hazardous waste as
derived from the TSD General Questionnaire, and 3) the survey's
final estimate of the total quantity of hazardous waste generated
which, as is explained in greater detail below and in Section
6.1, is derived through a combination of data obtained through
the Generator and TSD General Questionnaires. The quantity
estimates contrasted are then with the high degree of precision
achieved for estimated percentages of generator installations and
TSD facilities based on all respondents to the questionnaire.
2The generation estimate and methodology used to derive it are
detai Ipn in .QOO+-•; ^>-, & i -• «~ j. i_ ao.c
detailed in Section 6.1
47
-------
Table 3. Statistical Reliability of Estimates from the Generator
and TSD General Questionnaire
Questionnaire
Generator
TSD General
95% Confidence
Interval on
an Estimated
Percentage
+ 2.4%
+ 3.0%
Combination of Generator
and TSD General Used Only
to Derive 1981 Generation
Estimate
95% Confidence
Interval on
an Estimated
Quantity of
Hazardous Waste
Generated
(or Managed)
± 79.7% of the total
( + 48.8% of the total)
Not applicable
+ 49.7% of the total
To interpret the plus or minus factor indicated in
the table for an estimated quantity of hazardous waste generated
(or managed), the true value of the total quantity of hazardous
waste generated (or managed) is covered with 95 percent confidence
by an interval centered at the estimated value and extending
on either side of the estimated value. For example, if the
TSD general estimated quantity were 100 million gallons, the
corresponding 95 percent confidence interval would extend from
51.2 million gallons to 148.8 million gallons. Clearly, the
confidence interval is quite wide, and the estimated quantity
of total hazardous waste managed is not very precise. A similar
interpretation holds for the +49.7 and +79.7 percent factors
for the two estimates of the quantity of total hazardous waste
generated, which are even less precise.
To interpret the plus or minus factor indicated in
the table for an "estimated percentage of facilities" estimate
for generators, the true value of the percentage of installations
48
-------
with a particular characteristic is covered with 95 percent
confidence by an interval centered at the estimated percent
and extended on either side of the estimate by the percentage
shown below. For example, a conservative 95 percent confidence
interval for an estimate of 40 percent of the TSD facilities
having a particular characteristic (e.g., they receive waste
from off site), would extend from 37 percent to 43 percent of
all TSD facilities (40% +3.0%). It should be noted that these
"plus or minus" terms are established using an estimate of 50
percent (a proportion of .5).3 Such intervals provide at
least 95 percent confidence. A conservative approach was
selected since these limits may be applied as a tool to assess
the general precision of estimated percentages. If more specific
limits for particular items are desired, these need to be
individually computed.
Sampling errors for quantity estimates other than
those presented in Table 3 generally indicate less variability
than the +79.7 percent confidence interval for the Generator
Questionnaire estimate of hazardous waste generated. The
Generator Questionnaire estimate of hazardous waste generated
and shipped off site, for example, has a +30 percent confidence
interval at the 95 percent level. The estimated quantity
generated but not shipped off site, on the other hand, has a
confidence interval that exceeds +79.7 percent because of the
great variability of quantities generated by the generators
sampled from this population. This led to concern about the
accuracy of Generator Questionnaire estimates of the quantity
generated on site by this part of the generator population.
Tn^ifHH f? interval for other proportions would be smaller if
individually computed. A quick approximation can be obtained by
mu PnYMn? he conservative interval shown in Table 3 by 2 V(p)(l-p
where "p is the proportion. Examples are provided later in P
Section 4.1 and 5.1 footnotes.
49
-------
Fortunately, a second, less variable data source, the TSD
General Questionnaire, could be used to estimate the quantity
generated for these generators that also managed waste on
site. This led to the previously mentioned combined generation
estimate based on responses to the Generator Questionnaire
from generators without management facilities, and responses
to the TSD General Questionnaire from generators with management
facilities, yielding an overall interval for the 1981 generation
estimate of 4-49.7 percent at the 95 percent confidence level.
Because this was considerably lower than the +79.7 percent
confidence level from the Generation Survey estimate, and for
other reasons cited in Section 6.1, the combined survey estimate
served as the source of the final 1981 generation estimate
presented in this report.
For individual components or process types, the
accuracy on percentages of facilities with particular character-
istics ranged from +5 to +11 percent, as shown in Table 4, but
was better than +10 percent for all components except Land
Treatment. These confidence' intervals were based on the number
of facilities responding to each of the process type (component)
questionnaires, and the size of the universe. They are also
based on the conservative assumption that the estimated proportion
of the characteristic of interest (e.g., line their landfills),
is .50 (i.e., 50 percent). As previously footnoted, the confidence
intervals for other proportions can be determined by multiplying
the conservative interval shown in Table 4 by 2 V(p)(1-p),
where "p" is the proportion expressed as a decimal.
50
-------
Table 4. Number of Respondents and Statistical Reliability of
Estimated Percentages, by Process Type
95 Percent
Confidence
Interval on an
Number of Estimated
Respondents Percentage of
Process Type Active in 1981 Facilities
Injection Wells 73 +48%
Landfills 79 +87%
Land Treatment 37 + ^l 3%
Surface Impoundment 145 + 75%
Waste Piles 73 + 39%
Incinerators 117* +66%
Storage Containers 191 + 72%
Storage Tanks 233 + 6.4%
Treatment Tanks 121 + 3.6%
*The sample size for incinerators is 125, when one includes eight
eligible respondents to the questionnaire component who were not
active in 1981. (See note to Table 2).
The 95 percent confidence intervals presented in
Tables 3 and 4 are based on questions that required an answer
from all respondents to the questionnaire. If it is desired to
approximate 95 percent confidence intervals on the estimated
percentage of sites with a particular characteristic for subsets
of the respondents (e.g., landfills with liners), the "plus" or
"minus" term shown for the 95 percent confidence interval should
be multiplied by the factor, K, shown below in Table 5.
51
-------
Table 5. "K" Factors to Be Used in Obtaining 95 Percent
Confidence Intervals for Population Subsets
as a Factor, K, to multiply
:§_29Łu!ation terrain Tables 3 or 4
80 1.1
60 1.3
40 1.6
20 2 2
10 3.1
The "K" values presented above are reasonable approximations as
long as the subsample size is fairly large, say 30 or more, and
the estimated percentage is not extremely large or extremely
small. 4 Aiso, the use Qf ,,K factors,. for approximating conf:dence
intervals around quantity estimates is only appropriate if the
subpopulation has approximately the same distribution as the
population as a whole. If regional distributions of waste
quantities were similar, for example, the "K factor" could be
used to approximate the 95 percent confidence intervals around
the regional quantity estimates. Waste group quantities, however,
are likely to have very different distributions among waste
groups, and the "K factors" would not be appropriate for approximating
confidence intervals.
An example of how to obtain an estimate of variability
(i.e., the confidence interval) for a subgroup of a population
applying the "K factor" to an estimate of the whole population
(based on all respondents) is as follows: Table 3 shows that
percentage estimates for generators are accurate to ± 2.4 percent
52
-------
at the 95 percent confidence level. However, the subpopulation
of "generators" that managed all of their waste on site" is only
16 percent of all generators (based on 338 respondents, so the
concern about small sample sizes is satisfied and use of the K
factor is appropriate). Therefore using the formula K =VN/R
(see footnote 2), K = VT7 ffi -OR mi...,, .u
// ^ vi/.ib - 2.5. Thus, the accuracy for this
subpopulation of generator is:5
(2.5) (+ 2.4%) = + 6.0%.
3 • 4 Nonsampling Error
As indicated earlier, nonsampling errors are those
which result from sources other than that attributable to sampling
There were various potential sources of nonsampling errors in
these surveys. Although such errors are generally not quanti-
fiable, it is important to acknowledge these sources so that
users of the survey data may be aware of their possible effects.
Potential sources of nonsampling errors include:
nonresponse bias (discussed in Section 3.2); the misinterpretation
of questions; mistaken responses to questions; inadequate defini-
tions of terms or inappropriate assumptions inherent in the
questions; rounding errors; and errors in converting from wet to
dry measures (e.g., from gallons to metric tons) because the
general !y assumed relationship between wet and dry measures does
not hold in a particular case (e.g., the specific gravity of the
liquid differs considerably from the assumed specific gravity).
nr ^ n°ted that for items from the Generator, Tank and
TSD Gener
, n
neral Questionnaires, the K values in Table 5, or hose
based on the above formula, represent approximations. This is
due to the fact that sites selected for the generator tank and
TSD general samples were selected with variable Sampling rates"*
resulting in nonuniform weights. iates,
53
-------
All of these errors probably occurred to a limited extent. There
is no evidence to suggest that they introduced an inordinate
amount of bias/error into the survey results, with the possible
exception of nonsampling errors related to the proper inclusion/
exclusion of hazardous wastewaters and other waste in mixtures.
It is possible that up to 5 billion gallons of hazardous waste
in wastewaters that were treated exclusively in RCRA-exempt
NPDES tanks could not be eliminated by the TSD General Survey's
editing procedures, and as such may be erroneously included in
the TSD General Survey's quantities treated and managed. There
was also evidence that a substantial quantity of hazardous
wastewater mixture went unreported as only the dry weight of
hazardous materials was reported. The potential impacts of such
possible errors are discussed in more detail in the appropriate
quantity sections of this report.
54
-------
PART II (Sections 4-5)
NUMBERS OF GENERATORS
AND
MANAGEMENT FACILITIES
-------
-------
INTRODUCTION TO PART II:
NUMBERS OF HAZARDOUS WASTE
GENERATORS AND MANAGEMENT FACILITIES
This part is made up of two major sections. Section 4
and its subsections describe the number of hazardous waste
generators and various subpopulations of these generators.
Section 5 and its subsections describe the number of management
facilities that treat, store, or dispose of hazardous waste.
As previously stated, the statistical reliability of
data from this survey related to the number of facilities is
generally very good. It is especially good for generators,
owing mainly to the large sample (2,084) in the Generator
Survey. For generators as a whole, the 95 percent confidence
interval on a proportion of installations is + 2.4 percent.
For management facilities it is + 3.0 percent. Estimates for
subpopulations have wider confidence intervals, becoming less
precise as the subpopulation gets smaller. For a more detailed
explanation of the issue of precision, please refer to Sections
3.3 and 3.4.
55
-------
-------
4. NUMBER OF HAZARDOUS WASTE GENERATORS
Sections 4.1 through 4.5 define hazardous waste generators
and describe their population in terms of the number of facilities
in each EPA region, the industries to which they belong, the
hazardous waste groups they generate, whether or not they ship
some or all of their generated hazardous waste off site, and how
many recycle part of their hazardous waste.
A generator is defined in 40 CFR §260.10 as "any person,
by site, whose act or process produces hazardous waste identified
or listed in Part 261 of this chapter or whose act first causes
a hazardous waste to become subject to regulation." Among
wastes specifically excluded in Part 261 are:
• Wastes that pass through a sewer system to a
publicly owned treatment plant;
• NPDES permitted point source discharges of indus-
trial wastewater regulated under Section 402 of
the Clean Water Act, as amended;
• Certain trivalent chromium wastes, including
chrome trimmings and shavings, buffing dust
sewer screenings and wastewater treatment sludges
generated by various subcategories of the leather
tanning and finishing industry, that fail the EP
toxicity test for chromium, but do not fail any
other EP toxicity tests and do not fail the tests
for any other characteristics;
• Irrigation return flows;
• Solid agricultural wastes returned to the soils
as fertilizer (such as crop residuals or manure);
• Household wastes or materials derived from house-
hold wastes (e.g., refuse-derived fuel);
• Source, special nuclear, or by-product material
as defined by the Atomic Energy Act of 1954 a^
amended, 42 U.S.C. 2011 et seq.;
57
-------
• In situ mining wastes that are not removed from
the ground as part of the extraction process;
• Solid waste from extraction, beneficiation and
processing of ores and minerals, including coal;
• Mining overburden returned to the mine site;
• Drilling fluids, produced waters, and other wastes
associated with the exploration, development, or
production of crude oil, natural gas or geothermal
energy;
• Fly ash waste, bottom ash waste, slag waste, and
flue gas emission control waste generated primarily
from fossil fuel combustion;
» Wastes produced by small quantity generators
(generators generating less than 1,000 kilograms
in a month) where no more than 1 kilogram per
month of acutely hazardous waste is produced; and
• Certain recycled wastes (that is, wastes that are
being beneficially used or reused, or legitimately
recycled or reclaimed).
As of October 13, 1983, there were 51,345 establishments
on EPA's Notifier file indicating current, or possible future,
generation of nonexempt hazardous waste. In mid-1981, the year
of the survey data, nearly 60,000 such notifiers were listed in
EPA's files. The reduction between 1981 and 1983 occurred as a
result of EPA's effort to remove "protective notifiers" and
other nonregulated notifiers from its files. Greater elaboration
of this issue is provided in subsequent pages of this section.
The survey sample was drawn from an August 1, 1982
list of 55,739 notifiers. Based upon responses to the survey,
EPA now estimates that about one-quarter of these establishments
58
-------
actually generated RCRA-regulated quantities of hazardous waste
during 1981. EPA estimates that there were 14,098 active
generators of hazardous waste regulated under RCRA in 1981 (see
Figure 5), with the remaining 41,641 of the 55,739 notifiers in
that year falling into the various categories presented in Table 6
Table 6. Characteristics of Nonregulated Notifiers that
Responded to the Survey
Description
Non-generators
Potential Future Generators
Small Quantity Generators
100% Recyclers
Recently Retired Generators
Delisted Waste Generators
Nonregulated Notifiers
NSK, NEC*
Percent
43%
18%
18%
4%
3%
1%
13%
100%
*NSK = not specified by kind
NEC = not elsewhere classified
Nonregulated_Notif^ers - Nearly half of the nonregulated
notifiers sampled from the notifier and Part A files did not
generate hazardous waste during 1981, the five years preceding
1981, nor did they expect to generate hazardous waste in the
five years following 1981. This overfiling may have been done
for protective reasons to ensure compliance with RCRA regulations
for contingency planning, or because respondents misinterpreted
the regulations. Eighteen percent of the nonregulated notifiers
sampled had not generated hazardous waste during 1981, but expected
to within the next five years.
59
-------
Figure 5
PORTION OF NOTIFIERS THAT GENERATED RCRA-REGULATED
QUANTITIES OF HAZARDOUS WASTE IN 1981
41,641 Nonregulated
Notifiers (75%)
55,739 Notifiers
Source: HWDMS, 8-1-82
60
-------
Generators of hazardous waste in 1981 that were exempt
from RCRA as "small generators" because they generated no more
than 1,000 kilograms of hazardous waste, and not more than one
kilogram of acutely hazardous waste, during any month also
accounted for 18 percent of the nonregulated notifiers sampled.
Generators that recycled all their hazardous waste, generators
that retired within the last five years, and those that generated
delisted waste accounted for four percent, three percent, and
one percent respectively. The remaining nonregulated notifiers
(13%) included a small number of farm-exempt generators, but
for the most part, reported they were not RCRA-regulated generators
without further specifying their status.
In the sections that follow, we discuss the distribution
of number of generators by region of the country, by industry
group, and other breakdowns. it is important to note that very
different distributions may be found for the quantity of waste
generated, which is presented later in Section 6.
4<1 RgJLJQJial Distribution of Generators
While the actual number of generators actively generating
RCRA-regulated quantities of hazardous waste during 1981 is now
estimated to be significantly lower than initially thought,
the distribution of generators across EPA's ten regions remains
similar to the distribution of Notification submissions. The
regional distributions of two different notifier lists are
compared, in terms of percentages, to the regional distribution
estimated by the survey in Table 7. The 1981 survey sample
frame was drawn from the August 1, 1982 list of notifiers. The
most current EPA list of notifiers at the time this report was
being written was as of October 13, 1983. Looking at the
regions as three groups, those with the largest share of notifiers
61
-------
are estimated to have the largest share of active generators.
The estimated number of hazardous waste generators in each
region is presented in Figure 6.
Table 7.
Comparison of Notifier and Generator Distributions
in EPA Regions
Rank
1
2
3
4
5
6
7
8
9
10
Region
V
IX
IV
II
I
III
VI
VII
X
VIII
Total U.S.
Survey Estimated
Percent of Generators
1981
23.0%
14.1%
13.0%
11.8%
10.2%
9.9%
5.1%
2.4%
1.7%
100.0%
Percent of Notifiers
Identified on HWDMS
File as of:
8/1/82 10/13/83
22.4%
21.6%
10.2%*
13.1%
12.4%
8.0%
9.7%
14.3%*
4.6%
3.1%
2.2%
11.3%*
12.8%
14.4%*
9.4%
10.4%
12.0%*
3.3%
3.3%
1.5%
100.0% 100.0%
These numbers are outside the conservatively stated range of the
95 percent confidence interval surrounding the survey estimates
tor those regions (i.e., the percent of generators +24%)
62
-------
Figure 6
REGIONAL DISTRIBUTION OF HAZARDOUS WASTE
GENERATORS IN 1981
1433
ESTIMATED TOTAL NUMBER
OF GENERATORS ACTIVE IN 1981
14,098
63
-------
Region V had 3,240 generators (or 23% of the U.S. total),
decisively more than any other EPA region. Regions I-IV, VI and
IX had between 1,200 and 2,000 generators apiece with Region
VIII having the least (243).
The precision, at the 95 percent confidence level, of
the survey estimates for the regional percentage of the national
number of generators is conservatively stated as + 2.4 percent
for each regional estimate. Thus, for Region II, the 95 percent
confidence interval around the estimate 11.8 percent is from 9.4
percent to 14.2 percent (11.8 + 2.4).1 However, estimates about
the characteristics of facilities in a particular region would
be less precise (i.e., would have larger confidence intervals),
due to the smaller sample size for any given region. For example,
95 percent confidence limits on estimates of characteristics for
facilities found within Region II would be approximately +7.0
percent.
The^value of + 2.4 percent is a conservative estimate of
oh^! i Y.aSSring 5° Percent of the Population has a particular
characteristic (e.g., are in a particular region). The actual
value grows somewhat smaller as the proportion diverges from 50
percent. To adjust for different proportions and obtain more
precise confidence intervals, the conservative estimate (+? 4%
for numbers of generators) can be multipled bv 2^(p (l-p)~
where »p» is the proportion (expressed as a fraction of 1 ) of
facilities with a certain characteristic. For example, if 11 8
n »°hgene^a?rQ^Were ^ Regi°n IIr then s^stituting
p, the actual 95 percent confidence interval fnr that
regional estimate would be (+2.4%) (2) \l(. 118(1- us) or + 1 5
percent. Therefore, we can be 95 percent confident that the '
percentage of generators in Region II is 11.8 percent + 1 5
?Mrrn^°^ ^ °ther W°rdS' between 10.3 percent and 1373 percent
(Note that when p = .5, the confidence interval is + 2.4%??
Calculated with the "K factor" (see Table 5) as follows:
2.4) yi/p - (2.4) yi/.118 = +7.0
64
-------
4.2
Number of Generators by Industry Type
Hazardous waste generators were most prevalent in the
manufacturing industries (SIC 20-39). Only 15 percent were
nonmanufacturing generators, or unclassified. Industries
manufacturing or processing metals, electrical equipment, or
chemicals had the most generators as shown in Figure 7 below.
Figure 7. Number of Generators by Industry Type3
MANUFACTURING:
Fabricated Metal Products
Chemicals and Allied Products
Electrical Equipment
Other Metal-related Products
All Other Manufacturing
NONMANUFACTURING AND NSK:
SIC 34 ** 2,636
SIC 2# « 2,443
SIC 36 *
* 1,51$
SIC 33,35,3? * 2,222
SIC 20-27,29-32,38-39 « 3,208
2074
TOTAL GENERATORS = 14,093
Sampling error estimates were +25 percent or less
°
outside
65
-------
At the two-digit level, Major Group 34 of the Standard
Industrial Classification Codes, SIC 34, Fabricated Metal Products,
except Machinery and Transportation Equipment, had the most
generators, about 2,600, or 19 percent of the total. The
Chemicals and Allied Products industry (SIC 28) was ranked
second with about 2,400, or 17 percent of the generators. Next,
Electrical and Electronic Machinery, Equipment and Supplies (SIC
36) had about 1,500, or 11 percent of the generators. Metal-
related manufacturing industries, not mentioned above, had about
2,200, or 16 percent of the generators. These included Primary
Metal Industries (SIC 33), Machinery, except Electrical (SIC
35), and Transportation Equipment (SIC 37). Other manufactuiing
industries accounted for about 3,200, or 22 percent of the
generators. As previously mentioned, the remaining 2,100, or 15
percent of the generators, were from nonmanufacturing industries,
or industries not specified by kind (NSK).
4.3 Number of Generators by Waste Group Generated
The survey results provided estimates of the number of
generators generating specific types of hazardous wastes during
1981. Generators were asked to indicate the portions of their
total hazardous waste streams that were associated with major
groups of EPA hazardous waste codes. In addition, generators
were asked to report the EPA waste code for each hazardous waste
that was generated, although waste code-specific quantities were
not obtained. Figure 8 presents the estimates of the number of
generators in major EPA hazardous waste groups, developed from a
combination of the data provided by respondents to these two
questions. Note that a given generator may have generated more
than one type of waste, and that a given waste stream may have
been reported under more than one waste code (multiple characteristic
waste streams or mixtures). As a result, the sum of the generators
66
-------
Figure 8
NUMBER OF ESTABLISHMENTS GENERATING
EACH MAJOR WASTE GROUP
14,088 Tot*! Hazardous Waste Generators
6,117 D001 Igmtable Wastes
4,705 D002 Corrosive Wastes
999 D003 Reactive Wastes*
3,923 D004~~DGI?
E.P. Toxic Wastes
7,180 FOOT~F005
Spent Halogenated and Nonhalogenated
Solvents
2,309
F006 F019
Electroplating and Coating Wastewater Treatment
SudKcS and Cyanide-bearing Bath Solutions and
Sludges
1,439
K001 K106 Listed Industry Wastes from Specific Sources
P001 PI 23 Acutely Ha/ardous Wastes
4,062 U001-U247
orfDiscĄded Commercial Chemical
and Manufacturing Intermediates
State Regulated Hazardous Wastes
,720
Self defined Hazardous Waste,
0% 10%
25'
50% 75o/0
PERCENT OF ALL HAZARDOUS WASTE GENERATORS
Confidence Interval exceeds ± 25% at the 95% Confidence Level
100%
-------
across all of these waste groups substantially exceeds the total
of 14,098 in the population.
Just over half (7,180) of the total population of
14,098 generators indicated that they generated spent solvents,
both halogenated and nonhalogenated (EPA waste codes F001-F005),
during 1981. Generators of sludges from wastewater treatment
systems associated with electroplating and aluminum coating
operations and generators of cyanide-bearing quenching and
plating bath solutions and their sludges (F006-F019) accounted
for 16 percent (2,309) of the generator population, while only 10
percent (1,439) of the generators generated listed hazardous
wastes from specific industrial sources, such as spent pickle
liquor from steel finishing operations, pink/red water from TNT
operations, or dissolved air flotation float from the petroleum
refining industry ("K" prefix hazardous wastes).
Forty-three percent (6,117) of the generators generated
ignitable wastes (D001), a third (4,705) generated corrosive
wastes (D002), more than a quarter (3,923) generated wastes that
failed EPA's "extraction procedure" test for toxicity (D004-
D017), and 999 generated reactive wastes (D003) during 1981.
Each of these categories includes wastes that, while not specifically
listed in EPA's list of hazardous wastes, exhibit hazardous
characteristics (e.g., low flash points, high or low pH levels,
volitility or violent reactive tendencies with other substances
such as water, toxicity, etc.) and are thus regulated under RCRA
as hazardous wastes.
Just under 30 percent (4,062) of the generators reported
generating hazardous wastes that were spilled, discarded, or off-
specification commercial chemical products or manufacturing
chemical intermediates ("U" prefix waste codes), and slightly
more than 10 percent (1,418) indicated generating the subset of
68
-------
such products or intermediates that are regulated under RCRA as
acutely hazardous wastes ("P" prefix waste codes, subject to a
small generator exclusion level of only 1 kilogram per month).
In addition to the above EPA classications, more than
2,000 establishments generated wastes identified as hazardous
wastes by states, but not by EPA, and another 1,700 handled
wastes as hazardous wastes, even though they were not listed or
identified as hazardous wastes by EPA or their state.
4.4
Number of Generators Shipping Hazardous Waste Off Si
Although nearly all hazardous waste is managed to some
degree at the site where it is generated (see Section 6.5), Figure
9 illustrates that only one out of every six generators (just
over 2,300) manage their hazardous waste exclusively on site. Of
those 11,800 generators that ship hazardous waste to off-site
management facilities for treatment, storage (for more than 90
days), and disposal, roughly 3,100 manage part of their hazardous
waste on site. Approximately 8,700 generators ship all of their
hazardous waste off site. Some of the latter group reported
shipping more hazardous waste off site than they generated.
Shipments can exceed generation when inventories of accumulated
waste are reduced from one year to the next, or when a facility
becomes a transfer point for waste received from off site that is
reshipping.
Because the inventory changes and transfer shipments
mentioned above are unknown, and because respondents to the
survey may not have always rounded the quantity generated and the
quantity shipped consistently, the survey's specific estimates in
the "some off site" and "all off site" categories shown in
Figure 9 are probably less reliable than other characteristic
69
-------
Figure 9
NUMBER OF GENERATORS SHIPPING HAZARDOUS WASTE
OFF SITE IN 1981
NUMBER OF GENERATORS
ALL ON SITE
2,320
SOME OFF SITE
SOME ON SITE
o
3,074
ALL OFF SITE
16%
22%
62%
8,704
TOTAL = 14,098 Generators
70
-------
estimates provided in this report. Nonetheless, it is clear from
the survey results that most generators ship all of their hazardous
waste off site for treatment, storage, and disposal, and that
only a small portion of the population manages its hazardous
wastes exclusively on site.
4.5
Number of Generators Recycling Hazardous Wast-
Many generators recycle some or all of their
hazardous waste. Generators of certain hazardous wastes that
recycle 100 percent of their hazardous waste are not subject to
RCRA regulations and are therefore not included in this analysis
(see 40 CFR 261.6). Those recyclers subject to RCRA regulations
(i.e., non-small generators recycling less than 100% of their
hazardous waste) have been increasing over time as illustrated in
Figure 10. Over half of all generators are expected to recycle
hazardous waste after 1981.
Recycled wastes include those that are used or reused
such as for raw materials in production processes; or reclaimed
such as solvent redistillation, scrap metal reclaimed by second-
ary smelter, or wastes that are blended to make fuels. Of the
14,098 generator population, about 5,700 (41%) are estimated to
have generated hazardous waste that was recycled (either on site
or off site) before 1981. The number increased to nearly 6,100
generators recycling during 1981 and further increased to more
than 7,800 generators that expected to recycle after 1981 This
latter number represents 56 percent of the 1981 generators
compared to the 43 percent that reported recycling during 1981
71
-------
Figure 10
NUMBER OF GENERATORS RECYCLING
HAZARDOUS WASTES
Percent of All
Generators
100-
90-
80-
70-
50-
40
~v/
30-
20-
10-
0
5,743
6,072
7,865
PRIOR TO 1981
DURING 1981
AFTER 1981
Total Generators: 14,098
72
-------
The quantity information presented later in Section 6 6
indicates, however, that only a small portion of the total quantity
of hazardous waste is actually beneficially used, reused, recycled
or reclaimed. The data suggest, therefore, that while a large
number of generators recycle at least some hazardous waste, only
small portions of their waste streams are actually recycled. It
is not known, however, what portion of these waste streams are
recyclable. Many factors reduce or eliminate the possibility of
recycling, as discussed in greater detail in Section 6.6.
73
-------
-------
5.
DEFINITION AND NUMBER OF MANAGEMENT FACILITIES
Sections 5.1 through 5.5 define hazardous waste manage-
ment (TSD) facilities and describe their population in terms of
the number of facilities in each EPA region, the number of
commercial and other types of facilities, number of facilities
that treat, store, and dispose of hazardous wastes, and the
number of facilities operating specific treatment, storage, and
disposal processes.
Hazardous waste management facilities are those involved
in the "systematic control of the collection, source separation,
storage, transportation, processing, treatment, recovery, and
disposal of hazardous waste," as defined in 40 CFR §260.10.
Existing TSD facilities were required under RCRA to notify EPA
of their hazardous waste management activities by filing Part A
permit applications on or before November 19, 1980.
Some types of hazardous waste management are exempt
from RCRA regulation. Most notably these include some forms of
wastewater treatment and the storage of hazardous waste for less
than 90 days.
The hazardous waste management facilities of interest
in the surveys discussed in this report are those that actually
treated, stored, or disposed of hazardous waste in processes
that were regulated under RCRA during 1981. The survey results
show an estimated 4,818 facilities meeting these conditions,
with a precision of + 280 facilities at the 95 percent confidence
level. The estimate of 4,818 is substantially less than the
roughly 8,500 facilities for which EPA currently has Part A
applications on file in HWDMS. The difference between the two
counts merits some discussion.
75
-------
As explained in Section 2.2, prior to undertaking
these surveys, the Part A file in the HWDMS data base was
closely examined in order to determine whether it would provide
a reasonable sample frame from which useful samples of 1981
active TSD facilities could be selected. The following items
came to light:
• While Part A permit applications were required to
be fUed only by facilities "in existence" on or
before November 19, 1980, the actual wording of
the application forms themselves was somewhat
?denT?US;H The aPPllcations asked respondents to
identify the processes that "will be used" to
manage hazardous wastes. Thus, many applications
were submitted for facilities that werfnot actu-
ally operational at the time of submission during
nrp A ^Plications as a
precaution, even though they may knowingly not
have been actually managing hazardous waste at
that time or during 1981. For example, many
facilities operating nonhazardous waste manage-
ment processes (such as trash burning incinera-
tors or solid waste landfills) filed Part A
applications to cover those processes in the
?nh I-* hazardous wastes were ever introduced
into them either accidentally or out of neces-
sity (spill clean-ups, etc.). other facilities
th^t Fp*rt AaPPlicatl°ns against the possibility
that EPA would expand its hazardous waste listings
to cover additional wastes. ^tings
Finally, many facilities filed Part A permit
applications mistakenly, either due to misunder-
standings about the nature and scope of the RCRA
abou "th0'5 V?6 t0 ln-ff-ientPinformation
a the constituents in their waste Breams or
76
-------
For reasons such as those mentioned above, the HWDMS
file contains sites in addition to those actually active during
1981. Thus, it does not directly provide the total number of
facilities actually engaged in hazardous waste management at a
given point in time. The estimate obtained from the current
survey was established from a question which specifically asked
respondents selected via a probability sample, whether hazardous
wastes were actually treated, stored, or disposed of in processes
subject to regulation under RCRA in 1981. Thus, the survey
provides a direct estimate of the actual number of active
management facilities at that time.
As indicated in Section 1.3, however, the survey
estimate of the number of active TSD facilities in 1981 is not
intended to represent the number of facilities subject to regu-
lation under RCRA at that time or currently. The survey design
specifically excluded a variety of regulated facilities, includ-
ing: regulated sites that were temporarily inactive during
1981;! sites that no longer manage hazardous wastes but which,
due to prior hazardous waste management and lack of formal
closure, are still subject to regulation under RCRA; and sites
that may not manage hazardous wastes, or that manage hazardous
wastes exclusively in exempt processes, whose "interim status"
has not been formally withdrawn under RCRA.
Therefore, the survey systematically understates the
magnitude of the population of TSD facilities subject to regu-
lation under RCRA during 1981. The number of facilities subject
to regulation under RCRA at any point in time is represented by
the number of "valid" Part A permit applications on file at EPA
As an exception to this, eight eligible incinerator respondents
had hazardous waste incinerators that were temporarily shut down
or under construction during 1981, but were included in the
addlti°nal information about incinerators
77
-------
at that time. The number of these facilities has declined since
1980, as facilities that have been determined not to be subject
to the RCRA regulations have been formally withdrawn from the
regulated community. Until such official actions are taken,
however, facilities with valid Part A applications are regulated
under RCRA, whether or not they actually processed or handled
hazardous waste in 1981.
5-1 Regional Distribution of Hazardous Waste Management
Facilities
As indicated in the previous section, survey results
estimate that 4,818 facilities treated, stored, or disposed of
hazardous waste in processes regulated under RCRA during 1981.
As discussed earlier, however, this estimate should not be
interpreted to represent the number of TSD facilities actually
subject to regulation under RCRA in 1981; rather, the 4,818
facilities represent the subset of RCRA-regulated facilities
that actually processed or handled hazardous wastes in regulated
processes in 1981.
The map presented in Figure 11 depicts the distribu-
tion across EPA's ten regions of the number of TSD facilities
estimated to have managed hazardous waste in RCRA-regulated
processes during 1981. The number of management facilities is
concentrated in five of the ten regions. The largest two regions
(Regions V and IV) account for 40 percent of the estimated total
number of TSD facilities, while 75 percent of the facilities are
in the top five regions. Table 8 indicates the percentage
distribution of these 1981 active hazardous waste TSD facilities,
and compares this distribution with the percentage distribution
of TSD facilities identified on the HWDMS Part A file in late
1983 and in mid-1982, at the time the survey samples were drawn
from that file.
78
-------
Figure 11
REGIONAL DISTRIBUTION OF HAZARDOUS WASTE
MANAGEMENT FACILITIES IN 1981
354
ESTIMATED TOTAL NUMBER
OF ACTIVE TSD'S IN 1981: 4,818
79
-------
Rank
Region
1
2
3
4
5
6
7
8
9
10
->
V
IV
VI
II
III
IX
I
VII
X
VIII
Survey
Estimated
Percent of
TSD Facilities
1981
26%
14%
13%
13%
11%
7%
7%
4%
3%
2%
Percentage of
Part A Applications
Identified on
HWDMS File as of
8-1-83
22%
17%
13%
14%
10%
9%
fts-
o ^
4%
1%
2%
10-13-83
27%
13%
12%
11%
11%
9%
O o
9%
4%
2%
2%
From Table 8 it can be seen that there is a very close
correspondence between the survey and HWDMS regional distribu-
tions of TSD facilities. Thus, although the actual numbers of
TSD facilities reported on HWDMS are generally overestimates of
the number of active, regulated sites, the relative regional
concentration of such facilities appears to be accurately
reflected by the HWDMS data. The ranking of regions based on
the survey-estimated percentage of 1981 facilities is similar to
that obtained from a ranking based on the facilities reported on
HWDMS from Part A applications in 1982 and 1983. There is, at
most, a four percentage point difference between the percentages
for 1981 and either of the other two years for any region.
The precision at the 95 percent confidence level of
the survey estimates for the regional percentage of the national
number of TSD facilities is conservatively estimated at + 3.0
percent for each region. Thus, for Region II, the 95 percent
confidence interval around the estimate 13 percent is from
80
-------
10.0 percent to 16.0 percent (13 + 3.0).2 However, estimates
about the characteristics of facilities in a particular region
would be less precise (i.e., would have larger confidence
intervals), due to the smaller sample size for any given region.
For example, 95 percent confidence limits on estimates of
characteristics for facilities found within Region V would be
approximately +5.9 percent.3
<-\
The value +3.0 percent is a conservative estimate of variabilitv
assuming 50 percent of the population has a particular charac-
teristic (e.g., are in a particular region). The actual value
grows somewhat smaller as the proportion diverges from 50 percent
To adjust for different proportions and obtain more precise
confidence intervals, the conservative estimate (+3.0% for
numbers of TSD facilities) can be multiplied by 2V(p)(l-P)
where "p" is the proportion (expressed as a fraction of 1) of
facilities with a certain characteristic. For example, if 13
percent of TSD facilities were in Region II, then substituting
. 1.3 _ for p, the actual 95 percent confidence interval for that
regional estimate would be (+3.0)(2)V(.13) (1 -—IT) or + 2 0
percent. Therefore, we can be 95 percent confident that the
percentage of generators in Region II is 13 percent +2.0 percent
or in other words, between 11 percent and 15 percent. (Note
that when p = .5, the confidence interval is + 3.0%).
3 Calculated with the "K" factor (see Table 5) as follows:
(3.0) Vl/p = (3.0) Vl/,26 = 5.9
81
-------
5.2
Number of Managemen^_Faci_lities by Industry Type
Facilities engaged in the management of hazardous
waste are heavily concentrated in manufacturing industries (primary
SIC codes 2000-3999). An estimated 4,018 facilities (83.3% of
the total estimate of 4,818) are located at establishments classi-
fied as manufacturers. Figure 12 shows the distribution of the
number of management facilities across the largest industry
categories for which statistically reliable data is available.
Figure 12. Number of Management Facilities by Industry Type
MANUFACTURING:
Chemicals and Allied
Products
SIC 28 m 1,249
SIC 34 « 547
SIC 36 m 540
SIC
33,
r 3$,
r 3 / *S
804
SIC
20-
-27,
29-32,
38-39 » $78
800
Fabricated Metal Products
Electrical Equipment
Other Metal-related
Products
All Other Manufacturing
NONMANUFACTURING AND NSK:
TOTAL TSD FACILITIES = 4,818
The chemical industry has, by far, the'largest number
with 1,249 facilities managing hazardous waste, representing a
little over one-quarter of all such facilities. The fabricated
metal industry and the electrical machinery industry each account
for slightly more than 11 percent of the total number of facili-
ties. The remaining facilities are widely distributed over
various other industries.
82
-------
5-3 Number of Commercial Management Facilities
For the purposes of this study, a commercial facility
has been defined as a privately owned and operated facility that
receives more than half of its hazardous waste from firms with
which it is not associated by ownership (i.e., that it neither
owns nor is owned by). This definition was established in order
to provide comparability in the survey's estimates of commercial
activity with previous analyses of such operations. The specific
nature of the definition was determined through an analysis of
the distribution of percentages of interfirm shipments received
by facilities in the sample that indicated SIC 4953 (refuse
systems) as their primary SIC code. This differs somewhat from
the definition employed by the most authoritative source of
commercial hazardous waste management activity to date, "Hazardous
Waste Generation and Commercial Hazardous Waste Management
Capacity," prepared by Booz-Allen & Hamilton, Inc. and Putnam,
Hayes and Bartlett, Inc., a 1980 report updated in 1981 and
1982. The Booz-Allen report defined members of the commercial
hazardous waste management industry to be "facilities engaged in
the treatment and disposal of hazardous waste for a fee, but
does not include recovery operations, such as those buying and
selling solvents or storage and transfer stations which may be
handling wastes classified as hazardous." These facilities
solely managed hazardous waste — they did not generate any
themselves.
The major differences in these two definitions may be
summarized as follows. First, the definition used in this report
assumes that when a firm manages the waste of another, independent
firm (i.e., not connected through ownership), it receives finan-
cial compensation (a fee) and is thus commercial in nature.
Furthermore, some recovery operations and transfer stations that
were excluded in the Booz-Allen study are included as commercial
83
-------
facilities in this report, due to an inability to systematically
identify such operations in the survey data base and exclude
them. Finally, sites that generated hazardous waste were not
necessarily excluded by the definition of this report as long as
over half of the hazardous waste managed was generated by
independent firms. However, it is believed that much of the
hazardous waste generation occurring at these facilities is
related to hazardous waste management operations (e.g., landfill
leachate, treatment process sludges and residues, etc.), and
would not have caused such facilities to be excluded from the
Booz-Allen study. Nonetheless, it can not be ascertained from
the survey data that the primary business of these sites is
commercial waste management. The Booz-Allen study was targeted
specifically to such facilities.
The Booz-Allen study was undertaken at a time when
attention was focussed on amounts of hazardous waste shipped off
site to commercial facilities for management purposes. It was
thought that most generators shipped some or all of their
hazardous wastes off site for treatment, storage, and disposal.
This is consistent with the findings presented in this report
(see Section 4.4). However, another finding in this report is
that the overwhelming majority of hazardous waste quantities are
generated and managed on site by those generators which do not
ship the bulk of their hazardous waste off site. Thus, esti-
mates of hazardous waste generation and management activities
that are based solely upon an examination of the activities of
commercial off-site facilities are likely to be seriously under-
stated. While the Booz-Allen study attempted to address the
issue of on-site management, it was forced to rely on secondary
data.
84
-------
Under the definition used for this report, the survey
estimates indicate that there were 326 commercial facilities in
1981 (see Figure 13), representing roughly 7 percent of all TSD
facilities. The Booz-Allen estimate was 127 commercial facili-
ties in 1980. Differences in definitions account for some of
the difference in estimates. Nonetheless, the survey results
indicate substantially greater numbers of commercial facilities
than previously estimated. It should be noted that Booz-Allen
did not attempt to do a probability sample of all management
facilities in its effort to identify commercial facilities.
Rather, they contacted those firms that the best information
available (based on data from and discussions with EPA and industry
trade associations) indicated were likely to be involved in the
commercial management of hazardous waste. Since reliable data
bases on the generation and management of hazardous waste in the
U.S. are still in their formative stages, it is not surprising
that the estimates from the two reports differ.
In addition to the 326 commercial facilities estimated
by this survey in 1981, 37 (0.8%) publicly owned or operated
facilities are estimated to have received more than half of
their hazardous waste from other firms during 1981. Another 333
facilities were not specified by kind because of incomplete or
missing data. These may also include facilities that did not
fit clearly into either the private or public categories (e.g.,
quasi-public facilities).
Figure 13 also illustrates the prevalence of commercial
facilities relative to other types of management facilities.
Firms that manage predominantly their own wastes made up the
majority (85.5%) of management facilities. It should be pointed
out, however, that as many as 146 of these 4,122 facilities may
have managed some hazardous waste on a commercial fee basis,
85
-------
Figure 13
NUMBER OF COMMERCIAL VERSUS
OTHER TSD FACILITIES
Total TSD's: 4,818
COMMERCIAL
FACILITIES
326 (6.8%)
333
(6.9%)
MORE THAN 50% of waste
from other firms and publicly
owned oŁ operated
MORE THAN 50% of waste
from other firms and privately
owned and operated
50% OR LESS waste from
other firms
Other*
K
quasi-public facilities)
facillties
86
-------
even though the quantity of such waste was exceeded by the quantity
of hazardous waste managed noncommercially.
The Office of Solid Waste (OSW), in its efforts to
identify commercial facilities among those on file on HWDMS,
does not limit its definition in the way that both this report
and the Booz-Allen study have. Specifically, OSW defines as
commercial any facility that offers its services of hazardous
waste management as a business for a price. Commercial waste
management need not be the primary business activity at the
site. Nor are reclaimers or transfer stations excluded. This
is the broadest definition of commercial facilities; both the
population identified in Figure 13 and the Booz-Allen population
are subsets of this broader population. Using the assumption
stated above, that receipt of hazardous wastes from firms not
owning or owned by the facility is an indicator that hazardous
wastes are managed as a business for a price, the survey
estimates that as many as 509 facilities would have qualified as
commercial facilities in 1981 under OSW's broader definitions.
OSW's reason for employing this broad definition is to
ensure the inclusion of all quantities of hazardous waste managed
in a commercial manner and to identify the broadest population
of facilities that may be available to generators that do not
treat, store, or dispose of all of their hazardous wastes on
site. Alternatively, the definition developed for this report,
which allows comparability with previous studies, ignores
commercial quantities handled by firms at which the commercial
quantities comprise less than 50 percent of the managed waste.
However, the commercial quantities managed by these "non-commercial"
firms may be large (see Section 7.6), particularly for firms
managing exceptionally large quantities of waste. Thus, the OSW
definition of commercial facilities, which includes such facilities,
may be the most appropriate definition.
87
-------
5.4 Number of Facilities Treating, Storing, and/or
Disposing of Hazardous Waste
Survey results indicate that hazardous waste storage
is the most prevalent management activity regulated under RCRA.
Of the 4,818 TSD facilities, 89.2 percent are estimated to have
stored hazardous waste in regulated storage processes during
1981. (See Figure 14.) This estimate excludes any sites where
hazardous wastes were stored (accumulated) on site for less than
90 days in tanks or containers (exempt 90-day accumulation).
Storage is an essential aspect of the hazardous waste management
cycle. Section 5.4.2 below provides a comparative analysis of
the utilization of the different types of regulated storage
processes (containers, tanks, waste piles, and surface impound-
ments) for hazardous waste storage during 1981.
An estimated 1,495 facilities (or 31.0%) treated
hazardous wastes in processes regulated under RCRA during 1981.4
(See Figure 14.) However, a significant, but unestimable number
of additional facilities and generators also treated hazardous
wastes during 1981, but did so in processes that have been
specifically exempted or excluded from regulation under RCRA.
The most prevalent example of these exempt treatment processes
is hazardous wastewater treatment in tanks that are covered by
NPDES permits. As discussed in Section 2.3, lack of a valid ^
sample frame prevented the development of statistically valid
estimates of the number of facilities utilizing such exempt
procedures. Nonetheless, the frequency at which they were
observed during the survey process suggests that the number
6
This number may be slightly overstated to the extent that
respondents may have failed to exclude exempted wastewater
treatment tanks from the TSD General Questionnaire where all
treatment took place exclusively in these tanks.
88
-------
Figure 14
NUMBER OF FACILITIES WITH TREATMENT, STORAGE,
AND/OR DISPOSAL IN 1981
5,000 -
4,500 -
4,000 -
3,500 -
3,000 —
2,500 -
2,000 -
1,500 -
1,000 -
500 -
0 -
4,299*
1,495*
430*
TOTAL
TSD
FACILITIES
= 4,818
TREATMENT
STORAGE
DISPOSAL
Treatment + Storage + Disposal exceeds 4,818 due to multiple processing at facilities
Some treatment facilities shown here may be RCRA-exempt (see footnote on previous page).
89
-------
of regulated treatment facilities estimated by this survey
significantly understates the role played by treatment (both
regulated and exempt) in the management of hazardous waste.
Section 5.4.1 below presents a comparative analysis of the
utilization of the types of regulated hazardous waste treatment
processes (tanks, surface impoundments, and incinerators) for
RCRA-regulated hazardous waste treatment during 1981.
Based on the survey, only an estimated 430 facilities
(8.9%) actually disposed of hazardous wastes during 1981. (See
Figure 14.) However, clarifications are required to properly
interpret this number. To begin with, incineration is regarded
as a treatment process (thermal treatment) by EPA. Therefore,
in this survey, incineration is a form of waste reduction, which
is a form of treatment, not disposal. Thus, facilities that
incinerated hazardous wastes during 1981 are not included among
these 430 sites (unless, of course, they also engaged in
disposal). Furthermore, many other forms of hazardous waste
treatment represent "final" treatment: wastes are treated to
render them nonhazardous (e.g., neutralized). Once so treated,
these wastes are frequently discharged to surface water bodies'
or similarly "disposed" of. However, since the wastes are no
longer hazardous wastes, such "disposal" is not counted as
hazardous waste disposal under RCRA nor in this survey. If such
"final" treatment were to be regarded as disposal under RCRA,
the number of disposal facilities estimated by this survey would
be substantially larger.
Finally, the 430 disposal facilities estimated by this
survey do not include facilities that treated or stored hazardous
wastes in surface impoundments, nor do they include facilities
that stored hazardous wastes in waste piles during 1981 (unless,
of course, such facilities also engaged in disposal). m this '
way, the population of "disposal" facilities estimated in this
90
-------
section differs significantly from the population of "land
disposal" facilities subject to EPA's RCRA regulations. EPA
considers "land disposal" facilities to include all hazardous
waste landfills, land treatment areas, waste piles, and all
types of surface impoundments. Such facilities are subjected to
a variety of specific standards, including, but not limited to,
groundwater monitoring and response, and extensive closure and
post-closure care requirements. Not included among EPA's "land
disposal" facilities, which generally include only those
facilities that are required to perform groundwater monitoring
under RCRA, are underground injection wells and ocean disposers.
Both of these types of facilities are, however, included in the
population of "disposal" facilities estimated in this section,
along with landfills, land treatment areas, and disposal surface
impoundments. There were an estimated 1,049 "land disposal"
facilities (21.8% of the total) actively employing such processes
during 1981. If underground injection wells were to be added to
this group (those not already included due to the operation of
other disposal processes), the estimate would be 1,063 (22.1% of
the total).
Section 5.4.3 below presents a comparative analysis of
the various types of disposal processes observed during the
survey (underground injection wells, landfills, land treatment
areas, and disposal surface impoundments) utilized for hazardous
waste disposal in 1981.
The sum of the percentages presented for treatment
(31.0%), storage (89.2%) and disposal (8.9%) facilities exceeds
100 percent because many facilities employ more than one such
process in their hazardous waste management operations. Together
the survey's estimates of the numbers of treatment, storage, and '
disposal facilities provide an interesting picture of the RCRA-
regulated population of TSD facilities in 1981. More than half
91
-------
of the population does nothing more than store hazardous wastes
(at least 2,893 facilities, which was obtained by subtracting
from the 4,818 facilities all 430 disposal sites and, assuming
conservatively that none of the treatment processes are located
on the same sites as disposal processes, further subtracting all
of the 1,495 treatment sites). Some of these storage sites are
transfer stations (the survey did not distinguish transfer
stations, and therefore cannot estimate their numbers), others
are long-term storage sites or sites that employ waste piles or
surface impoundments. But the majority of these storage facil-
ities would appear to be tanks and container areas employed on
site by generators to store hazardous waste, for periods of more
than 90 days, prior to shipment off site for treatment and dis-
posal.
As indicated in Section 4.4 above, most generators
(62%) ship all their hazardous wastes off site for treatment,
storage and disposal. One possible conclusion that might be
implied in the findings presented in this section is that most
facilities also ship, or intend to ship (since they do not treat
or dispose on site) all their hazardous wastes off site for
treatment and disposal. This conclusion cannot, unfortunately
be confirmed directly from the survey itself, because the facility
questionnaires did not obtain data on shipments from the facil-
ities. Nonetheless, the numbers presented in this section suggest
that a great many, if not the majority, of the facilities regu-
lated under RCRA during 1981 performed only an interim role in
the management of hazardous wastes.
5'4'1 Number of Facilities Treating Hazardous Waste, hv
Treatment Process Type
The survey results revealed that treatment of hazardous
waste in tanks was the most commonly used treatment process
92
-------
during 1981. Approximately 41 percent (or 609) of the 1,495
treatment facilities treated hazardous waste in tanks. About
one out of every eight (12.6%) of the 4,818 TSD facilities
employed treatment tanks in their management of RCRA-regulated
hazardous wastes. In addition, many more facilities indicated
that they treat hazardous wastewaters in tanks covered under
National Pollutant Discharge Elimination System (NPDES) permits.
As is indicated in Section 1.3, such tanks are excluded from
regulation under RCRA, and the survey was not designed to provide
estimates of the numbers of facilities operating tanks under
NPDES permits during 1981. The large number of NPDES tank
facilities and/or processes observed among survey respondents,
however, suggests that treatment in tanks, as a general hazardous
waste management technology, is employed more heavily than the
estimate of RCRA_-regulated tank treatment sites would indicate.
Treatment of hazardous waste in surface impoundments
emerged as the second most commonly used treatment process during
1981. Survey estimates indicate that treatment surface impound-
ments were in use at 27.4 percent (or 410) of the 1,495 treat-
ment facilities. About one out of every twelve (8.5%) of the
4,818 TSD facilities employed treatment surface impoundments in
their management of RCRA-regulated hazardous wastes during 1981.
Unlike treatment tanks, treatment surface impoundments that are
employed to treat hazardous wastewater under NPDES permits are
not excluded from RCRA regulation. Thus the estimated number of
facilities with treatment surface impoundments includes facili-
ties that are only engaged in NPDES wastewater treatment in
surface impoundments, while the estimated number of facilities
with treatment tanks excludes those facilities that are only
engaged in NPDES wastewater treatment in tanks. Therefore, the
estimate of the number of hazardous waste treatment surface
impoundments is not likely to understate the role surface
impoundments play in hazardous waste treatment.
93
-------
Incineration was the least frequently employed treatment
alternative among the three treatment technologies specified in
the survey questionnaire. About 16 percent (or 240) of the
1,495 treatment facilities were estimated to have used incinera-
tion in their treatment of hazardous waste during 1981. One out
of every twenty (5.0%) of the 4,818 TSD facilities employed
incinerators for hazardous waste treatment. While there were
fewer facilities employing incineration than any other hazardous
waste treatment alternative, it should be noted that the estimated
number of facilities that used incineration to treat hazardous
waste was greater than the number of facilities that employed
the most commonly used disposal method, landfilling. (There
were an estimated 240 incinerator sites during 1981, compared to
199 landfill sites; see Section 5.4.3.)
More than 25 percent of the treatment facilities
indicated that they also treated hazardous waste in processes
other than the three treatment processes specified in the survey
questionnaire. Other treatment processes listed by respondents
included open burning, explosion, treatment in containers, and
treatment in waste piles. About 8 percent (one out of every 12)
of the 4,818 TSD facilities listed "other treatment" as one of
their hazardous waste management technologies. In most cases,
facilities reporting "other" treatment methods also reported the
more common methods (tanks, impoundments and incinerators). In
addition, some technologies classified by respondents as "other"
may prove to be classifiable as treatment in tanks or impoundments.
Figure 15 on the following page, presents the estimated
number of sites employing RCRA-regulated tank, surface impound-
ment, incinerator, and other treatment processes in their
management of hazardous waste during 1981. Table 9, following,
presents the percentages of treatment facilities and TSD
facilities that used each treatment process type.
94
-------
Figure 15
NUMBER OF FACILITIES TREATING HAZARDOUS WASTE IN 1981
BY TREATMENT PROCESS TYPE
TREATMENT
TANKS
TREATMENT
SURFACE
IMPOUNDMENTS
INCINERATORS
TOTAL
TREATMENT
FACILITIES
= 1,495*
OTHER
TREATMENT
*Some
of these treatment facilities may be RCRA-exempt (see Section 5.4 footnote)
95
-------
Table 9.
Percentage of Treatment Facilities and of All TSD
Facilities Employing Each Treatment Technology in 1981
Process
Type
Treatment
Tanks
(N=609)
Treatment
Surface
Impoundment
(N=410)
Incinerators
(N=240)
Other
Treatment
(N=392)
Percent of 1,495 Treatment
Facilities Employing
Process Type*
40.7%
27.4%
16.1%
25.8%
Percent of 4,818 TSD
Facilities Employing
Process Type
12.6%
8.5%
5.0%
8.1%
*The sum of the percentages in this column exceeds 100 percent because
some facilities employed more than one type of treatment.
96
-------
5•4•2 Number of Facilities Storing Hazardous Waste.
by Storage Process Type
Survey results indicated that storage in containers
was the most commonly used type of hazardous waste storage during
1981. "Container" is defined by 40 CFR, Subpart B §260.10 as
"any portable device in which material is stored, transported,
treated, disposed of, or otherwise handled." Fifty-five gallon
metal drums are commonly used as hazardous waste containers.
Other containers include 25 gallon plastic drums, tank cars,
tank trucks, and, for some acutely hazardous waste, glass vessels.
An estimated 85 percent (or 3,577) of the 4,299
hazardous waste storage facilities used storage in containers as
a method of storing hazardous waste. Almost three out of four
(74.2%) of all 4,818 TSD facilities used containers for storing
hazardous waste. In addition, many more TSD facilities indicated
that they accumulated hazardous waste generated on site, for
less than 90 days, in containers. As is indicated in Section
1.3, on-site accumulation of hazardous waste for less than 90
days in containers is excluded from most regulation under RCRA.
The TSD facility survey was not designed to estimate the number
of facilities using on-site accumulation in containers. Estimates
from the TSD Generator Questionnnaire, however, suggest that,
during 1981, about 51 percent of the 14,098 hazardous waste
generators used containers to accumulate hazardous waste for
less than 90 days.
Storage in tanks emerged as the second most commonly
used type of hazardous waste storage during 1981. This storage
method was used by 33.2 percent of the 4,299 hazardous waste
storage facilities during 1981. Better than one out of every
four (29.6%) of the 4,818 TSD facilities stored hazardous waste
in tanks during 1981. In addition, many more facilities indicated
97
-------
that they used tanks to accumulate hazardous waste generated on
site for less than 90 days. As indicated in Section 1.3, on-site
accumulation of hazardous waste in tanks for less than 90 days
is excluded from most regulation under RCRA. The TSD facility
survey was not designed to estimate the number of facilities
using on-site accumulation in tanks. Estimates from the Generator
Questionnaire, however, suggest that, during 1981, about 15
percent of the 14,098 hazardous waste generators used tanks to
store hazardous waste for less than 90 days.
Hazardous waste storage tanks that are used exclusively
as part of a wastewater treatment system covered under a National
Pollutant Discharge Elimination System (NPDES) permit are exempted
from regulation under RCRA. This survey of RCRA-regulated facil-
ities was not designed to provide estimates of the number of
facilities that operated storage tanks under NPDES permits during
1981. However, the large number of NPDES facilities and/or
processes observed among survey respondents (both in the Tank
Questionnnaire data file and in the Generator Questionnaire data
file) suggests that storage in tanks, as a general hazardous
waste management technology, is employed much more frequently
than the estimate of RCRA-regulated storage in tanks suggests.
About one-eighth of the 4,299 storage facilities (12.8%
or 552 facilities) stored hazardous waste in surface impoundments
during 1981. About one in every nine (11.5%) of all 4,818 TSD
facilities stored hazardous waste in surface impoundments in
1981. The 90-day storage exclusion rule that applies to on-site
accumulation of hazardous waste in tanks and containers (see
above, and Section 1.3) does not apply to generators who store
hazardous waste in storage surface impoundments. Thus, unlike
the counts for containers and storage tanks, the counts for
storage surface impoundments include impoundments in which
hazardous waste was stored for any amount of time, by any type
98
-------
of facility, and are therefore not likely to understate the role
played by surface impoundments in hazardous waste storage.
Furthermore, unlike storage tanks, storage surface
impoundments that are employed to store hazardous wastewater
under NPDES permits are not excluded from RCRA regulation.
Thus, the estimated number of facilities with storage surface
impoundments includes facilities that are only engaged in NPDES
wastewater storage in surface impoundments, while the estimated
number of facilities with storage tanks excludes those facilities
that are only engaged in NPDES wastewater storage in tanks.
Hazardous waste storage in waste piles was the least
frequently used storage method among the four storage methods
listed in the TSD General Questionnaire. Approximately 4.0
percent (or 174) of all of the 4,299 storage facilities and 3 6
percent of all 4,81Ł TSD facilities used waste piles as a method
of storing hazardous waste during 1981. Less than 1 out of
every 25 TSD facilities stored hazardous waste in waste piles in
1981. At the time this survey was developed, waste piles were
regarded by the EPA as only a storage process. Waste piles are
now recognized as a possible disposal process, and are therefore
regulated under more stringent standards (e.g., requiring ground
water monitoring). since the survey dealt only with storage
waste piles, estimates of the number of waste piles used for
treatment and/or disposal cannot be made.
Approximately 3 percent of the 4,299 storage facilities
(3.2% or 139 facilities) stored hazardous waste in storage
methods other than containers, tanks, impoundments or waste
piles. These facilities represent 2.9 percent of all 4,818 TSD
facilities.
99
-------
Figure 16 presents the number of container, tank,
surface impoundment, waste pile and other storage facilities
that stored hazardous waste in 1981. Table 10 presents the
percentage of storage facilities and TSD facilities that employed
each storage process type.
5'4*3 Number of FacjJJ.U^Disggsin^ Waste, JDV
Disposal Process Type
According to survey results, the most frequently used
method of disposal of hazardous waste during 1981 was landfilling
Almost one half (46.3% or 199) of the 430 disposal facilities
used landfills for disposal of hazardous waste during 1981
About 1 in 25 (4.1%) of all 4,818 TSD facilities actively iand-
filled hazardous wastes in 1981.
Disposal in surface impoundments emerged as the next
most frequent method of hazardous waste disposal in 1981. This
technology was used by 27.0 percent (or 116) of the 430 disposal
facilities. About 1 out of every 40 (2.4%) of the 4,818 TSD
facilities used surface impoundments for disposal of hazardous
waste.
Disposal in underground injection wells was the third
most frequently used method of hazardous waste disposal in 1981,
with approximately one-fifth of the disposal facilities using
this method. About 20.2 percent (or 87) of the 430 disposal
facilities used injection wells for disposal in 1981. This
represents approximately 1 out of every 55 (1.8%) of the 4 818
TSD facilities. As indicated in Section 7.7.3, however, under-
ground injection is estimated to have accounted for the
overwhelming majority of the quantity of hazardous wastes
disposed of during 1981.
100
-------
Figure 16
NUMBER OF FACILITIES STORING HAZARDOUS WASTE
IN
1981, BY STORAGE PROCESS TYPE
CONTAINERS
STORAGE
STORAGE
IMPOUNDMENTS
WASTE P,LES
OTHER
MORAGE
"Containers. + Tanks + Impoundments + Piles + Other exceeds 4 9QQ
some facilities used more than one storage method exceeds 4-^yy
101
-------
Figure 17
NUMBER OF FACILITIES DISPOSING OF HAZARDOUS
WASTE IN 1981, BY DISPOSAL PROCESS TYPE
440 -
420 -
400 -
380 -
360 -
340 -
320 —
300 -
280
260 -
240 -
220 -
200 -
180 -
160 -
140 -
120 -
100 -
80 -
60 -
40
20
0
TOTAL
DISPOSAL
FACILITIES
-430**
199*
116'
87*
70=
7*
LANDFILLS
DISPOSAL
bUKhACE
IMPOUNDMENTS
INJECTION
LAND
APPLICATION
(TREATMENT)
OTHER
DISPOSAL
+ Other exceeds 430 because
**Does
not include waste piles or surface impoundments used for treatment or storage of hazardous
waste.
104
-------
Table 11. Percentage of Disposal Facilities and of all TSD
Facilities Employing Each Disposal Technology in 1981
Process
Type
Landfills
(N=199)
Disposal
Surface
Impoundments
(N=116)
— — . __
Injection Wells
(N=87)
Land Appli-
cation
(Treatment)
(N=70)
Other Disposal
(N=7)
— , .
Percent of t
Facilities E
Process Type
4
2
2
1
46.3%
27.0%
20.2%
16.3%
1.6%
Percent of 4,818 TSD
Facilities Employing
Process Type
4.1%
2.4
1.8%
1.5'
0.1%
the PerSentageS in this column exceeds 100% because sonv
employed more than one type of disposal.
105
-------
Table 12. Age Distribution of TSD Facilities
Age (in years)
Percentage Cumulative
in Class Percentage
" 29-0 29.0
2"\ 11-4 40.4
*-* 15-6 56.0
7~15 18.7 74 7
16-30 14.2 88*9
31-40 6.4 '
40+ 4.7
In 1981 over half of the facilities' waste management
operations were six years old or less. In fact, almost 30 percent
of the facilities commenced waste management operations in 1980
or 1981. Three-fourths of the facilities began waste management,
at most, 15 years ago, while nearly 90 percent began within the
past 30 years. This heavy concentration of new waste management
facilities corresponds to the recent attention and concern given
to the problems of handling hazardous waste, but may also be due
to the pre-RCRA closure of older facilities. The survey was not
designed, however, to estimate the number of such closures.
5.5.2 Ownership Status of TSD Facilities
The ownership of facilities involved in the treating,
storing, or disposing of hazardous waste in 1981 was predominantly
private. Table 13 shows the ownership status of the facilities.
Privately owned TSD facilities accounted for 90 percent
of the population. Those facilities solely owned by the Federal
government represented approximately four percent of the popula-
tion. Solely state and solely local owned facilities accounted
for about one percent each, while the remaining three percent
fell in a miscellaneous or "other" categorization of ownership.
106
-------
Table 13. Ownership Status of TSD Facilities
Ownership Status
Solely Federal government 4>3
Solely State government j*3
Solely local government 1"1
Privately owned Qn*n
Other
*Does not add to 100 due to rounding.
5.5.3 Operator_Status_of TSD Facilities
As with ownership, the operation of facilities treat-
ing, storing, and disposing of hazardous waste was predominantly
private in 1981. Table 14 shows the operator status of the
facilities.
Table 14. Operator Status of TSD Facilities
Operator Status Per^ejitage_^f_jrotal
Solely federal government 3 2
Solely state government 1 *2
Solely local government 0*5
Privately operated Q9*A
Other 27
The distribution of operator status is very similar to
that of ownership status. Private operators account for over 92
percent of the population, while government operation is repre-
sented by 3 percent Federal, 1 percent State, and one-half percent
local. The remaining 3 percent is accounted for by "other"
approaches to handling the operation of TSD facilities.
107
-------
Comparing the distributions of ownership and operator
status, the survey results indicate that a portion of government
owned facilities are actually operated by the private sector.
Examples include private research facilities located at govern-
ment installations, private industries located at military instal
lations, and the contracting out of the operations of waste
management facilities to private enterprises by State and local
governments. The end result of these factors, combined with the
heavy concentration of facility ownership in private hands, is
that the day-to-day responsibility for the treatment, storage,
and disposal of hazardous waste in the United States is being
managed almost exclusively by private enterprise.
108
-------
PART III (Sections 6-7)
QUANTITIES
OF
HAZARDOUS WASTE
-------
-------
INTRODUCTION TO PART III:
QUANTITIES OF HAZARDOUS
WASTE GENERATED AND MANAGED
As detailed in Section 1.2, one of the purposes of
conducting the survey was to develop estimates of the quantity
of RCRA-regulated hazardous wastes generated during 1981, and
the quantities of hazardous wastes treated, stored, and disposed
of in RCRA-regulated processes during 1981. Part II (Sections
4 and 5) of this report summarized the survey's major findings
concerning the numbers and characteristics of the generators
and TSD facilities regulated under RCRA. Part III (Sections 6
and 7) presents a summary of survey results pertaining to the
various quantities of hazardous waste regulated under RCRA
during 1981. Survey-based estimates of the quantities of
hazardous waste generated in 1981 are presented initially
(Section 6), followed by a presentation of the estimated
quantities of hazardous waste that were managed (treated,
stored, and disposed of) in processes subject to regulation
under RCRA during 1981 (Section 7).
Before presenting the hazardous waste quantity estimates,
however, it is necessary to present a brief overview of those
aspects of the survey design that particularly affected the
quantification of hazardous waste generation and its subsequent
treatment, storage, and disposal under RCRA. The survey design
played an important role in defining the specific nature of the
quantity estimates developed from the obtained data, and should
be understood clearly in the interpretation of the findings
presented in Sections 6 and 7 below.
109
-------
Statistical Reliability of Quantity Estimates
As detailed in Section 3.3, the survey's estimates
concerning quantities of hazardous waste generated and managed
in 1981 are subject to substantially greater statistical
uncertainty than are its estimates of population characteristics
This greater uncertainty results, primarily, from two factors:
1) The survey samples were designed to provide more
2) The populations of hazardous waste generators and
TSD facilities were both found to be highly skewed
in terms of the quantities of hazardous wastes
they generate and manage; so skewed, in fact,
that very small proportions of the populations
account for nearly all of the quantities generated
and managed (see Figures 3 arid 4).
These two factors, generally categorized as "sampling
error" (see Section 3.1), are responsible for the wide confidence
intervals surrounding the quantity estimates presented in the
following sections.
Sampling error had a particularly severe impact on the
quantity estimates derived from the Generator Questionnaire.
As discussed briefly in Section 3.3 and summarized in Table 3,
sampling error related to the design of the Generator Survey
and the skewed distribution of the generator population resulted
in a confidence interval of plus or minus 79.7 percent surrounding
the Generator Questionnaire's estimate of total quantity of
hazardous waste generated during 1981. This confidence interval
is 63 percent wider than the interval surrounding similar estimates
derived from the TSD General Questionnaire. The reasons behind
the differences between these confidence intervals are essentially
related to the greater impact of sampling error in the Generator
Survey than in the TSD Facility Survey: sampling error was
110
-------
reduced in the survey of TSD facilities because, by focusing on
sites with TSD facilities, the TSD survey included a higher
proportion of the larger sites than did the Generator Survey.
Since the larger sites account for the greatest proportions of
the quantities generated and managed, their heavier representation
in the facility sample served to improve the precision of the
estimates.
The severe impact of sampling error in the Generator
Survey prompted the development of an alternate mechanism for
deriving an estimate of the total quantity of hazardous waste
generated during 1981. This alternate mechanism, described in
detail in Section 6, essentially divides the generation estimate
into two components: the quantity of hazardous waste generated
by generators that do not operate on-site TSD facilities, and
the quantity generated by generators that do operate on-site
TSD facilities. Since the TSD facility survey is acknowledged
to provide more precise quantity estimates for the subpopulation
of sites with TSD facilities, a proxy generation estimate was
derived from the TSD General Questionnaire and substituted in
place of the same estimate provided by the Generator Questionnaire
for that segment of the population. The Generator Questionnaire
is still utilized to derive the other piece of the total generation
estimate, that from generators without on-site TSD facilities.
These two estimates are then added together to form the estimate
for the total quantity of hazardous waste generated in 1981.
Development of this alternate mechanism for estimating
the total quantity of hazardous waste generated in 1981 resulted
in a significant reduction in the uncertainty surrounding that
estimate, due to greater reliance upon that part of the survey,
the TSD facility sample, that was subject to reduced sampling
error. The alternate estimation mechanism also provides for
111
-------
greater consistency between the estimates of hazardous waste
generation and its subsequent management, as discussed in detail
in Sections 6 and 7.
Nonsampling error also is believed to be involved to a
greater extent in relation to the quantity estimates (from both
the Generator and TSD General Questionnaires) than in relation
to the population characteristic estimates presented in Part II
As discussed in detail in the following section, greater
nonsampling error is associated with the quantity estimates due
to the high degree of complexity inherent in trying to measure
the quantities of hazardous waste subject to regulation under
RCRA.
As a result of the greater uncertainties (related to
sampling and nonsampling error) associated with the survey's
quantity estimates, and due to the fact that the survey was
designed to provide national estimates, regional breakdowns of
the estimates of the quantities of hazardous wastes generated
and managed in 1981 are not provided. Regional breakdowns of
the numbers of generators and TSD facilities are provided in
Part ii, despite the national nature of the survey design, due
to the high level of precision obtained for estimates of
population characteristics. Such precision was not obtained
for the quantity estimates and, as such, presentation of
regional breakdowns of the quantity estimates could be
seriously misleading.
Other quantity data presented in the following sections
should be interpreted and utilized within the context of the
discussion of statistical reliability issues presented in
Sections 3.3 and 3.4.
112
-------
Definition of Hazardous Wastes Included in Survey
Estimates
Two features of EPA's RCRA hazardous waste regulatory
program significantly impacted the quantity estimates presented
in the following sections: 1) rules governing the mixtures of
hazardous and nonhazardous wastes; and 2) exemptions and exclusions
of specific wastes and specific waste management processes from
regulation under Subtitle C of RCRA. Each of these issues
contributed to the complexities faced in attempting to measure
the quantities of hazardous wastes regulated under RCRA during
1981; each contributed to increased uncertainty surrounding the
survey's quantity estimates in the form of "nonsampling error"
(see Section 3.1). The ramifications of RCRA's exemptions and
exclusions will be addressed initially, followed by a discussion
of the impacts of the mixture rules on the survey's quantity
estimates.
Exemptions and Exclusions
As stated in Section 1.3, the national survey was
designed to estimate the quantities of hazardous waste that
were generated in 1981 and were or were intended to be managed
subsequently in processes subject to regulation under RCRA.
This represents a very narrow definition of the types of
hazardous wastes that are included in the survey's quantity
estimates, and needs to be understood clearly in order to
properly interpret the findings presented in the following
sections.
113
-------
Specifically, two categories of waste streams are not
included in the survey's hazardous waste quantity estimates:
1) Wastes that have been exempted or excluded from
regulation under RCRA as hazardous wastes; and,
2) RCRA-regulated hazardous wastes that were generated
in 1981 but that were not, at any point in the
management process, treated, stored, or disposed
of in processes subject to regulation under RCRA.
Most prominent among the waste streams that have been excluded
from regulation as hazardous wastes under RCRA are those waste
streams sent to publicly owned treatment works (POTW's)l and
waste streams mixed with domestic sewage [see 40 CFR §261.4(a)(l)
and 40 CFR §265.1(c)(3)]. Other such excluded waste streams
are listed specifically in the introduction to Section 4. The
survey design specifically excluded such waste streams from the
estimated quantities of hazardous waste generated in 1981.
More importantly, however, the survey design also
excluded from its estimates quantities of RCRA-regulated hazardous
wastes that were managed exclusively in processes exempted or
excluded from regulation under RCRA. The most prevalent examples
of such survey-excluded hazardous waste streams are those that
were treated exclusively irf wastewater treatment processes
whose discharges are covered under NPDES permits and where the
treatment occurred exclusively in tanks (as opposed to surface
impoundments). Such tank treatment systems are excluded from
regulation under RCRA, as detailed in EPA's November 17, 1980
Federal Register announcement (a copy of which is included in
Appendix C), even though the waste streams treated therein may
still be hazardous wastes as defined under RCRA.
IpOTW's are defined in Section 502(4) of the Clean Water Act.
-------
The survey excluded such quantities of hazardous waste
from its statistical estimates of 1981 hazardous waste generation
due to an inability to properly measure such quantities at the
management stage in the hazardous waste cycle. This inability
results from the absence of a valid sample frame from which to
select a sample of such processes for purposes of developing
statistical data. Whereas owners and operators of RCRA-regulated
hazardous waste treatment, storage, and disposal processes are
required to file Part A permit applications indicating the
presence of such processes at particular sites, owners and
operators of excluded processes need not have submitted Part A
applications. Without a valid sample frame, it is not possible
to determine the number of such processes in operation or the
quantities of hazardous waste that are managed exclusively in
such processes. Some Part A applications were submitted for
such processes, inadvertently, due to lack of information about
the exclusions or because other processes at the site required
submission of Part A applications and the excluded processes
were included on the site's form, even though they were not
required to be included. However, EPA has no way of determining
the number of exempt processes for which Part A applications
were not filed.
In order to preserve the integrity of the quantity
estimates intended to be developed through the survey, all
hazardous waste streams that were managed exclusively in exempt
or excluded processes were excluded from the survey, at both
the management and generation stages of the hazardous waste
cycle. The reason for excluding such quantities from the
generation estimates was that the survey was designed to produce
comparable estimates of 1981 hazardous waste generation and its
subsequent management (treatment, storage, and disposal). In
designing the survey along these lines, OSW and Westat intended
to observe essentially the same 1981 national hazardous waste
115
-------
stream in both the generation and management surveys; that is,
it was intended that the TSD Facility Survey would account for
and describe the management of all, or most, of the hazardous
wastes estimated to have been generated in the Generator Survey.
These design criteria contributed to the design and
structure of the survey questionnaires and samples. Limitations
on the ability to collect quantity data at the management stage
of the hazardous waste cycle also limited the nature and scope
of the data that were requested at the point of generation.
Since one of EPA's primary goals in conducting the survey was
to estimate the impact of its regulatory standards on the
regulated population, the survey was not designed to collect
information about handlers, waste streams, or waste management
activities that fall outside the scope of the RCRA regulatory
program.
The decision rule for including hazardous wastes in
the survey essentially stated that those wastes that did not
encounter, or were not intended to encounter a RCRA-regulIted
hazardous waste treatment, storage, or disposal process were to
be excluded from the survey's scope, while any hazardous wastes
that were managed, at__a_niL^o_int in the treatment, storage, and
disposal cycle, in processes subject to regulation under RCRA,
were to be included in the survey's scope. The following two'
examples help to clarify the impact of these restrictions on
the survey estimates presented in the following sections.
Example A: Situation
Firm A generated two million gallons of a corrosive
wastewater stream during 1981. Since the pH of
the stream was less than two, the wastewater
stream would bo considered a hazardous waste
under RCRA. However, the firm employs a wastewater
Treatment system that consists entirely of a
116
-------
series of tanks that neutralize the wastewater
stream and remove other contaminants. The tanks'
discharge into a nearby river is covered under an
NPDES permit, and is therefore excluded from
regulation under RCRA. The wastewater stream
feeds directly into the NPDES tank treatment
system and does not pass through any other
treatment or storage process.
Quantities Included in purvey
Since none of the hazardous waste generated by
Firm A passed through processes regulated under
RCRA during 1981, none of the two million gallons
generated would have been included in either the
estimate of hazardous waste generation or the
various estimates of hazardous waste management,
including total hazardous waste managed, hazardous
waste entering treatment, and hazardous wastes
treated in tanks.3
Example B: Situation
Firm B generated a waste stream identical to that
generated by Firm A. However, Firm B employs a
wastewater treatment system that utilizes surface
impoundments instead of tanks. Firm B's treatment
system has an NPDES permit, but since the treatment
occurs in surface impoundments, as opposed to
tanks, the system is not eligible for the RCRA
NPDES exemption.
Quantities_In_cluded in Survey
Since the hazardous wastes generated by Firm B
passed through a treatment process regulated
under RCRA during 1981, the entire hazardous
waste stream generated by Firm B would have been
included in the estimate of hazardous waste
generation and' in the various estimates of
hazardous waste quantities managed in RCRA-
regulated processes during 1981.
Generated quantities, in this case, could only be excluded
using information from an associated Tank Questionnaire or
supplied by respondents in marginal notes, waste stream
descriptions, or telephone interviews during the data editing
and cleaning process. For many generators, however, there was
no way of knowing from the survey data that these quantities
should be excluded.
-------
EPA would have received a Notification form from Firm A
indicating hazardous waste generation. As such, Firm A would
have been eligible to be sampled as a generator. However,
Firm A is not required to have submitted a Part A permit applica-
tion to EPA indicating hazardous waste treatment, storage, or
disposal. Therefore, it would not have been possible to survey
Firm A as a TSD facility. since generation and management
activities were the foci of largely independent surveys,
situations like these would have contributed to the estimates
of hazardous waste generation while not contributing to the
estimates of hazardous waste quantities managed, had the survey
not been restricted as indicated. The result would have been
incomparable estimates of hazardous waste generation and its
subsequent management. Furthermore, since not all firms were
aware of this exemption at the time they filed their Part A
applications or even at the time they responded to the survey,
the resulting estimates would have been based upon a mixture of
regulated and nonregulated waste streams, leaving little or no
clarity in the obtained data and substantially reducing the
value of the derived estimates. By restricting the focus of
the survey to those hazardous waste streams actually managed in
processes regulated under RCRA during 1981, the survey yields
valuable, clearly defined information about a specific aspect
of the RCRA regulatory program, even though it purposely fails
to answer other questions of interest.
Mixtures of Hazardous and Nonhazardous Wastes
The final issue requiring attention prior to the
presentation of the survey's 1981 hazardous waste quantity
estimates centers upon the impact of the RCRA "mixture rule" on
those estimates. According to 40 CFR 261.3(a)(2)(ii), a solid
waste (as defined in §261.2) is defined as a hazardous waste if
118
-------
it is a mixture of a solid waste (including "aqueous" solid
wastes) and one or more hazardous wastes (as listed in Subpart D
of Part 261 and not excluded under §260.20 and §260.22). Further
specifications to this rule are provided in §261.3. However,
the essence of the rule is that nonhazardous waste material
becomes hazardous waste when it is mixed with hazardous waste,
and must be managed as such. The implication of this rule is
that in some cases, initially small quantities of hazardous
wastes can result in huge quantities of hazardous waste through
their mixture with large volumes of nonhazardous wastes or
substances. Examples of such occurrences were observed in the
data obtained through the Generator Questionnaire, as well as
in the survey of TSD facilities, including cases where trace
quantities of specific listed hazardous wastes were mixed with
large volume industrial process waters, resulting in very large
volumes of hazardous waste generated.
Hazardous waste quantities reported by many, if not
most, of the largest generators and facilities observed in the
survey were affected significantly by the mixture rule. The
mixing process, however, varied considerably from case to case,
with the mixture occurring anywhere from deep within the industrial
processes themselves through the storage and treatment process
and up to the point of disposal. The survey questionnaires
were not, however, designed to provide separate quantity data
for the various portions of these mixtures (i.e., obtained data
do not indicate percent solids, percent listed wastes, percent
water, etc.). The RCRA regulations are clear, however, in
their intent to require these mixtures to be managed as hazardous
wastes. [Certain exceptions to the rule are provided in 40 CFR
§261.3(a)(2)(iv).] It is therefore appropriate, in fact
mandatory, that these quantities be included within the
estimates of hazardous waste generation provided in the following
sections. As indicated previously, however, respondent error
119
-------
in correctly reporting quantities affected by the mixture rule
(both underreporting and overreporting) may not always have
been detected during the coding process, and therefore contributes
to the uncertainty surrounding the estimates of 1981 hazardous
waste generation.
Another mixture issue affecting the TSD quantity
estimates developed through the survey relates to the management
of nonhazardous wastes as hazardous wastes in hazardous waste
treatment, storage, and disposal processes. Since respondent
error in reporting such quantities could not always be identified,
a more precise label for the hazardous waste management quantity '
estimates presented in this section would be that of "wastes
managed as hazardous wastes."
120
-------
6. QUANTITIES OF HAZARDOUS WASTE GENERATED
The findings presented in this section of the report
are designed to estimate the quantities of hazardous wastes
generated in the U.S. by activities of generators that generated
RCRA-regulated quantities of hazardous waste during 1981 and
whose hazardous wastes were or were intended to be managed
subsequently in treatment, storage, and disposal processes
subject to regulation under RCRA. Section 4 of this report
cites the definitions of RCRA-regulated hazardous waste generators,
and further identifies those types of generators and those waste
streams that have been specifically excluded or exempted from
regulation under RCRA.
The introduction to this part of the report (Part III:
Sections 6 and 7), describing specifically the impact of RCRA's
complex rules and exemptions on the nature of these quantity
estimates, should be read carefully in conjunction with any
interpretation or use of the estimates presented in this section.
As noted in the introduction to Part III, sampling
error was extremely large for quantity estimates derived from
the Generator Questionnaire. The Generator Questionnaire's
estimates were designed to serve as the basis for most, if not
all, of the estimates presented in this section. The large
sampling error resulted in extremely wide confidence intervals
around the generation quantity estimates; confidence intervals
were nearly two thirds larger than those surrounding similar
estimates from the TSD General Questionnaire, which was designed
to produce estimates describing the post-generation stages of
the hazardous waste management cycle.
121
-------
As a result of these very large confidence intervals,
an alternate mechanism was developed to derive the survey's
estimate of the total quantity of hazardous waste generated in
1981. This alternate mechanism, which draws upon data collected
in both the Generator and TSD General Questionnaires, is described
in detail in Section 6.1, below, in conjunction with the presenta-
tion of the estimate itself. it is important to note at this
point, however, that this alternate mechanism could not be refined
sufficiently to serve also as the basis for the additional
estimates provided in this section. Accordingly, the estimates
presented in Sections 6.2 through 6.6 continue to be based
exclusively upon data obtained through the Generator Questionnaire
and are thus subject to substantially greater uncertainty than
other quantity estimates presented elsewhere in this report.
Further discussion of the relationship between these additional
estimates and the estimate of total 1981 hazardous waste generation
is presented at the end of Section 6.1, and should be referred
to in conjunction with their interpretation and use.
As indicated above, the survey's estimate of the total
quantity of hazardous waste generated during 1981 is presented
in Section 6.1. The additional breakdowns of this estimate,
corresponding, in part, to the breakdowns of the numbers of
generators presented in Sections 4.1 through 4.5, are presented
in Sections 6.2 through 6.6. Section 6.7 then concludes the
summary of the survey's findings regarding 1981 hazardous waste
generation with a discussion and assessment of the various
factors contributing to potential increases and decreases in
annual generation rates of hazardous wastes subject to regulation
under Subtitle C of RCRA.
122
-------
6-1 1981 Hazardous Waste Generation Estimates
Based upon the definitions specified in the introduc-
tion to Part III of this report and the survey results, an
estimated 71 billion gallons (approximately 264 million metric
tonnes) of RCRA-regulated hazardous waste were generated in the
United States during 1981. Because of the wide variation in the
quantities generated by individual generators and other factors
mentioned in Section 3.3 and detailed below, this estimate is
rather rough. A more precise statement of the statistical reli-
ability of this estimate, based upon calculations of sampling
error, would be that we are very confident (95% confident) that
the true quantity of hazardous waste generated in 1981 was between
35.7 and 106.3 billion gallons. Despite this wide confidence
interval, the results of the survey clearly indicate that there
were significantly greater quantities of hazardous waste gener-
ated in 1981 than previously estimated.
Prior to discussing the statistical details surrounding
this estimate, however, it is important to highlight the fact
that this 71 billion gallon estimate differs considerably from
the preliminary survey findings released by EPA on August 30,
1983. When the survey results were initially tabulated, the
quantity of hazardous waste generated during 1981 was estimated
to have been 40 billion gallons (approximately 150 million
metric tonnes). Even at that time, however, EPA believed, and
indicated in its public presentation of the preliminary findings,
that the 40 billion gallon estimate had a greater likelihood to
understate, rather than overstate, the actual quantity of hazardous
waste generated in 1981. Two factors led to this belief:
1) Even though the +79.7 percent confidence interval
associated with the 40 billion gallon estimate
indicated that the actual quantity generated in
1981 could have ranged from 8 billion to 72 billion
123
-------
gallons, actual generation data in hand (without
applying the statistical weighting factors) from
the respondents to the Generator Questionnaire,
when added to the quantities believed to have
been generated by respondents to the TSD General
Questionnaire that did not receive the Generator
Questionnaire, substantially exceeded the lower
bound of that confidence interval; and,
2) The estimate of the total quantity of hazardous
waste managed by TSD facilities during 1981,
derived from the TSD General Questionnaire,'was
substantially larger than the 40 billion gallon
generation estimate, even though most of the
facilities included in the TSD sample turned out
to be located at the sites of hazardous waste
generators.
Subsequent to the release of the preliminary findings,
the 40 billion gallon estimate was revised as a result of editing
to a still preliminary 42 billion gallon estimate. Further
analysis of the data obtained through both the Generator Question-
naire and the TSD General Questionnaire, however, revealed that
this 42 billion gallon estimate (that appeared in early drafts
of this report) continued to substantially understate the actual
quantity of hazardous waste generated. Accordingly, efforts
were undertaken to develop an alternate mechanism for deriving
the total quantity estimate. These efforts proved to be
successful, and resulted in the production of the survey's
final estimate of 71 billion gallons of hazardous waste generated
in 1981.
Essentially, the difference between the survey's
preliminary and final estimates rests in the fact that the
survey's 42 billion gallon preliminary estimate was based
exclusively upon data obtained from generators through the
Generator Questionnaire, while the final estimate of 71 billion
gallons presented in this report is derived through a combination
of appropriately weighted data obtained through both the Generator
124
-------
and TSD General Questionnaires. The reasons behind the shift
away from total reliance upon the Generator Questionnaire and
the decision to utilize data obtained through the TSD General
Questionnaire in estimating the total quantity of hazardous
waste generated in 1981 are described in detail below.
The Generator Questionnaire was mailed to a randomly
selected sample of RCRA regulated hazardous waste generators
(see Section 2.3 and Appendix A). Each respondent to the
Generator Questionnaire was asked to report the total quantity
of hazardous waste generated during 1981. At the conclusion of
the survey field period, responses from eligible respondents
(RCRA-regulated hazardous waste generators) were entered into
computer data files and assigned weights, based primarily upon
their probability of being selected for the sample (nonresponse
adjustments were also used to refine individual base weights,
as detailed in Appendix A).
The 42 billion gallon preliminary estimate was then
derived through a simple summation of the weighted responses
(each respondent's generation quantity multiplied by its assigned
weight). Identical procedures were used to derive all of the
generator statistics presented in Section 4, as well as for all
of the TSD facility statistics (characteristics and quantities)
presented in Sections 5, 7, and 8 of this report (of course,
the TSD facility statistics were derived from the TSD General
Questionnaire and its associated process-specific component
questionnaires, not from the Generator Questionnaire).
Had the quantities of hazardous waste generated by
individual generators been approximately normally distributed,
this procedure for estimating the total quantity generated by
that population would have produced a reasonably reliable
estimate, since the survey enjoyed an exceptionally high response
125
-------
rate, a large sample size was obtained (2,084 eligible respon-
dents to the Generator Questionnaire), and a reasonably valid
sample frame (EPA's Notification and Part A/Telephone Verifica-
tion files) was used to select the statistical sample. As
indicated in Section 3.3, however, the size measures for the
populations of RCRA-regulated generators and TSD facilities
were not found to be normally distributed. Rather, both popu-
lations were found to be highly skewed; so skewed, in fact,
that nearly all of the total quantities of hazardous wastes
generated and managed were accounted for by very small propor-
tions of the respective populations (see Figures 3 and 4).
As a result of these highly skewed populations, and
due to the fact that the survey samples were designed to provide
more accurate characteristic data (for reasons described in
Section 2.3), the survey's quantity estimates are subject to
high degrees of sampling error, or statistical uncertainty.
Accordingly, the survey's quantity estimates have much wider
confidence intervals associated with them than do the survey's
characteristic estimates (e.g., the estimated numbers of gener-
ators and facilities, etc.).
The quantity estimates derived from the Generator
Questionnaire, however, are substantially more affected by the
problem of sampling error than are similar quantity estimates
derived from the TSD General Questionnaire. As indicated in
Table 3, the +79.7 percent confidence interval surrounding the
Generator Questionnaire's 42 billion gallon preliminary genera-
tion estimate is 63.3 percent wider than the confidence interval
surrounding the TSD General Questionnaire's estimate of the
total quantity of hazardous waste managed by TSD facilities
during 1981. Analyses of the factors responsible for the
extremely high degree of sampling error associated with the
126
-------
quantity estimates derived from the Generator Questionnaire led
to the development of the alternate approach for deriving the
estimate of the total quantity of hazardous waste generated
during 1981.
The generator survey's greater potential for understating
the total amount of hazardous waste generated nationally, is a
result of the fact that the generator survey sample included a
smaller number of the more important larger generators. Due to
the highly skewed size distribution found in the population of
hazardous waste generators, inclusion of a greater proportion
of the larger generators in the generator sample would have
improved the generation estimate. The TSD facility sample did,
however, include larger numbers of the important larger facilities.
The occurrence of a greater number of larger sites in the TSD
facility sample than in the generator sample is due to two
factors:
1) Treatment, storage, and disposal facilities, for
reasons related primarily to economies of scale,
tend more frequently to be located at the sites
of larger generators as opposed to smaller
generators; and,
2) The generator sample was made up of far more
generators that do not operate on-site TSD facilities
than of generators that do operate on-site TSD
facilities.
The composition of the generator sample generally
reflected the characteristics of the generator population as a
whole. As indicated in Section 2.3, both the generator and TSD
process component samples were drawn using equal probability of
selection (within strata) sampling techniques, rather than
probability of selection proportionate to size (PPS) sampling
techniques. PPS sampling, by sampling larger sites with a
greater likelihood of selection than in an equal probability
127
-------
approach, tends to produce more accurate quantity estimates.
PPS sampling could not be used in this survey, however, due to
the absence of adequate size data at the time the samples were
drawn. Thus, neither sample attempted to focus on the important
larger generators and facilities. Under equal probability
sampling, the resulting samples of eligible respondents tend to
resemble, in their composition, the major distributional charac-
teristics of the populations from which they were drawn.
Figure 18 illustrates the overlap that exists among
the populations of generators that submitted RCRA Notification
forms (approximately 55,000) and TSD facilities that submitted
Part A permit applications (approximately 10,000 at the time
the survey samples were drawn; see Section 5.1). While only a
small proportion of the notifying generators indicated (through
additional submission of Part A applications) that they operated
on-site TSD facilities, most of the TSD facilities indicated on
their Notification forms (Part A applicants are also required
to file Notification forms) that they were located on the site
of hazardous waste generators.
Equal probability sampling (within strata, see Appendix A)
from the generator population resulted in Generator Question-
naires being sent to a sample composed of 10,667 generators
without on-site TSD facilities, and 553 generators with on-site
TSD facilities, as would have been generally expected given the
distribution of the generator population as illustrated in
Figure 18.
Similarly, as would also have been expected based upon
the population overlap illustrated in Figure 18, most of the
2,599 facilities that received TSD General Questionnaires were
located on the sites of hazardous waste generators. Only 553
of these recipients, however, also received the Generator Ques-
tionnaire, since only 553 Generator Questionnaires were sent to
generators with on-site TSD facilities.
128
-------
Figure 18
OVERLAP AMONG POPULATIONS OF
GENERATORS AND TSD FACILITIES
REGULATED UNDER SUBTITLE C OF RCRA
i, r Ł,*>*•'* A * ^* •*>*„ v\
' t «*^*«-f •» ^^*>*"'v»'^ *>v-. «
^y"* //"* "* *^ *J** * *^L
l%" ^ T y *V* » *"*-»t^ -1 -I1*L*
GENERATORS
129
-------
Figure 19 illustrates the overlap that occurred between
the initial samples for the Generator and TSD General Questionnaires
and the overlap between the eligible respondents to those question-
naires (whose responses serve as the basis for all of the estimates
presented in this report).
The TSD General Questionnaire did not request facilities
to indicate the quantity of hazardous waste that was generated
at the site, since its purpose was to quantify the post-generation
management stages of the hazardous waste cycle. It did, however,
request facilities to report: (a) the total quantity of hazardous
waste managed (treated, stored, or disposed of) at the site
during 1981; and (b) quantities of hazardous wastes were received
from off site during 1981. Through a comparison of the responses
obtained from the 266 respondents that completed both the Generator
and TSD General Questionnaires (see Figure 19), it was determined
that subtracting quantities received from off site from total
quantities managed in the TSD General Questionnaire produced a
fairly reliable "proxy" for the quantity of hazardous waste
generated at the site during 1981. Among the 266 sites completing
both questionnaires, the proxy generation value was frequently
equivalent to the quantity generated as reported specifically
in the Generator Questionnaire. Furthermore, the sum of the
TSD General Questionnaire's proxy generation values across the
266 sites came close to equalling the sum of the actual generation
values reported in the Generator Questionnaire across the same
266 sites.
The proxy generation values were then computed for all
respondents to the TSD General Questionnaire. In so doing, the
survey obtained substantially greater numbers of "generator"
observations since, as noted above, most of the sampled TSD
facilities were located at the sites of hazardous waste generators.
Initially, these observations were used only to calculate a
130
-------
Figure 19
OVERLAP AMONG GENERATOR AND TSD SAMPLES
AND AMONG ELIGIBLE RESPONDENTS
QUESTIONNAIRE SAMPLES
ELIGIBLE RESPONDENTS
Generator Questionnaires
11,220
Generators without
on-jite TSD Facilities
10,667
TSD General
Questionnaires
2,559
TSD Facilities that
did not receive
Generator Questionnaires
Generator TSD General
Questionnaires Questionnaires
2,084 1,462
Generator
Questionnaire
Only 1266
UTS
TSD General
Questionnaire, but no
Generator Questionnaire
Generators with on-site
TSD Facilities that
received both
questionnaires
Both TSD General
and Generator
Questionnaires
131
-------
lower bound for the survey's preliminary estimate of the quantity
of hazardous waste generated. This lower bound represented the
minimum quantity of hazardous waste that was "known" to have
been generated by the respondents to the survey alone, and
approached the Generator Questionnaire's preliminary estimate
for the total quantity generated by the entire population of
14,098 hazardous waste generators. Nevertheless, the Generator
Questionnaire's preliminary estimate continued to be regarded
as the only valid statistical estimate, although likely under-
stated, of hazardous waste generation produced by the survey.
Subsequent to EPA's release of the preliminary findings,
however, an error was discovered in the computer program used
to calculate the lower bound. TSD facilities that indicated
that they did not receive any hazardous wastes from off site
during 1981 had been mistakenly omitted from the calculation of
the lower bound. Since these facilities did not receive any of
the hazardous wastes that they managed from off site, it was
clear that all of the quantities reported in their TSD General
Questionnaire responses had to have been generated on_sUe.
Once the computer program error was corrected, the recalculated
lower bound - the known minimum quantity of hazardous waste
generated during 1981 - exceeded the preliminary estimate for
the total quantity of hazardous waste generated during 1981,
necessitating the development of an alternate approach for
deriving a reliable estimate for the total quantity generated.
The approach used in this report for estimating the
total quantity of hazardous waste generated in 1981 was developed
using some of the generation data obtained from the Generator
Questionnaire and also using the TSD General Questionnaire's
proxy generation indicator. The approach is logically consistent
in its structure, takes advantage of additionally available
"generator" observations, and has produced an estimate that is
132
-------
subject to substantially reduced statistical uncertainty over
the preliminary estimate. Furthermore, the 71 billion gallon
estimate derived using this approach compares closely with the
estimate of the total quantity of hazardous waste managed in
processes regulated under RCRA by TSD facilities during 1981
(see Section 7.1).
Essentially, the population of hazardous waste generators
can be divided into two mutually exclusive populations:
• The population of generators that operate on-site
hazardous waste treatment, storage, and" disposal
facilities; and
• The population of generators that do not operace
on-site hazardous waste treatment, storage, a'ncT
disposal facilities.
As indicated earlier in this section, RCRA-regulated treatment,
storage, and disposal operations tend to be located more frequently
at the sites of larger hazardous waste generators as opposed to
smaller generators, primarily due to reasons related to economies
of scale. Conversely, the larger hazardous waste generators
tend to be included within ;he population of generators that
operate on-site TSD facilities. Furthermore, the implication
of the skewed size distribution found to characterize the population
of hazardous waste generators is that omission or inclusion of
a few of these large generators with on-site TSD facilities
in a probability sample can significantly affect any quantity
estimates derived therefrom,
The Generator Questionnaire, by not being targeted
specifically at this important segment of the population, incurred
excessive sampling error and related wide confidence intervals.
The sample of 2,084 generators responding to the Generator
Questionnaire contained only 266 generators that operated RCRA
133
-------
authorized TSD facilities on site. The TSD General Questionnaire,
however, was targeted at this important group of generators,
even though its intent was not to obtain information about
generation itself. Of the 1,462 respondent facilities, 1,370
including the 266 sites included in the generator sample, were
observed as generators of hazardous waste through the TSD General
Questionnaire's proxy generation indicator.
The estimate of the total quantity of hazardous waste
generated during 1981 is therefore best derived by adding together
the independently obtained estimates of the quantities of hazardous
waste generated by each of the two populations of generators.
The Generator Questionnaire provides an estimate of 5.4 billion
gallons of hazardous waste generated by the population of gener-
ators that do not operate RCRA authorized on-site TSD facilities.
The TSD General Questionnaire provides an estimate of 65.6
billion gallons of hazardous waste generated by the population
of generators that do operate on-site TSD facilities. Since
these two populations are mutually exclusive, and since together
they represent the entire population of hazardous waste generators
regulated under RCRA, the survey's estimate of the total quantity
of hazardous waste generated during 1981 can be obtained by
adding these two estimates together for a total of 71 billion
gallons generated, as indicated in the right hand bar (Final
Estimate) in Figure 20. The left hand bar (Preliminary Estimate)
is presented in Figure 20 for graphic comparison of the 42
billion and 71 billion gallon estimates, and to indicate how
each division of the generator population can also be quantified
exclusively using data obtained in the Generator Questionnaire.
Note that the segments for generators without TSD facilities
are identical in each estimate: both are derived from the
Generator Questionnaire. This difference between the estimates
rests entirely with the different generation estimates for
generators with TSD facilities on-site. All indicators suggest
that the TSD sample provides a better estimate for the quantity
generated by this group.
134
-------
Figure 20
to
z
o
110
100
90
80
70
2 60-
u_
O
CO
O 50-
40-
30-
20-
10-
0
95% C.I.
95% C.I.
(t 49.7%)
42 Billion gallons
(156 MMT)
71 Billion gallons
(264 MMT)
C.I. = Confidence interval
MMT = Million metric tonnes
L,"".J = From generators
without TSD facilities
(Generator Survey)
L^-^l = From generators
with TSD facilities
(Generator Survey)
- From generators
with TSD facilities
(TSD General Survey)
79.7%)
\
K
36.6 -
'V Billion
gallons .x
5.4 Bill
Billion
5.4
PRELIMINARY
ESTIMATE
FINAL
ESTIMATE
- Lower Bound -
44 Billion gallons
(164 MMT)
*Quantities generated were estimated by subtracting the amount recieved from
off site from the amount managed, since the TSD General Questionnaire did not include
a direct measure of generation.
135
-------
Figure 20 also presents the statistical confidence
interval at the 95 percent level. As indicated in Section 3.3,
this confidence interval is substantially narrower than the
interval associated with the preliminary generation estimate
obtained exclusively from the Generator Questionnaire. It is
important to note, however, that the survey's final estimate of
the total hazardous waste generation does fall within the confidence
interval surrounding the preliminary estimate, indicating that
the two estimates are not completely inconsistent.
Figure 20 also presents the recalculated "lower bound"
to the survey's 71 billion gallon estimate of 1981 hazardous
waste generation. The sum of the actual responses to the Generator
Questionnaire (except for the 266 sites that also completed the
TSD General Questionnaire) plus the sum of the proxy generation
values for all of the respondents to the TSD General Questionnaire
equals 44 billion gallons, and serves to reduce the range of
uncertainty at the lower end of the 71 billion gallon's confidence
interval at the 95 percent level.
The survey's 71 billion gallon (264 million metric
tonnes) hazardous waste generation estimate substantially exceeds
the 11 billion gallon (41 million metric tonne) estimate most
commonly referenced to date.1 Even the known lower limit to
the survey's estimate exceeds by fourfold the previously estimated
quantity of hazardous waste generated in the U.S. Strict
comparison of these estimates may not be entirely appropriate,
however, due to potential differences in the nature and types
of hazardous wastes included in each. In particular, the previous
estimate may not have included mixtures of hazardous and
Booz Allen, Supra, p. III-6.
136
-------
nonhazardous wastes in the same way that such mixtures are
included here. On the other hand, however, the 71 billion
gallon estimate specifically excludes, wherever they could be
identified, quantities of RCRA-regulated hazardous waste that
were generated in 1981, but that were not, or were not intended
to be, managed subsequently in processes regulated under RCRA
(see the section on Exemptions and Exclusions in the introduction
to Part III).
The major disadvantage of the alternate approach to
estimating the total quantity of hazardous waste generated is
that it could not be refined sufficiently to serve in a similar
capacity for the other generation estimates presented in this
section. An extensive attempt was made, for example, to develop
a similar proxy value to identify the quantities of hazardous
wastes that were shipped off site by TSD facilities during
1981. Respondents to the TSD General Questionnaire were not
requested to report such shipments, and efforts to derive a
proxy value did not meet with sufficient success when compared
with answers by the 266 respondents to the Generator Questionnaire
Thus, the estimates presented in Sections 6.2 through 6.6 continue
to be based exclusively upon data obtained through the Generator
Questionnaire, and their absolute values, that are the basis of
the percentages presented, are generally based upon the 42
billion gallon estimate derived from the Generator Questionnaire.
Accordingly, the estimates presented in the Sections
6.2 through 6.6 are presented in percentage form only. Implicit
in this approach is the assumption that the breakdowns of the
71 billion gallon estimate follow the same distribution as do
the breakdowns of the 42 billion gallon preliminary estimate.
This assumption may not be appropriate in all cases. The
remainder of the estimates presented in this section are,
therefore, subject to greater statistical uncertainties than
137
-------
are other quantity estimates presented elsewhere in the report,
and should be interpreted cautiously.
Finally, it is important to point out that, as indi-
cated in Table 3, extremely narrow confidence intervals were
obtained for nonquantity estimates derived from the Generator
Questionnaire, and those estimates are unaffected by the issues
affecting the quantity estimates. The survey sample was designed
for reasons discussed in Section 2.3, to produce accurate esti-
mates of population characteristics. All available evidence
suggests that it was highly successful in meeting that objective.
6•2 Size Distribution of Generators
As indicated in Section 3.3 and in Section 6.1, the
population of hazardous waste generators was found to be highly
skewed in terms of the quantities of hazardous waste generated
by individual generators. A few very large generators are
believed to account for ninety or more percent of the estimated
71 billion gallons of hazardous waste generated during 1981.
Figures 3 and 4 in Section i.3 graphically illustrate the simi-
larly skewed distribution of the population of hazardous waste
TSD facilities.
Graphic illustrations of the skewed distribution of
the generator population, based upon the Generator Questionnaire
look very similar to the figures presented for TSD facilities
As indicated in Section 6.1, however, the Generator Questionnaire
sample does not adequately represent the quantities of hazardous
waste generated by the larger generators that operate on-site
TSD facilities. Thus, graphic illustrations based exclusively
on the Generator Questionnaire sample may not accurately represent
the true size distribution of the generator population, and are
therefore not presented in this report.
138
-------
Nonetheless, it is clear, based upon the data obtained
through the Generator and TSD General Questionnaires, that the
size distribution of the population of hazardous waste generators
regulated under RCRA is highly skewed. Furthermore, the survey's
findings with respect to the highly skewed size distribution of
the generator population are consistent with other studies
conducted in individual states. The state of Massachusetts,
for example, in its 1981 compilation of state hazardous waste
reports, found an almost identical size distribution of generators
in Massachusetts during 198L.2
6. 3 Quantities Generated by Industry Type
Keeping in mind that a few large facilities in the
sample dominate the generator quantity estimates (see the previous
section) and that wide confidence intervals are associated with
such data (see Section 4.2), this section will cautiously present
estimates of quantities of hazardous waste generated by industry
type. The estimates presented in this section are, however,
based entirely upon the Generator Questionnaire, which under-
states the total quantity of hazardous waste estimated to have
been generated during 1981 (see Section 6.1). The impact of
that understatement is discussed below in relation to the
industry estimates themselves. Since these quantity estimates
are drawn from the Generator Questionnaire, however, the
statistical uncertainty surrounding them is likely to be large
(see Section 3.3). Their statistical uncertainty is increased
even further since they represent subsets of that questionnaire's
national estimates and are drawn from smaller sample sizes
Massachusetts Department of Environmental Management, Bureau of
Solid Waste Disposal, Hazardous Waste Management in Massachusetts,
Statewide Environmental Impact Report, August 1982, p~~14~.
139
-------
(requiring the application of the »K" factors) in determining
their confidence intervals, as discussed in Section 3.3 and
summarized in Table 5. For this reason it makes more sense to
discuss larger subsets of the data rather than individual four-
digit SIC codes that were collected in the survey.
Two industry groups that stand out are within the
manufacturing sector (SIC 20-39). Manufacturing as a whole
accounts for more than 90 percent of the total quantity of
hazardous waste generated. As shown in Figure 21 the chemical
and petroleum industries (SIC 28 and 29) alone account for more
than 70 percent of total generation. This is almost entirely
accounted for by SIC 28, the chemical industry. The average
quantity of hazardous waste generated by chemical industry
establishments (based on the Generator Survey) was almost four
times larger than the average for all generators. This explains
why the chemical industry, with only 17 percent of the generators,
generated 68 percent of all the hazardous waste generated in
1981. The petroleum refining industry (SIC 29) accounts for
about three percent of the hazardous waste generated in the
U.S.
Another prominent group in the manufacturing sector
was metal-related industries (SIC 33-37). The two largest
hazardous waste generating industries of this group were Machinery,
except Electrical (SIC 35), 10 percent; and Transportation
Equipment (SIC 37), six percent. Together with Primary Metals
(SIC 33), Fabricated Metal Products (SIC 34) and Electrical and
Electronics Machinery Equipment and Supplies (SIC 36), these
metal-related manufacturing industries generated approximately
22 percent of all hazardous waste generated in 1981.
140
-------
Figure 21
QUANTITIES OF HAZARDOUS WASTE GENERATED
IN 1981 BY INDUSTRY TYPE
Chemical & Petroleum
SIC 28-29
(71%)
Metal-Related
Industries
SIC 33-37
(22%)
0th0r
Industrie
141
-------
AH remaining manufacturing, nonmanufacturing, and
industries not specified by kind accounted for about seven
percent of U.S. hazardous waste generation. More than half of
this was generated by the Motor Freight Transportation and
Warehousing Industry (SIC 42).
The Generator Questionnaire understates quantities of
hazardous waste generated by generators with on-site TSD facilities
(see section 6.1), which tend to be larger industrial operations
To the extent that larger industrial operations with on-site
hazardous waste TSD facilities are concentrated in certain
industries, the quantities of hazardous waste generated by
those industries, and their proportions of total U.S. generation
are most likely understated by the Generator Questionnaire's
estimates. Thus, proportions instead of quantity estimates
are indicated in Figure 21.
As noted above, the average quantity generated by
chemical industry plants is four times larger than the average
for all other industries, suggesting a high concentration of
larger operations and the likelihood that the Generator Question-
naire understates the quantity of hazardous waste generated in
that industry. Furthermore, as indicated in Section 7.4 showing
quantities of hazardous waste managed during 1981, the chemical
and petroleum industries account for an even greater percentage
(85%) of the total quantity managed. In particular, the petroleum
inaustry, with only three percent of the generation total attributed
to it accounts for 19 percent of quantities managed. The comparisons
suggest that the quantity of hazardous waste generated by the
chemical and petroleum industries, and the proportion of total
U.S generation they account for, may be significantly understated
in the Generator Questionnaire's estimates and in Figure 21
142"
-------
6.4 Quantities of Hazardous Waste Generated by Type of
Waste Group
One of the breakdowns of generated hazardous waste
quantities most in demand is by waste group. Unfortunately, an
analysis of results from the Generator and TSD General Question-
naires indicates that the study did not yield a conclusive
estimate of the quantity of hazardous waste generated by waste
type.
Low credibility in the waste group quantity breakdowns
derived from the Generator Questionnaire was the result of many
factors. First, respondents were asked to report the waste
groups as percentages of their generation. Many respondents
only roughly approximated these percentages. Second, statistical
reliability of subpopulations of the quantity generated is
likely to be lower than the already low statistical reliability
of the Generator Questionnaire's estimate of the total quantity
of hazardous waste generated, especially since there are many
different waste groups, reducing the sample size from which the
estimates are based (see Table 5 and discussion of "K" factors
in Section 3.3). Third, the estimate of the quantity of hazardous
waste generated derived from the generator survey is understated,
thus understating the resulting quantities broken down by waste
groups.
Of particular concern to this third factor is that the
quantities of hazardous wastes generated by generators with
on-site TSD facilities are those that are understated the most.
Since these generators tend to be larger than those without TSD
facilities, they may also exhibit different industry and waste
group characteristics than the smaller facilities without their
own management facilities. The sampling approach and skewed
nature of the population combine to undermine the ability to
143
-------
accurately describe this important portion of the generator
population. TO the extent that these generators with on-site
TSD faculties do generate various types of waste in proportions
that differ from the proportions generated by other generators,
the percentage breakdowns by waste group derived from the Generator
Questionnaire may not accurately describe the "true" waste
group composition of the total U.S. hazardous waste stream.
This concern is reinforced by the fourth, and final
factor: the percentage breakdown by waste group for quantities
generated, based on the Generator Questionnaire, differs substan-
tially from the same breakdown for quantities managed based on
the TSD General Questionnaire, as indicated graphically in
Figure 22.
Although, for the reasons stated above, the Generator
Questionnaire cannot be used as the basis for a breakdown of
waste group quantities, the TSD General Survey also falls short
as a standard, for reasons explained in Section 75 As a
consequence, statements based on these surveys about quantities
of hazardous waste generated (or managed) by waste group cannot
be made with much precision.
6-5 ' ite
Ver_sus_0.tfSite
While the survey estimates that 84 percent of the
14,098 generators ship some or all of their hazardous wastes
off site (see Figure 9), the vast majority of the guantities of
hazardous waste are managed onsite. Figure 23 presents the
Generator Questionnaire's estimates of the proportions of generated
hazardous wastes that are subsequently managed at on-site TSD
facilities vs. the proportions shipped off site. Off site
144
-------
Figure 22
WASTE GROUP PERCENTAGE COMPARISONS BETWEEN
THE 1981 GENERATOR AND TSD GENERAL SURVEYS
_J
67
"F" or "K" Waste
Spent Solvents, Process
Sludges and Listed
Industry Wastes
"U" Waste Off-Specification or Discarded Commercial Chemical
1 Products and Manufacturing Intermediates
0.2
"P" Waste Acutely Hazardous Wastes
D001 Ignitable Wastes
46
D002 Corrosive Wastes
35
52
D003 Reactive Wastes
16
10
D004-D017 E.P Toxic Wastes
Unspecified (including State Regulated and
Self-defined Hazardous Wastes)
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
*36 of the 52% is represented by one generator with a highly dilute waste stream
170
KEY
Quantity of waste group handled as a percentage of total amount
of all waste groups managed' as per the TSD General Survey
Quantity of waste group generated as a percentage of total amount
generation, as per the Generator Survey
145
-------
Figure 23
nnAIT HA2ARDOUS WASTE GENERATED:
QUANTITY MANAGED ON SITE VERSUS
SHIPPED OFF SITE FOR MANAGEMENT
Managed on site
Shippeckpff site
Interfirm
shipments
Intrafirm captive
or unspecified shipments
146
-------
shipments are then further broken down between interfirm ship-
ments (i.e.f shipment by firms to facilities owned by other
firms) and intrafirm shipments (i.e., shipments by firms to
"captive" facilities that they own).
As indicated in the figure, data supplied by generators
indicates that approximately 96 percent of all generated hazardous
wastes are managed on site, with only four percent being shipped
off site for treatment, storage, and disposal.3 Interfirm
shipments, which are assumed generally to represent "commercial"
shipments, account for 82 percent of the total quantity shipped,
with shipments by firms to their own captive facilities making
up the remaining 18 percent of hazardous wastes shipped off
site.
Although the dominance of on site vs. off site manage-
ment of quantities of hazardous waste contrasts sharply with
the comparison of the number of generators managing on site and
the number shipping off site, the phenomenon^ is easily explained.
Large generators manage their wastes on site because, among
other factors, they experience economies of scale in constructing
and operating on-site TSD facilities and because the costs of
shipping and managing large quantities of hazardous waste off
site are prohibitive. Economies of scale encourage generators
of large quantities of hazardous waste to invest on site in
high-volume waste management technologies that result in low
unit costs for handling and disposal. Thus, one would expect
on-site management portion may be even larger than the 96
percent indicated from the Generator Survey since that survey
understates quantities generated by generators with on-site TSD
facilities (see Section 6.1). Generators with on-site TSD
facilities are less likely to ship their wastes off site. Were
they adequately represented in the generator sample, their on
site management practices would likesly increase the estimate of
the proportion managed on site and decrease the estimate of the
proportion shipped off site.
147
-------
to find greater quantities managed on site than shipped off
site, even though the greater numbers of generators ship their
hazardous wastes off site than manage them on site.
This finding is also supported by the understanding of
the distribution of hazardous waste generators across the size
(quantity generated) spectrum. As previously discussed in
Section 6.2, a few very large generators account for substantial
portions of the total quantity of hazardous waste generated
across the U.S., overwhelming the quantities of hazardous waste
produced by the more numerous, smaller generators. Since these
large generators are most apt, for reasons described above, to
nnanage their wastes on site, it is not surprising that such
quantities also overwhelm the quantities shipped off site for
treatment, storage, or disposal by the many smaller generators.
The final point to be highlighted with respect to the
information presented in Figure 23 relates to comparisons between
the results produced by this mail survey and estimates published
previously from other studies In order to make such comparisons,
however, the proportions presented in Figure 23 need to be
translated into estimates of the actual quantities of hazardous
wastes shipped off site and managed on site during 1981 As
noted at the end of Section b.l, quantity estimates have been
withheld in favor of the presentation of proportions throughout
this section because the Generator Questionnaire understates
the estimate of the total quantity of hazardous waste generated
during 1981. while the Generator Questionnaire produces an
estimate of 42 billion gallons, this report estimates that
there were actually 71 billion gallons of hazardous waste generated
m 1931 (see Section 6.1). Thus, quantity estimates from the
Generator Questionnaire could only be presented as proportions
since their actual values are inconsistent with the survey's 11
148
-------
billion gallon estimate. A decision was also made to not apply
the Generator Questionnaire's proportion estimates to the 71
billion gallon estimate throughout this section, due to uncer-
tainties about the representativeness of those proportions (the
Generator Questionnaire understates the quantities generated by
large generators with on-site TSD facilities).
For purposes of the comparisons to be made in this
section, however, the proportions derived from the Generator
Questionnaire (as presented in Figure 23) will be applied to the
71 billion gallon estimate to derive estimates of the actual
quantities needed to make such comparisons. Specifically, this
process produces an estimate of 2.8 billion gallons of hazardous
waste shipped off site during 1981, of which 2.3 billion gallons
were shipped commercially (i.e., interfirm shipments) and approxi
mately one half of a billion gallons were shipped by firms to
their own captive TSD facilities (i.e., intrafirm shipments).
While the survey's estimate of 71 billion gallons of
hazardous waste generated during 1981 is approximately six
times greater than the most frequently referenced previous
estimate (published in the Booz-Allen report),4 its 2.3 billion
gallon estimate of quantities shipped to commercial facilities
is actually very similar to the Booz-Allen estimate of 1.9
billion gallons (7.2 million metric tonnes), of hazardous waste
managed by commercial facilities.5 Furthermore, the difference
between these two estimates can be traced in part to the somewhat
broader range of facilities included in the survey's definition
4Booz-Allen, Supra, p. III-6: Exhibit III-4, col. 5 total.
5Booz-Allen, p. V-20.
149
-------
of commercial facilities (see Section 5.3). Purther accommoda-
tion of the difference between the two estimates is found in
the fact that the survey's 2.3 billion gallon estimate may
somewhat overstate the actual quantity of hazardous waste shipped
to commercial facilities, due to the fact that the proportion
estimate from which it is derived may itself be slightly over-
stated (see footnote 3).
The high comparability between these separate estimates
speaks well of each study's efforts to quantify commercial
hazardous waste activities. At the same time, however, the
substantial difference between their estimates of the total
quantity of hazardous waste generated reaffirms doubts concern-
ing the scope of most previous hazardous waste studies. In the
past, most attention has been paid to off site shipments and
particularly, to quantities of hazardous waste shipped to com-
mercial facilities. AS a result, good estimates appear to have
been generated by these studies concerning the characteristics
and magnitudes of this subset of the population, but few insights
were gained concerning similar data, particularly quantity
data, for on-site hazardous waste management and the population
as a whole. The major differences between the conclusions of
this and earlier studies appear to be related to the previously
unobserved activities of large hazardous waste generators with
on-site TSD facilities.
150
-------
6.6 Recycling of Quantities Generated by Location of Recycling
Facilities
Despite the relatively large proportion (over 40%) of
RCRA-regulated generators that indicated that they generated
hazardous wastes that were recycled during 1981, only four
percent of the hazardous wastes generated in 1981 were recycled.6
It should be noted, however, that generators of certain hazardous
wastes that recycle 100 percent of these wastes are not RCRA-
regulated. Thus, they are not included in the estimates cited
here. These results suggest that while many generators recycle,
they tend to recycle only small portions of their hazardous
waste streams. Over 80 percent of this recycling was done on
site. Less than one percent of the off site shipments went to
captive (owned by the generator) facilities; the remainder went
to commercial recyclers. The average amount sent off site for
recycling was substantially smaller than the average amount
recycled on site. Figure 24 displays graphically the proportions
of RCRA-regulated hazardous waste that were recycled in 1981.
Direct comparisons of the survey's estimates of quantities
of hazardous wastes recycled with total quantities generated
may be somewhat misleading. As stated in the introduction to
Part III of this report, significant portions of the quantities
included in the generation estimate are mixtures of hazardous
wastes with industrial process waters (the resultant mixtures
are hazardous wastes). Such mixtures are likely to reduce the
6This proportion may be understated or overstated. The proportion
estimate is based upon the Generator Questionnaire, which under-
states the quantities of hazardous waste generated by generators
with on-site TSD facilities. To the extent that these generators
recycle greater proportions of their hazardous wastes than does
the generator population as a whole, the proportion is understated.
To the extent that these generators recycle smaller proportions
of their hazardous waste, the proportion is overstated.
151
-------
Figure 24
HAZARDOUS WASTES RECYCLED IN 1981
(% on site vs. off site)
Hazardous Wastes
Not Recycled
96%
Recycled
4%
152
-------
feasibility of recycling, since the materials sought in the
recycling process — the hazardous constituents -- represent
only a small portion of the total volumes. A more meaningful
comparison would be between quantities recycled and quantities
"recyclable" (i.e., the quantities of the hazardous constituents
themselves). The survey data do not, however, disaggregate the
quantities in such mixtures, preventing presentation of such
comparisons in this report.
6. 7 Future Hazardous Waste Generation Regulated Under RCRA
The survey was designed to establish estimates of the
RCRA-regulated quantities of hazardous waste generated during
1981 and the quantities of hazardous waste managed in processes
regulated under RCRA during 1981. Although quantitative data
relating to anticipated future hazardous waste generation rates
were not obtained through the survey, it is appropriate nonethe-
less to conclude this section of the report with a discussion
of the factors that are likely to contribute to increases and/or
decreases in the annual quantities of regulated hazardous waste
generated in future years.
The quantity of RCRA-regulated hazardous wastes generated
at any point in time is essentially affected by two factors:
• The specific nature and scope of the RCRA hazardous
waste regulatory program at that time; and
• The nature of the industrial and other activities
that actually result in hazardous by-products.
153
-------
C°n9ress is considering a number of statutory
changes to RCRA. Some of these changes, upon enactment, would
have a considerable effect on the range of activities covered
under the RCRA regulatory program and, therefore, upon the
quantities of RCRA-regulated hazardous waste generated in the
future. important among the statutory changes under consideration
are proposals to reduce the "small quantity generator" definition
for generators of hazardous waste, while firm estimates of the
numbers of small quantity generators that would be affected by
such a change are not yet available (see Section 9, Future
Studies), it is assumed that thousands of currently exempt
firms would be brought into the RCRA-regulated community under
these provisions. Furthermore, while information is also not
yet available to estimate the additional quantities of hazardous
waste that would be regulated under RCRA as a result of such a
provision, it seems clear that an increase in the size of the
RCRA-regulated population will serve to increase the quantities
of regulated hazardous waste generated in the future.
Other changes currently under consideration by the
Congress and at EPA involve the removal of exemptions for wastes
burned as fuels, and removal of certain exemptions covering
recycling activities. All of these changes to the RCRA regulatory
program would be expected to result in greater quantities of
wastes regulated under RCRA in future year.
Implementation of industrial pretreatment standards by
EPA and States may also affect future hazardous waste generation
Currently, an unestimable quantity of otherwise hazardous wastes
are exempted from regulation under RCRA because they pass legally
through public sewer systems into Publicly Owned Treatment
Works (POTW-s). While the quantities of such wastes cannot be
estimated by this survey, information from other sources suggests
154
-------
that such quantities are indeed large. The Massachusetts Bureau
of Solid Waste Disposal, for example, estimates that the quantity
of hazardous waste generated annually in Massachusetts could
increase by more than 50 percent with the implementation of
pretreatment standards and the installation of industrial
pretreatment processes by currently exempt generators.7 Similar
experiences in other States would add considerably to the
quantities of hazardous waste regulated under RCRA in future
years.
Increased industrial output, prompted by improved
economic conditions, could also serve to increase hazardous
waste generation in future years. The 1981 national economy
was in an economic slump, resulting in reduced industrial output.
The survey was not able to determine the relationship between
changes in the levels of industrial production and changes in
the quantities of hazardous waste generation. However, it is
generally believed that there exists a direct relationship
between these two factors. Thus, hazardous waste generation
would be expected to increase during periods of improved
economic conditions.
EPA is currently proposing the listing of additional
wastes as hazardous wastes. To the extent that such listings,
now and in the future, are actually promulgated, the quantities
of hazardous waste regulated under RCRA are expected to increase
accordingly.
Hazardous Waste Management in Massachusetts, pp 7-13
155
-------
Finally, increased clean-up of abandoned or closed
hazardous waste sites will generate additional quantities of
hazardous waste that require proper disposal under RCRA. The
survey observed some such quantities being generated during the
1981 calendar year. Implementation of the "Superfund" program
by EPA and States has advanced since 1981 and is expected to
continue to advance in coming years, with resulting increases
in the quantities of hazardous waste to be managed.
While the factors stated above are among the factors
that are expected to result in increases in hazardous waste
generation in future years, a number of other factors are
expected to contribute to decreases in future hazardous waste
generation. Prominent among these are proposals in Congress
and some States to adopt specific taxes on the generation and/or
disposal of hazardous waste. These taxes are often intended,
among other things, to provide economic disincentives toward'
the actual generation of hazardous waste. These and other
existing or future economic factors are expected to encourage
industries to engage in greater "source reduction" and "source
separation" to eliminate or reduce their generation of hazardous
waste.
Further regulatory actions by the Congress and EPA
that may encourage or require reductions in hazardous waste
generation include proposed bans on the land disposal and
underground injection of certain hazardous wastes. As disposal
options for hazardous waste become more limited and more
costly, generators will be encouraged to reduce their hazardous
waste generation.
156
-------
Finally, one of the purposes behind the enactment of
the Resource Conservation and Recovery Act was to encourage the
conservation of natural resources through the use, reuse,
recycling, and reclamation of materials contained in industrial
and other waste streams. As the value of natural resources
increases, economic incentives for such recycling activities
are expected to increase, resulting in increased efforts by
generators to turn their waste streams into useful, valuable
commodities.
It is difficult, if not impossible, to draw any firm
conclusions concerning future hazardous waste generation rates
based upon data available from the survey. The interactions of
the factors described above, together with countless other
factors not mentioned or even currently anticipated, are difficult
to predict. Furthermore, the uncertainties surrounding this
survey's 1981 generation estimates will make the measurement of
any short-term changes in hazardous waste generation rates even
more difficult to evaluate.
•157
-------
-------
7. QUANTITIES OF HAZARDOUS WASTE MANAGED
The quantity estimates and breakdowns presented in
this section pertain to a very specific population of hazardous
waste management (TSD) facilities and hazardous waste management
activities. The types of facilities and processes specifically
included in the survey were defined in Section 5, notably includ-
ing only those TSD facilities and those treatment, storage, and
disposal processes that were subject to regulation under RCRA
and that actively processed hazardous waste during 1981 (as
opposed, for example, to having maintained a hazardous waste
landfill without actually having entered new quantities of
hazardous waste therein during 1981).
The introduction to this part of the report (Part III:
Sections 6 and 7), describing specifically the impact of RCRA's
complex rules and exemptions on the nature of these quantity
estimates, should be read carefully in conjunction with any
interpretation or use of the estimates presented in this section.
The quantity estimates presented in this section were
developed from questions included in the Treatment, Storage, and
Disposal General Questionnaire, and six of the eight technology-
specific management questionnaires. Estimates of total quantities
of hazardous waste managed, total quantities treated, stored,
and disposed of at facilities, and estimates of quantities managed
in waste piles, incinerators, and miscellaneous processes (not
specified by kind, or not elsewhere classified), were estimated
from the TSD General Questionnaire results, as were other facility-
wide estimates. Estimates of the quantities of hazardous wastes
managed in tanks, surface impoundments, containers, injection
wells, landfills, and land application were based on the tech-
nology (component) questionnaires. All of the estimates presented
pertain exclusively to 1981 calendar year operations.
159
-------
The latter stages of the hazardous waste management
cycle, involving its treatment, storage, and eventual disposal,
do not lend themselves to quantification in the same manner as
the initial, or generation stage does. Whereas the "point of
generation" (where measurements of hazardous waste quantities
generated should be obtained) can be conceptualized and identi-
fied, a single corresponding "point of management" for such
quantities frequently cannot be decided upon. Rather, generated
hazardous wastes are frequently managed in multiple stages:
initially stored after generation, then treated and/or disposed
of. Accordingly, the decision as to the point at which the
measurement, or quantification, of the management stage should
take place is not as clear cut as in the case of generation.
In response to this problem, the TSD facility survey
questionnaires obtained quantity information at various points
in the management stages of the hazardous waste cycle. Initially,
respondents were asked to identify the total quantity of hazardous
wastes managed during 1981. Specifically, this question asked
respondents to report the total quantity of hazardous waste that
passed "through the front door" of the waste management facility
(even if the facility was located on the site of generation),
counting only once those quantities of hazardous waste that were
managed in multiple stages. Thus, wastes that were initially
stored or treated prior to disposal were to be counted at the
moment that they entered such storage or treatment processes,
and not to be counted a second, third, or fourth time upon their
removal from such storage or treatment and their advancement to
subsequent or final stages of processing. For generators that
managed all generated hazardous wastes on site during 1981, it
was expected that the quantity reported in response to this
160
-------
question would be equal to, or would approximate, the quantity
of hazardous waste generated at that site during 1981.1
After obtaining in this manner the total quantity of
hazardous waste managed during the calendar year, the question-
naires then went on to ask for quantity estimates corresponding
to each individual stage of the hazardous waste management pro-
cess, specifically calling for the repeated counting of quanti-
ties that were managed in multiple stages. Separate quantities
were requested for hazardous wastes that were entered into storage,
entered into treatment processes, and actually disposed of at
the facility during 1981, without regard to any other management
processes applied previously or subsequently to specific waste
streams included in these quantities. It was expected, therefore,
that the sum of these quantities (treatment plus storage plus
disposal) would exceed the quantity of hazardous waste generated
at a facility wherever wastes were managed in multiple stages or
through multiple processes.
The estimates presented in the following sections
should, therefore, be interpreted carefully. Subtle differences
are built into each specific estimate, as each was developed to
meet rather precisely defined information needs. Furthermore,
due to the complex nature of the questions asked, it was not
always possible to identify respondent errors in the obtained
data, particularly with respect to the double counting or nondouble
counting of waste quantities. While many such errors were detected
and corrected in processing the survey responses, an unestimable
number of respondent errors are believed to remain in the obtained
TSD facility quantity data, despite the fact that extensive
Subsequent analysis of responses from generators that completed
both the Generator Questionnaire and the TSD General Question-
naire found a high correlation between total quantities reported
generated and total quantities managed. Exceptions to this
relationship were observed, however, and generally resulted in
cases where generators managed hazardous waste in processes
excluded from regulation under RCRA.
161
-------
internal logic and consistency edit checks were designed into
the questionnaires (see the survey field report contained in
Appendix B). In the final analysis, many aspects of the manage-
ment stage of the hazardous waste cycle lack the conceptual
clarity and precision required to develop "hard" answers to
questions that would appear, at first glance, to be simple and
straightforward.
Estimates of hazardous waste quantities managed have
been developed for the entire TSD facility population and for
each of the subpopulations of facilities and specific waste
management processes whose numbers and types were detailed in
Sections 5 through 5.5. These survey estimates are presented in
the following Sections 7.1 through 7.7. An analysis of hazardous
waste treatment, storage, and disposal capacity estimates devel-
oped through the survey can be found in Section 8.
7.1 Quantity of Waste Managed as Hazardous by TSD
Facilities in 1981
Based on survey results, the estimated quantity of
waste managed as hazardous waste under RCRA in the United States
in 1981 was 71.3 billion gallons (265 million metric tonnes).
As in the case of generators, there exists a wide variation in
the size of TSD facilities, with some facilities managing huge
quantities of waste as hazardous waste (confirmed by independent
EPA callbacks to Westat's largest respondents). However, as
indicated in Section 6.1, since by definition it was targeted
exclusively on sites with TSD facilities (most of which were
also sites of hazardous waste generators), the TSD facility
sample included substantially greater numbers and proportions of
the important (in terms of the precision of derived quantity
estimates) large quantity managers (and generators) than did the
162
-------
sample for the Generator Questionnaire. As indicated in Table 3,
the 95 percent level statistical confidence interval surrounding
the primary quantity estimate derived from the Generator Question-
naire was 63.3 percent wider than the confidence interval
surrounding the 71.3 billion gallon estimate of total quantities
managed as hazardous waste during 1981 derived from the TSD
General Questionnaire.
Accordingly, as indicated in Table 3, the 95 percent
statistical confidence interval surrounding the survey's estimate
of the total quantity of wastes managed as hazardous wastes
during 1981 is +48.8 percent. As illustrated in Figure 25, this
interval ranges +36 billion gallons on either side of the 71.3
billion gallon estimate. However, as explained in Section 6.1,
the lower end of this interval can be "bounded," thereby reducing
some of the uncertainty about this estimate. A summation of the
unweighted responses to the TSD General Questionnaire (adding
together the actual answers provided by the 1,462 respondent
facilities, without applying the statistical weighting factors
used to project from the sample to the universe) produces a
lower bound of 45.3 billion gallons, and is indicated by the
arrow in Figure 25.
The survey's final estimate of 71.3 billion gallons of
waste managed as hazardous waste during 1981 is somewhat reduced
from the preliminary estimate of 80 billion gallons released by
EPA on August 30, 1983. The reduction is due entirely to further
editing of the data obtained through the TSD facility survey
subsequent to the release of the preliminary findings. As
indicated in the introduction to Part III of this report, the
definition of hazardous wastes intended to be included in the
national survey is fairly precise, requiring sophisticated
evaluations of the responses to the survey questionnaires in
order to identify and remove quantities of waste not intended to
163
-------
Figure 25
QUANTITY OF WASTE MANAGED AS
HAZARDOUS WASTES BY TSD FACILITIES IN 1981
120
- 100
71.8 Billion
gallons
- 80
- 70
- 60
^- 40
- 20
Confidence
Intervals
95%
71.3 Billion gallons=265 Million metric tonnes
* Known lower limit on quantity managed
164
-------
be included under the survey's definitions. Most prominent
among such wastes, as noted earlier, are those managed exclusively
in processes not subject to regulation under RCRA. Subsequent
to the August 30 release of the preliminary findings, additional
methods were discovered for identifying such quantities, among
others, in the obtained data. Application of those methods, in
the form of a second round of computer edit and internal consistency
checks, resulted in the reduction of the survey's management
estimate from the preliminary figure of 80 billion gallons to
the final estimate of 71.3 billion gallons presented in this
report.
In addition to the statistical uncertainty ranges
presented above, there are a number of nonsampling error sources
whose effect and magnitude cannot be precisely estimated. These
include: respondent error in reporting quantity; misunderstanding
of RCRA-covered waste (especially wastewater, 90-day storage and
mixture rule exclusions); possible respondent'misunderstanding
of storage and treatment as aggregate to date rather than 1981
additions (i.e., inclusion of stocks on hand carried over from
previous years); and processing errors, (e.g., unit conversions,
coding, keypunch). Many of these factors will be the subject of
further study, especially the wastewater exclusions. The failure
of some respondents to exclude exempted treatment tank wastewater
in the TSD General Survey combined with an inability to identify
such cases systematically, has resulted in an estimated over-
statement of roughly 5 billion gallons in the total quantities
treated. Such misreporting also affects the estimate of total
quantities managed and all breakdowns of that quantity that are
presented in Section 7.1 through 7.7.
165
-------
7.2 Size Distribution of TSD Facilities
As indicated in Section 3.3, the distribution of quantity
of hazardous waste managed is quite skewed. A few extremely
large facilities dominate the measures from all others. The two
largest sites sampled contribute 25 percent of the estimated
total of 71 million gallons in the country. The sites represent-
ing the largest one percent in the country provide approximately
72 percent of the estimated national total. The largest two
percent provide 87 percent of the total. Figure 26 illustrates
the size distribution for management facilities in the country.
For future survey efforts desiring to obtain more
precise estimates on total quantities managed nationally, prior
identification and selection of the largest management facilities
will improve the precision of quantity estimates. However, the
identification of an appropriate, reliable size measure applicable
to all or most facilities in the universe may require some
investigation due to the fact that different size measures have
greater applicability to some processes than to others (see
Section 2.3).
7.3 Comparison of Quantities of Hazardous Waste Generated
and Managed
As stated in the introduction to Part III of this
report, the national survey was designed to produce comparable
estimates of the 1981 hazardous waste generation and its subsequent
management (treatment, storage, and disposal). In designing the
survey along these lines, OSW and Westat intended to observe
essentially the same national 1981 hazardous waste stream in
both the generation and management surveys; that is, it was
166
-------
Figure 26
COMPARISON OF CUMULATIVE DISTRIBUTIONS OF QUANTITY OF HAZARDOUS WASTE
MANAGED IN 1981 AND THE NUMBER OF TSD FACILITIES
100
80 H
70
ct un SO
UJ <
> 3
« § 40
~r
50
~T
80
r
90
20 JO 40 50 60 70
CUMULATIVE PERCENTAGE OF TSD MANAGEMENT FACILITIES IN US
(FROM LARGEST TO SMALLEST FACILITY)
)00
167
-------
intended that the TSD facility survey would account for and
describe the management of all, or most, of the hazardous waste
estimated to have been generated in the generator survey.
In the final analysis, the survey design can be seen
as largely successful in meeting this objective: as indicated
in Figure 27, the survey's estimate of the total quantity of
hazardous waste generated is only slightly less than its estimate
of the total quantity of wastes managed as hazardous waste during
1981. At the time that preliminary results of the survey were
released, however, the difference between these two estimates
was fairly substantial, due primarily to the fact that the
preliminary estimate of the total quantity of hazardous waste
generated in 1981 was 40 billion gallons (as opposed to the
final estimate of 71 billion gallons used in this report).
Section 6.1 discusses in detail the reasons why the preliminary
estimate, derived solely from the Generator Questionnaire, under-
stated the actual quantity of hazardous waste that was generated
during 1981.
The final 71 billion gallon estimate for 1981 hazardous
waste generation, used in this report, was derived by adding
together the quantity of hazardous waste estimated through the
Generator Questionnaire to have been generated by generators
that do not operate on-site TSD facilities and the quantity
estimated, through a proxy generation indicator in the TSD General
Questionnaire, to have been generated by generators with on-site
TSD facilities. As detailed in Section 6.1, a proxy generation
indicator was developed from the TSD General Questionnaire by
subtracting the quantity each facility reported as having been
received from off site (i.e., not generated on site) from the
total quantity each reported managing during 1981. The logic
supporting this calculation is that quantities of hazardous
waste that were managed at a site and that were not received
from off site must have been generated on site.
168
-------
Figure 27
QUANTITY GENERATED VERSUS MANAGED
30-1
70
60~
50
40 -
30-
20-
QUANTITY
GENERATED:
71 Billion
Gallons
QUANTITY
MANAGED:
71.3 Billion
Gallons
5.4
<] 65.6* [>
Quantity generated and managed on site
by generators with on-site TSD facilities
Quantity managed by TSD facilities
that was received from off site
Quantity generated by generators
that do not operate on-site TSD facilities
* Estimated from TSD General Questionnaire by subtracting
quantities received from off site from total quantities
of hazardous waste managed
169
-------
Figure 27 presents the survey's estimates of the total
quantity of hazardous waste generated during 1981 and the total
quantity managed as hazardous waste in that year. Each estimate
is broken into its component parts: for generation, quantities
generated by generators (a) with and (b) without on-site TSD
facilities; for management, quantities managed that were (a)
generated on site and (b) received from off site. The figure
illustrates that the major portion of each estimate comes from
the same source: the TSD General Questionnaire-derived proxy
for quantities generated and managed on site by generators with
on-site TSD facilities. With such a large percentage (approximately
98%) of each estimate being derived from the same source, it is
not surprising that the resultant total estimates for quantities
generated and managed are similar.
The correlation between the estimates of quantities of
hazardous waste generated by generators without on-site TSD
facilities and quantities received from off site by TSD facilities
deserves some discussion. On the surface, these two quantity
estimates would be expected to be similar: generators without
on-site TSD facilities must be shipping all of their generated
hazardous wastes off site to RCRA-regulated TSD facilities; such
quantities would therefore be expected to show up in reports by
TSD facilities of quantities of hazardous waste received from
off site. A number of factors revealed through closer examination
of those estimates, however, reveal that these numbers should
not, in fact, correlate as closely as they do.
To begin with, not all of the quantities that RCRA-
regulated TSD facilities receive from off site originate from
RCRA-regulated hazardous waste generators. Specifically, 2.9
billion of the 5.7 billion gallons received by facilities from
off site was indicated by them to have originated from small
quantity generators that are not subject to regulation under
170
-------
RCRA (see Section 4). Even though quantities received from
small quantity generators are not, technically, RCRA-regulated
hazardous wastes, they are included in the survey's estimates
under the term "wastes managed as hazardous wastes" (see intro-
duction to Section 7). Thus, the TSD facilities actually reported
receiving only 2.8 billion gallons (5.7 billion - 2.9 billion =
2.8 billion gallons) of hazardous waste from RCRA-regulated
generators. This 2.8 billion gallon figure, derived from the
TSD General Questionnaire, correlates exactly with the quantity
estimated in Section 6.5 to have been shipped off site by RCRA-
regulated generators.
Figure 28 presents the estimates of quantities generated
by generators without TSD facilities and quantities received
from off site by TSD facilities in greater scale and more detail.
The right hand bar in the figure illustrates the breakdown of
the quantities received by TSD facilities between those originating
from RCRA-regulated generators (2.8 billion) and those received
from generators not regulated under RCRA (i.e., small quantity
generators). The 2.8 billion gallons is then broken down further
between estimated quantities received from generators with TSD
facilities and generators without TSD facilities. This further
breakout is based upon the estimate from the Generator Question-
naire of the quantity shipped off site by generators without TSD
facilities (800 million gallons). The remaining 2 billion gallons
that facilities reported receiving (2.8 billion - 800 million =
2.0 billion gallons) from RCRA-regulated generators is assumed
to originate from generators with on-site TSD facilities who,
nonetheless, still ship some hazardous wastes off site. Many of
the generators that do operate on-site TSD facilities also ship
hazardous wastes off site to other TSD facilities. Many of the
TSD facilities that are located on the site of generators are
actually only storage facilities (see Section 5.4). These
generators store their wastes for more than 90 days, and as such
171
-------
Figure 28
9 -
8 -
7 -
5 -
3 -
2 -
4.6
Quantity Generated
by Generators
Without TSD
Facilities
5.4 Billion
Gallons
Quantity Received
By TSD Facilities
5.7 Billion
Gallons
2.8
Managed at sites that
did not file Part A
permit applications
Measurement error
In 90 day storage
Shipped off site
Received from small
quantity generators
Received from generators
with on-site TSD facilities
Received from generators
without TSD facilities
172
-------
appear as TSD facilities, then ship them off site for treatment
or disposal. Additional generators with on-site treatment or
disposal facilities also frequently ship residues from those
processes to other facilities for final disposition.
The left hand bar in Figure 28 breaks out the estimated
quantity of hazardous waste generated by generators without TSD
facilities. The dotted line connecting the two bars indicates
that the 800 million gallons shipped off site by these generators
is assumed to account for 800 million gallons of the 2.8 billion
gallons estimated to have been received by TSD facilities. This
is only an assumption, however, since data collected from facilities
are not broken to this degree.
Logic suggests that generators without RCRA-authorized
TSD facilities must ship all of their generated hazardous wastes
off site. As indicated above, however, this group of generators
only reported shipping 800 million gallons of hazardous waste
off site during 1981. The final disposition of the remaining
4.6 billion gallons of hazardous waste generated by generators
without TSD facilities cannot be determined from the survey
results. Some portions of that quantity can be explained, however,
and generally quantified.
To begin with, not all of the 4.6 billion gallons
actually had to be shipped off site during 1981. The RCRA regu-
lations allow storage of hazardous wastes in containers and
tanks on site for up to 90 days without submission of Part A
applications. Thus, any wastes generated during the last 90
days of 1981 need not have been shipped off site by the end of
1981, and therefore would not be included in the generator's
reports of 1981 off site shipments. Assuming equal generation
rates in each quarter of the calendar year, up to 25 percent of
the 4.6 billion gallons (1.2 billion gallons) generated in 1981
173
-------
could legitimately have not been shipped off site by the end ot
1981 by generators that did not file Part A permit applications.
Two additional factors account for the remaining 3.4
billion gallons (4.6 billion - 1.2 billion = 3.4 billion gallons)
generated by this group that would be expected to have been
shipped off site during 1981. The first can generally be classified
as measurement error in the survey data. Some of the quantities
reported as generated by these generators may actually have been
treated in exempt processes (for which Part A permit applications
need not have been submitted) or shipped to POTW's, and would
not have been included in the survey's estimates for quantities
generated could they have been identified as such (see introduction
to Part III). Inclusion of such quantities in the final survey
data contributes to the uncertainty surrounding its estimates of
total hazardous waste generation and management in the form of
nonsampling error (see Sections 3.4 and 6.1).
Data supplied through the survey, however, suggests
that some of these generators may actually be treating, storing,
or disposing of hazardous wastes on site, even though they have
not submitted Part A applications covering such activities as
required under RCRA. The apparent on-site management practices
of these generators would account for most of the remaining
differences between the 5.4 billion gallons generated and the
800 million gallons shipped off site by the generators without
RCRA-authorized (i.e., for which Part A permit applications have
been submitted) TSD facilities. Information concerning these
on-site management practices has been forwarded to EPA's enforce-
ment offices for further investigation.
Thus, while the estimates of hazardous waste generation
by generators without TSD facilities and quantities received
from off site by TSD facilities appear on the surface to correlate
174
-------
very well, that correlation is apparently misleading. Nonetheless,
the correlations that should be found in the data collected
independently from generators and facilities does appear to hold
in the survey's findings. The estimated quantity of hazardous
waste shipped off site by RCRA-regulated generators (2.8 billion
gallons) is estimated to equal almost exactly the quantity
estimated to have been received by TSD facilities from RCRA-
regulated generators.
7.4 Quantities Managed by Industry Type
The survey data indicate that close to two-thirds
(66%) of all hazardous waste management in 1981 was undertaken
by the chemical industry, while only a little over 25 percent of
the management facilities were associated with the chemical
industry. The petroleum refining industry accounts for 19 percent
of the hazardous waste managed with only four percent of the
management facilities. Thus, the two largest industries (chemical
and petroleum refining) have roughly 30 percent of the management
facilities, but account for 85 percent of all hazardous waste
managed. Clearly, management is heavily concentrated in these
two industries with a few very large management facilities found
within each. Metal-related industries as a group (SIC 33-37)
accounted for about seven percent of the hazardous waste managed
in 1981.
The only nonmanufacturing industry among the top nine
in terms of quantities of hazardous waste managed was SIC 49,
the electric, gas, and sanitary services. This sector accounts
175
-------
for approximately three percent of the management facilities and
also three percent of the hazardous waste managed. This industry
includes SIC Code 4953, refuse systems.2
Figure 29 presents the distribution of quantities of
hazardous waste managed for major categories of industry (as
determined by primary SIC code).
7. 5 Quantities Managed by Type of Waste Group
As previously stated in Section 6.4, survey-derived
data on hazardous waste quantities broken down by waste group,
although in great demand, are inconclusive. This conclusion is
based, in part, on the previously discussed (see Section 6.4)
disagreement between the survey's two primary sources of waste-
specific quantity data (the Generator Questionnaire and the TSD
General Questionnaire). Some of the other reasons why the survey
results describing waste group quantities are inconclusive are
enumerated in Section 6.4, with respect to quantitites of waste
types generated. Other reasons for why the survey's waste group
quantity estimates are inconclusive are presented below in relation
to quantities of waste types managed by TSD facilities. This
section will present some very rough estimates of quantities
managed by type of waste group and describe the limitations of
these data.
The TSD General Questionnaire allows a more detailed
breakdown of waste groups than does the Generator Questionnaire.
Whereas waste quantities are grouped into broad categories in
2
-'Many commercial refuse systems may not have reported under SIC
49 because it was not theii" primary business, or because they
failed to report an SIC code. "Commercial," as defined in this
report, takes in much more than SIC 4953.
176
-------
Figure 29
QUANTITIES OF HAZARDOUS WASTE MANAGED
IN 1981 BY INDUSTRY TYPE
Chemical & Petroleum
Industries
SIC 28-29
(85%)
Related
Industries
Other
Industries
SIC 33-37
(7%)
177
-------
the Generator Questionnaire (e.g., "F" and "K" wastes are lumped
together), quantities managed for individual waste codes can be
retrieved from the TSD General Questionnaire. One of the major
questions included in the TSD Questionnaire asked respondents to
describe the flow of specific waste streams (designated by EPA
waste codes) through the various treatment, storage, and disposal
processes used in hazardous waste management at their facility.
Up to the ten largest wastes could be reported in the question
(a limit imposed to reduce respondent burden). The question
asked respondents to report the total quantity of each waste
stream "handled" at their facility, followed by the specific
quantities of each stream that were treated, stored, and/or
disposed of in the management cycle.
One of the major problems, however, in the design of
the above mentioned question is that it used the term "handled,"
instead of "managed," in asking for the total quantity of each
waste stream that was treated, stored, and/or disposed of at the
facility in processes regulated under RCRA. (See the introduction
to Part III for definitions of waste streams intended to be
included in the survey). Some respondents, however, are believed
to have interpreted the term "handled" more broadly than intended
in the design of the question. For example, in addition to the
desired "managed" quantities, some respondents included in their
reports of quantities "handled:" (a) quantities of hazardous
waste generated or transported that were not managed on site;
(b) quantities included in stocks carried over from generation
in previous years; or (c) quantities handled exclusively in
RCRA-exempt processes, particularly in the case of aqueous
corrosive waste streams that may have been treated exclusively
in RCRA-exempt wastewater treatment tanks (see introduction to
Part III). The questionnaire editing process attempted to identify
and remove quantities that were reported under this broader
interpretation (i.e., quantities confirmed to have been handled
178
-------
exclusively in exempt wastewater treatment tanks were subtracted
from respondents reported quantities). However, due to imprecise
question wording and the complicated nature of the editing process,
some such undesired quantities are believed to remain in the
final survey data. While the amount of such undesired quantities
is believed to be small, their specific quantity cannot be deter-
mined. Furthermore, while their amount may be small relative to
the 71.3 billion gallons estimated to have been managed in 1981,
such undesired quantities could significantly affect estimated
quantities for individual waste groups. As a result of the
occurrence of these undesired quantities, the estimate of the
total quantity "handled" for a given waste stream is believed to
somewhat overstate the desired (but, as a result, unknown)
estimate of the total quantity "managed" for that stream.
Therefore, Figure 30 presents two estimates, derived
from the TSD General Questionnaire, for each waste group. The
estimated total quantity "handled" is expressed for each waste
group at the right of each bar, in billions of gallons. As
noted above, however, this number tends to overstate the amount
managed. The lower estimates, identified by the dark shaded
portion of each bar, are intended to estimate the minimum
quantity managed for each waste group, and are the result of
choosing the largest of the individual quantities treated, stored,
or disposed for each waste stream reported. These minimum
estimates clearly understate the amount managed (ignoring
sampling and nonsampling error for the moment) since treatment,
storage, and disposal operations do not completely overlap each
other. Part of a corrosive waste stream, for example, may be
deep well injected, while another part is chemically treated.
The lower estimates can be quantified by referring to the scale
at the bottom of the graph, also expressed in billions of gallons.
179
-------
Figure 30
QUANTITIES OF HAZARDOUS WASTE HANDLED
BY MANAGEMENT FACILITIES IN 1981
BY TYPE OF WASTE GROUP
(billions of gallons)
3.2 F001-F005 Spent Halogenated and Nonhalogenated Solvents
2.6 F006-F019
Electroplating and Coating Wastewater Treatment
Sludges and Cyanide-bearing Bath Solutions and
Sludges
Listed lndustry Wastes
Specific Sources
2.9 U001-U247 Off-Specification or Discarded Commercial Chemical
Products and Manufacturing Intermediates
.2 P001-P123 Acutely Hazardous Wastes
1.4 D001 Ignitable Wastes
D002 Corrosive Wastes
3 2 D003 Reactive Wastes
11 1 D004-D017 E.P. Toxic Wastes
7 Unspecified (including State Regulated and
Self-defined Hazardous Wastes)
10
15
20
25
30
33.0
35
Dark shaded portion represents the estimated minimum quantity managed (derived
from the larger of quantities treated, stored, or disposed for each waste code, summed
over all facilities).
Total bar indicates the quantity handled.
Note: Important notes are listed on the following page.
180
-------
Notes to Figure 30
1.) A high degree of uncertainty is associated with the quan-
tities shown due to questionnaire design, sampling and
nonsampling error, and the lack of correlation with similar
data derived from the Generator Questionnaire (see Fiqure
22) .
2.) The sum of the amounts handled by the different waste groups
exceeds the 71.3 billion gallons managed in the U.S. for two
reasons. First, some respondents may have interpreted the
term "handled" more broadly than intended in the survey
design by including: additional quantities generated or
transported that were not (i.e., treated, stored for more
than 90 days, or disposed of) managed on-site; quantities in
storage generated in previous years that were carried over
as stocks in hand; and quantities managed in RCRA-exempt
processes (see Introduction into Part III). Second, some
hazardous waste quantities were reported by respondents
under multiple hazardous waste codes, resulting in some such
quantities being counted in more than one of the categories
presented in the figure. Quantities reported under multiple
hazardous waste codes generally fell into two groups:
a.) An estimated 6.6 billion gallons of hazardous waste
were represented by respondents as mixtures of two or
more hazardous wastes that could not be disaggregated.
Where such mixtures included wastes from more than one
category (e.g., U122 and FOOD, the entire amount was
included in each appropriate waste group presented in
the figure (e.g., F001-F005 and U001-U247). Some mix-
tures included numerous waste codes and, thus, could be
counted many times; and
b.) Multiple classification also occurred wherever respon-
dents reported a single waste (i.e., not a waste mix-
ture) using more than one hazardous waste code. Many
solvents, for example, are also ignitable and were
sometimes reported as D001 as well as one of the F001-
F005 solvent waste codes, again resulting in double
counting.
181
-------
In the case of the uppermost bar, representing quantities
of spent halogenated and nonhalogenated solvents, classified
under EPA waste codes F001 through F005, both estimates are the
same. For acutely hazardous wastes, the quantities were too
small to be graphically depicted as two estimates, but on a
larger scale, the lower estimate would be about three-quarters
of the amount handled, which is depicted in the full bar. For
corrosive wastes (D002), however, the two estimates are
considerably different.
The actual amount managed (as opposed to "handled")
falls someplace in between the two estimates illustrated in
Figure 30 (again, ignoring sampling and nonsampling error).
Both estimates, however, may understate the quantity estimates
for the "U" and "P" waste groups that are characteristically
managed in small quantities and may not have been adequately
represented by some of the 12 percent of all TSD facilities that
have 10 or more waste streams.
Another problem in the analysis of waste group data is
that there is a tradeoff between data quality and detail. The
more detailed the breakdown, the more extreme the sampling error
attached to the estimate is likely to be. Confidence intervals
for individual waste groups were not calculated, but are expected
to be very wide, indicating low statistical reliability. For
this reason, the waste groups depicted in Figure 30 reflect
convenient conceptual groupings and do not imply any acceptable
degree of statistical reliability.
Taking sampling error, nonsampling error, and the
limitations of the questionnaire into account, however, we cannot
be confident that these quantity estimates reflect the actual
waste group quantity distribution in 1981, except in very general
terms. We can conclude, for example, that the quantity of
182
-------
corrosive wastes is probably greater than any other waste group,
acutely hazardous waste quantities less than any other waste
group, and that there are probably more "K" wastes (listed industry
wastes from specific sources) and E.P. toxic wastes than ignitable
(D001), or "U" wastes. Attempts to draw more detailed or specific
conclusions from those waste-specific quantity data are, however,
strongly discouraged.
7-6 Quantities Managed by Commercial Management
Facilities
The number of facilities that can be characterized as
commercial has been discussed previously (see Section 5.3).
Table 15 reviews the distribution of commercial and other manage-
ment type sites and shows the corresponding quantity measures.
Although commercial facilities make up nearly seven
percent of all management sites, the data indicate they account
for only about two percent of the actual hazardous waste managed,
On the other hand, those facilities characterized as receiving
50 percent or less of the hazardous waste they managed from
other firms make up 85.5 percent of the total number of all
facilities. The data indicate that these facilities account for
91.4 percent of the managed hazardous waste. The category NSK
(not specified by kind) came into being because of missing data
resulting from the construction of the different management
classes established here.
183
-------
Table 15. Quantity Distribution for Commercial and Other Types of TSD Facilities
Estimated Quantity
Percent of
Estimated Estimated
Number of Total (mi
Type of Facility
Commercial (privately
owned and operated —
more than half of waste
from other firms)
Publicly owned or
operated and:
More than half of waste
from other sites
i—•
oo
^ Half or less of waste
from other sites
Other: NEC
NSK
Estimated
Number of
Facilities
326
37
4,122
9
324
Percent of
Estimated
Total
Number
6.8%
0.8%
85.5%
0.2%
6.7%
(million
gallons)
1,300
24
65,000
2
4,700
(thousand
metric
tonnes)
4,800
89
240,000
7
17,000
Percent of
Estimated
Total
Quantity
1.9%
0.1%
91.4%
0.1%
6.7%
-------
7•7 Quantities of Hazardous Waste Treated, Stored and
Disposed During 1981.
Most of the 71.3 billion gallons3 of hazardous waste
managed at TSD facilities during 1981, was treated; approximately
two thirds (66.5% or 47.5 billion gallons; see Figure 31)3 of
the hazardous waste managed by TSD facilities was treated. During
1981 there was a total of 1,495 TSD facilites that treated
hazardous waste. The average quantity treated at each facility
was 31.7 million gallons (see Table 16). Figure 31 presents the
quantity of hazardous waste treated in 1981, in comparison to
the quantities of hazardous waste stored and disposed. As the
figure indicates, about thirty percent more waste was treated,
than was stored; and more than three times as much waste was
treated as was disposed. It should be pointed out, however,
that significant portions of the quantity of treated hazardous
wastes were also stored and/or later disposed. A strict comparison
of these numbers is therefore discouraged. Section 7.3.1 contains
a detailed discussion of the quantities of hazardous waste treated
by each treatment technology.
About half of the hazardous waste that was managed by
TSD facilities during 1981 was stored. Of the 71.3 billion
gallons of managed waste, 36.5 billion gallons (or 51.1%) was
stored at TSD facilities. During 1981, there were 4,299 TSD
facilities that stored hazardous waste, and the average quantity
stored per facility was 8.5 million gallons (see Table 16.)
Figure 31 presents the quantity of hazardous waste stored in
1981, compared to the quantities treated and disposed. About 30
Quantities treated and managed are overstated by roughly 5
billion gallons due to the failure of some respondents to
exclude exempted treatment tank wastewater in the TSD General
Survey.
185
-------
Figure 31
QUANTITIES OF HAZARDOUS WASTE TREATED,
STORED AND DISPOSED IN 1981
TOTAL QUANTITY OF
C/1
Z
o
CQ
3U —
40 -
30 -
20 -
10 -
1
47.5*
WASTE MANAGED =
71.3 Billion gallons*
36.5*
14.7*
QUANTITY QUANTITY QUANTITY
TREATED STORED DISPOSED
* Treated + Stored + Disposed exceeds 71.3 billion gallons managed due to
multiple processing of some wastes. Roughly 5 billion gallons oT RCRA-exempt
wastewater are included in the treatment and management total.
186
-------
Table 16. Total Quantities Managed, Average Quantities Managed per Facility, and
Number of Facilities Treating, Storing and Disposing of Hazardous Waste
in _Lyoi
Average Quantity
Pr°CeSS M Tot*1 Q^ntity Managed Per Facility, Number of Facilities
Managed, By Process By Process Engaged In Each
(Billions of Gallons) (Millions of Gallons) Process
Treatment 47.5*
31.7
Storage 36 5 o c
y Jb A ~>
00 ^ -L<± . / 34.3 430
*Roughly 5 billion gallons of exempted wastewater are included due to the failure of somp
respondents to exclude these amounts from the TSD General Survey? ±ail^^e of some
-------
percent more waste was treated than stored, and about 60 percent
less waste was disposed than stored. Section 7.3.2 contains a
detailed discussion of the quantities of hazardous waste stored
in each storage technology.
One fifth of the 71.3 billion gallons of hazardous
waste that was managed during 1981 was disposed by TSD facilities
during 1981. Approximately 14.7 billion gallons (or 20.6% of
the managed hazardous waste) was disposed by 430 TSD facilities.
The average quantity disposed per facility was 34.3 million
gallons (see Table 16). Figure 31 presents the quantity of
hazardous waste disposed in 1981, compared to the quantities
treated and stored. More than twice as much waste was stored,
and more than three times as much waste was treated, than was
disposed. Most of the quantities stored, however, were subse-
quently treated and disposed of, rendering strict comparisons of
the quantities treated, stored, and disposed of less meaningful.
The sum of the quantities of waste that were treated,
stored and disposed is greater than the total quantity that was
managed, due to multiple processing of some wastes. An important
note, however, is that the sum of hazardous wastes treated and
disposed (62.2 billion gallons) significantly exceeds the estimated
total quantity of hazardous waste generated during 1981. Thus,
even when double counting of quantities treated and disposed of
is accounted for by assuming that all wastes disposed of were
first treated, the survey findings fail to indicate a short fall
between the quantity of hazardous waste generated and the quantities
subsequently finding their way into final treatment and disposal
processes regulated under RCRA in 1981.
188
-------
7.7.1 Quantities of Hazardous Waste Treated in 1981, by
Treatment Process Type
Treatment surface impoundments handled the greatest
single portion of hazardous waste treatment, treating an estimated
16.6 billion gallons (62 million metric tonnes) during 1981. As
shown in Figure 32, there was almost twice as much hazardous
waste treatment in surface impoundments as there was in tanks.
There was nearly forty times more treatment in surface impoundments
than there was in incinerators. There were about 410 treatment
surface impoundment facilities treating hazardous waste during
1981. They treated an average of 40.6 million gallons (151,000
metric tonnes) of hazardous waste per facility (see Table 17).
However, as is shown in Table 18, a very large percentage of the
total quantity of hazardous waste that is treated in surface
impoundments is treated by a very small number of facilities.
Half of that waste was treated by only six percent of the facilities
treating their hazardous waste in surface impoundments during
1981. Forty-two percent of those facilities accounted for 99
percent of the hazardous waste treatment in surface impoundments.
As was indicated in Section 5.4.1, treatment surface
impoundments that are employed to treat hazardous wastewater
under NPDES (National Pollutant Discharge Elimination System)
permits are not excluded from RCRA regulation (although NPDES
permitted treatment tanks are excluded). Thus, the estimated
quantity of hazardous waste treated in surface impoundments
includes hazardous wastewater treated by facilities engaged in
NPDES wastewater treatment in surface impoundments, while the
estimated quantity of hazardous waste treated in treatment tanks
excludes hazardous wastewaters treated by those facilities engaged
in NPDES wastewater treatment in tanks.
189
-------
Figure 32
QUANTITIES OF HAZARDOUS WASTE TREATED IN 1981, BY
TREATMENT PROCESS TYPE
(billions of gallons)
20 -i
15 -
o
_!
_l
<
c/i
Z
o
00
10 -
5 -
16.60C
(62 MMT)
TOTAL QUANTITY TREATED
= 47.45 Billion gallons
(176 Million metric tonnes)
8.73"
(32 MMT)
4.58C
(17 MMT)
0.45 u
(1.7 MMT)
J_
TREATMENT
TANKS
TREATMENT
SURFACE
IMPOUNDMENTS
INCINERATORS
OTHER
TREATMENT
Note: Footnotes are listed on the following page.
190
-------
Notes to Figure 32
quantity treated includes roughly five billion gallons
in the TSD^en^r1? 1°™*** °f RCRA— ^ wastewater misrlpor?ed
in the TSD General Survey. Due to this and the large samplina
errors associated with quantity estimates of the Tank, ISrfac!
Impoundment and TSD General Surveys, the total quantity trealld
^h g^SKer ^ the SUm °f the Passes even though the processes
should be greater to the extent they overlap. Processes
^nimaK6d^y multi^1yj-n(3 the number of facilities with treatment
tanks by the mean of Tank Questionnaire question 18, quantity of
hazardous wasl-p t-naat-^H i ^ 4-o~i^ ' 4uani.j.T:y or
hazardous waste treated in tanks.
Estimated by multiplying the number of facilities with treatment
nai^^^^fr^^L "!« ™? _?L?«f««. ^P°-dment Question-
Estimated by multiplying the number of facilities with active
incinerators in 1981 by the mean of TSD General Questionnaire
question 17, quantity of hazardous waste treated, for incinerated
h^ndtV Y;h ThlS ^ bS unde^tated since only the 10 wastes
handled in the greatest quantity were reported.
Estimated by multiplying the number of facilities with other
treatment in 1981 by the mean of the TSD General Questionnaire
question 17, quantity of hazardous waste treated, for wastes
treated by other treatments only. Combinations with tank surface
impoundments or incinerators were not included, and so waste
streams may also have been excluded where there were more than
10 waste streams in a facility.
191
-------
Table 17. Total Quantities Treated, Average Quantities Treated and Number of Facilities
Treating Hazardous Waste by Each Treatment Process Type
Treatment
Process
Type
Treatment
Tanks
Treatment
Surface
Impoundments
Incinerators
Other
Treatment
Total Quantity Treated
By Process Type
(Billion (Million
Gallons) Metric
Tonnes )
8.73
16.60
L
0.45
4.58a
32
62
1.7
17a
Average Quantity Treated
By Process Type
(Million (Thousand
Gallons) Metric
Tonnes )
14.3
40.6
1.9
12. 9a
53
151
7
48a
Number of Facilities
Treating By Each
Process Type
609
410
240
392b
LBased on 355 facilities that were not in combination with the above processes
'includes 37 facilities that were in combination with the above processes.
-------
Table 18.
Size Distributions in the Treatment Technologies: Proportions of the
Population Accounting for 33 Percent, 50 Percent and 99 Percent of the
Quantities of Waste Treated in Each Technology
Treatment
Technology
Treatment
Tanks
OJ
Treatment
Surface
Impound-
ments
Incinera-
tors
Percent of the Popula-
tion That Accounts For
33% of the Waste
Treated in Each
Technology
1%
3%
1%
Percent of the Popula-
tion That Accounts For
50% of the Waste
Treated in Each
Technology
2%
6%
4%
Percent of the Popula-
tion That Accounts For
99% of the Waste
Treated in Each
Technology
30%
42%
47%
-------
Treatment tanks accounted for the next greatest quantity
of hazardous waste treatment, with an estimated 8.7 billion
gallons (32 million metric tonnes) of treatment during 1981. As
Figure 32 indicates, there was almost twice as much hazardous
waste treatment in surface impoundments, compared to tanks. But
roughly twenty times more waste was treated in tanks than was
incinerated, and almost twice as much treatment occurred in
tanks, compared to "other treatment" methods. There were about
609 treatment tank facilities in 1981, and they treated an average
of 14.3 million gallons (53,000 metric tonnes) of hazardous
waste per facility. The average treatment quantity for treatment
tanks is about one third of the average treatment quantity for
treatment surface impoundments. However, as Table 18 indicates,
the bulk of the total quantity of hazardous waste treated in
tanks was treated by a very small proportion of the facilities.
About two percent of the facilities that treated hazardous waste
in tanks in 1981 treated 50 percent of the waste that was treated
in tanks. Just 30 percent of the facilities accounted for 99
percent of hazardous waste treatment in tanks.
Many TSD facilities indicated that they treated hazard-
ous wastewaters in tanks that were covered under NPDES permits.
As is indicated in Section 1.3, such tanks are excluded from
regulation under RCRA, and this survey was not designed to
provide estimates of the quantities of hazardous wastewater
treated in tanks operating under NPDES permits in 1981. However,
a large number of survey respondents reported that they operated
treatment tanks under NPDES permits in 1981. This suggests
that, had hazardous wastewater treatment quantities been included
in the quantity estimate for treatment tanks, the estimate would
have been much larger.
194
-------
Incinerators treated the least amount of hazardous
waste of the listed treatment processes. About 450 million
gallons (1.7 million metric tonnes) of waste was incinerated
during 1981. As is shown in Figure 32, about twenty times more
waste was treated in tanks, and about forty times more waste was
treated in surface impoundments. Nearly twice as much hazardous
waste (807 million gallons or 3.0 million metric tonnes) was
disposed in landfills (see Section 7.3.3). Even the category
used to specify other unlisted types of waste treatment is eight
times the quantity that was incinerated. Table 17 indicates
that an average of 1.9 million gallons (17,000 metric tonnes)
was incinerated per incinerator facility during 1981. However,
as Table 18 indicates, less than 4 percent of the population of
incinerator facilities accounts for half of the total quantity
of hazardous waste incinerated, -and the top 47 percent of incin-
erator facilities account for 99 percent of all hazardous waste
incineration.
Treatment methods other than the three treatment methods
specified in the survey questionnaire were used to treat approxi-
mately 4.6 billion gallons of hazardous waste during 1981. There
were an estimated 392 facilities using these treatment methods,
to treat an average of 12.9 million gallons of hazardous waste
per facility. Other treatment methods included open burning,
explosion, treatment in waste piles, and treatment in containers.
In addition, some treatment technologies that were classified by
survey respondents as "other," but unspecified, may prove to be
classifiable as a type of treatment in tanks or impoundments.
The estimated total quantity treated (see Section 7.3)
from the TSD General Survey is 47.5 billion gallons (176 million
metric tonnes). The sum of treatment quantities estimated from
the technology component surveys (29.4 billion gallons or 109
195
-------
Figure 33
QUANTITIES OF HAZARDOUS WASTE STORED IN 1981 BY
STORAGE PROCESS TYPE
(billions of gallons)
TOTAL QUANTITY STORED
= 36.40 Billion gallons
(135 Million metric tonnes)
14.10C
(52 MMT)
CO
O
_j
<
(D
u_
O
to
CD
0
STORAGE
TANKS
STORAGE STORAGE
CONTAINERS SURFACE
IMPOUNDMENTS
WASTE
PILES
OTHER
STORAGE
Note: Footnotes are listed on the following
page.
198
-------
Notes to Figure 33
Estimated by multiplying the number of facilities with
"" "
.
plus the quantity ,nt.«d tnto ,tor,g, In conXner,™ J9™'
2. fs±!rI=Kttj.trsj:l'ssr1ias.-5^-? ^
previous; years in their estimates of quantities stored.
orwas
199
-------
Table 19. Total Quantities Stored, Average Quantities Stored and Number of Facilities
Storing Hazardous Waste by Storage Process Type
Treatment
Process
Type
Storage
Tanks
Total Quantity Stored,
By Process Type
(Billion (Million
Gallon Metric
Tonnes)
5.10
19
Average Quantity Stored,
By Process Type
(Million (Thousand
Gallons Metric
Tonnes )
3.57
(Carry
over
= 0.2,
Entered
= 3.6
13
Carry over
= .7
Entered
= 13.4)
Number of
Facilities Stored By
Each Process Type
1,428
o
o
Storage
Containers
0.16
.6
.045
(Carry
over
= .009,
Entered
= .039
17
(Carry over
= .03
Entered
= .15)
3,577
Storage
Surface
Impoundments
14.10
52
25.6
95
552
Waste Piles
0.39
1.5
2.2
174
Other Storage
.26a
.97a
1.9a
-------
Furthermore, unlike storage tanks, storage surface
impoundments that are employed to store hazardous wastewater
under NPDES permits are not excluded from regulation under RCRA.
Thus, the estimated quantity of waste stored in surface impound-
ments includes hazardous wastewaters that are stored in surface
impoundments operated under NPDES permits, while the estimated
quantity of hazardous waste stored in tanks excludes hazardous
wastewaters that are stored in tanks operated under NPDES permits.
Tanks were used to store the second largest quantity
of hazardous waste. About 5.1 billion gallons (19 million metric
tonnes) were stored by 1,428 facilities. Approximately 180
million gallons (0.4%) of the 5.1 billion gallons were carried
over in storage from 1980 (see Figure 33 and Table 19.) As was
indicated above, the estimated quantity of hazardous waste stored
in tanks does not include hazardous wastewaters stored in tanks
operated under NPDES permits, nor does it include quantities of
hazardous waste accumulated on site under the 90-day rule.6
Storage tank quantities were dominated by two facilities
in the sample that represented almost three-quarters of all
RCRA-regulated hazardous waste entering storage tanks in 1981.
The largest facility sampled represented 52 percent of the total
quantity entering storage, but only 0.3 percent of the storage
tank facility population. Ninety-nine percent of storage tank
quantities were accounted for by only 27 percent of the storage
tank population (see Table 20).
hazardous waste under the 90-day rule during
201
-------
Storage containers were used by 3,577 facilities to
store about 160 million gallons (600,000 metric tonnes) of
hazardous waste. Seventeen precent of this, or 27 million
gallons (100,000 metric tonnes) was carried over in storage. As
was indicated above, the quantity of hazardous waste stored in
containers does not include quantities of hazardous waste accumu-
lated on site under the 90-day rule. About 51 percent of the
14,098 hazardous waste generators used containers for on-site
accumulation of hazardous waste under the 90-day rule in 1981.
While storage in containers was the most frequently used technology
among all hazardous waste management technologies, it ranked
smallest among the four storage technologies surveyed in terms
of quantity managed.
Quantities of hazardous waste entering storage in
containers were also concentrated in a relatively small number
of facilities, although the distribution was not as highly skewed
as in storage tanks. Table 20 indicates that a third of the
containers' waste was stored by one percent of the facilities
entering hazardous waste into storage containers during 1981.
Two percent of the population accounted for half the waste and
two thirds percent of the population accounted for 99 percent of
the waste entering container storage.
Waste piles were used by 174 facilities to store an
estimated 390 million gallons (1.5 million metric tonnes)7 of
waste. Storage tanks were used to store more than ten times
that quantity of waste, and surface impoundments stored more
than thirty times that quantity. As was true with surface
7The estimates of the quantity stored in waste piles was developed
by multiplying the number of facilities with in-use waste piles
from the Waste Pile Questionnaire, by the mean quality of hazardous
waste entered into storage in waste piles, from Question 17 of
the TSD General Questionnaire, which is limited to the ten wastes
handled in greatest quantity.
202
-------
Table 20.
Size Distributions in the Storage Technologies: Proportions of the
Population Accounting for 33 Percent, 50 Percent and 99 Percent of
the Quantities of Waste Stored in Each Technology
O
U)
Storage
Waste
Storage
Tanks'
Containers'
Storage
Surface
Impound-
ments
Percent of the Popula-
tion That Accounts For
33 Percent of the Waste
Stored in This Technology
1%
1%
Percent of the Popula-
tion That Accounts For
50 Percent of the Waste
Stored in This Technology
2%
2%
Percent of the Popula-
tion That Accounts For
99 Percent of the Waste
Stored in This Technology
27%
67%
27%
The distribution of quantities entering storage shown here should be similar to auant- i t- i -«,
stored since quantities carried over are relatively small (See Table 20? quantities
^ °'3 percent °f
and 52 percent of
-------
impoundments, waste that is "accumulated" on site in waste piles
for less than 90 days is not excluded from regulation as hazardous
waste storage under RCRA. Thus the estimated quantity of hazardous
waste stored in waste piles does include waste that was stored
on site for less than 90 days. However, the estimated quantity
managed by waste piles does not include waste that may have been
treated or disposed in waste piles, since at the time of the
survey, waste piles were regarded by the EPA as only a storage
process.
There were 139 facilities that used storage methods
other than impoundments, tanks, containers or piles in 1981.
These facilities handled about 260 million gallons (approximately
one million metric tonnes)8 of hazardous waste.
7.7.3 Quantities of Hazardous Waste Disposed in 1981, by
Disposal Process Type
While the largest number of disposal facilities have
landfills, 199 facilities, the largest quantity of waste is
disposed in underground injecton wells (see Figure 34 and Table
21). Approximately 8.6 billion gallons (32 million metric tonnes)
of waste managed as hazardous waste was disposed in injection
wells in 1981, which is more than ten times the quantity disposed
in landfills. The average quantity disposed per facility for
the 87 facilities that used injection wells for disposal was
99.0 million gallons (370,000 metric tonnes). As is shown in
Table 22, 10 percent of the injecton wells facilities were
8The quantity estimated for other storage methods is an under-
estimate derived from question 17 of the TSD General Questionnaire,
which is limited to the ten wastes handled in greatest quantity,
and limited to waste that entered storage in 1981.
204
-------
Figure 34
QUANTITIES OF HAZARDOUS WASTE DISPOSED IN 1981, BY
DISPOSAL PROCESS TYPE
(billions of gallons)
20 -i
15 -
O
CO
10 -
5 -
TOTAL QUANTITY DISPOSED
= 14.70 Billion gallons
(55 Million metric tonnes)
8.60 d
(32 MMT)
5.10C
(19 MMT)
0.81°
(3 MMT)
0.10°
(.4 MMT)
0.02 e
[.07 MMT)
INJECTION LANDFILLS DISPOSAL LAND OTHER
WELLS SURFACE APPLICATION DISPOSAL
IMPOUNDMENTS (TREATMENT)
Note: Footnotes are listed on the following page.
205
-------
Notes to Figure 34
Estimated by multiplying the number of facilities with injection
well disposal by the mean for Injection Wells Questionnaire
question 4, quantity of waste disposed by injection wells in
1981.
L
Estimated by multiplying the number of facilities with landfills
by the mean for Landfills Questionnaire question 5a2, quantity
of hazardous waste disposed in landills in 1981.
Q
Estimated by multiplying the number of facilities with disposal
surface impoundments by the mean for Surface Impoundment Question-
naire question 5, quantity of hazardous waste disposed in surface
impoundments in 1981.
Estimated by multiplying the number of facilities with land
application by the mean for Land Treatment Questionnaire
question 6, quantity of hazardous waste land treated during
1981.
Q
Estimated by multiplying the number of facilities with other
disposal by the mean of the TSD Questionnaire question 17if
quantity of hazardous waste disposed in 1981, for waste disposed
by other methods only. This represents an underestimate of the
total quantity of hazardous waste disposed by other methods
because question 17i obtained data only for the ten wastes
handled in greatest quantity.
206
-------
Table 21. Total Quantities Disposed, Average Quantities Disposed and Number of Facilities
Disposing of Hazardous Waste by Each Disposal Process Type unities
— —
Disposal
Process
Type
Injection
Wells
Landfills
Disposal
Surface
Impoundments
Land
Application
(Treatment)
Other
Disposal
Total
Quality Disposed
By Process Type
(Billion (Million
Gallons) Metric
Tonnes )
8.60
— — — . .. ,. _
0.81
5.10
0.10
0.02
32
— — — - — — — . —
3
19
.4
.07
Average
Quantity Disposed
By Process Type
(Million (Thousand
Gallons) Metric
Tonnes )
99.0
4.1
44.0
1.4
3.3
370
15
160
5
12
Number of
Facilities Disposing
By Each Process Type
87
199
116
70
7
-------
Table 22. Size Distributions in the Disposal Technologies: Proportions of the Population
Accounting for 33 Percent, 50 Percent and 99 Percent of the Quantities
Waste of Disposed in Each Technology
Disposal
Technology
Percent of the Popula-
tion That Accounts For
33 Percent of the
Waste Disposed in
Each Technology
Percent of the Popula-
tion That Accounts For
50 Percent of the
Waste Disposed in
Each Technology
Percent of the Popula-
tion That Accounts For
99 Percent of the
Waste Disposed in
Each Technology
NJ
o
oc
Injection
Wells
Landfills
Disposal
Surface
Impound-
ments
Land
Applica-
tion
6%
2%
10%
3%
63%
52%
2%
3%
20%
7%
13%
65%
-------
responsible for fifty percent of the total quantity of disposal
in injection wells. It is significant to noter however, that
the injection well population was the least skewed of all the
hazardous waste treatment, storage, and disposal process popula-
tions. A greater percentage of the injection well sites are
"larger" than was observed in any of the other processes.
The second largest quantity of hazardous waste was
disposed in disposal surface impoundments. Approximately 5.1
billion gallons of waste was disposed in 1981 (see Figure 34 and
Table 21). The 116 facilities that disposed of waste in surface
impoundments disposed an average of 44.0 million gallons (160,000
metric tonnes) of hazardous waste per facility. However, three
percent of the facilities with disposal surface impoundments
disposed of 50 percent of the waste disposed in surface impound-
ments, and 99 percent of the waste was disposed by just 20 percent
of the facilities. In fact, most of the disposal in surface
impoundments was observed at a single extremely large facility.
Were it not for the inclusion of this facility in the sample,
the estimated quantity of hazardous waste disposed in surface
impoundments would be substantially reduced.
About 810 million gallons (3 million metric tonnes) of
hazardous waste were disposed in landfills during 1981. As is
shown in Figure 34 this represents about one-tenth of the waste
that was disposed in injection wells and about one-sixth of the
waste that was disposed in surface impoundments. Disposal in
landfills, however, was about eight times greater than disposal
by land application (treatment). The 199 facilities using
landfills for hazardous waste disposal disposed an average of
4.1 million gallons (15,000 metric tonnes) of hazardous waste
per facility during 1981. But, as is shown in Table 22, a small
209
-------
number of facilities were responsible for a large proportion of
the waste disposed in landfills. Approximately three percent of
the facilities disposed of 50 percent of the hazardous waste
landfilled.
As is shown in Figure 34 and Table 21, about 100 mill-
ion gallons (400,000 metric tonnes) of hazardous waste was
disposed by land application (treatment) during 1981. The 70
facilities using land application for hazardous waste disposal
disposed of an average of 1.4 million gallons (5,000 metric
tons) of waste per facility during 1981. However, as is shown
in Table 22, only 13 percent of the land application facilities
accounted for 50 percent of the disposal by land application.
About 20 million gallons (70,000 metric tonnes) of
hazardous waste were disposed by other methods during 1981. The
estimated 7 facilities using other disposal methods disposed of
an average of 3.3 million gallons (12,000 metric tonnes) of
waste by other methods during 1981.
210
-------
PART IV (Section 8)
CAPACITY
-------
EPA's HWDMS file of Part A permit applications;1
to the extent that these out-of-scope facilities
actually manage or intend to manage hazardous
waste, additional capacity beyond that indicated
by the findings presented in this section could be
available.
• The data represent capacity on January 1, 1982.
Additional capacity may have been "used up" in the
past two years. Also, additional capacity may
have been "created" at existing facilities.
• The reported capacity represents the respondents'
own perceptions of additional quantities of
hazardous waste that "could have been treated,
stored or disposed of in 1981." Presumably, this
included only existing capacity; no attempt was
made to obtain potential for expanding capacity.
• There is some uncertainty as to the meaning of the
respondents' answers to the capacity questions.
While the answers clearly represent annual "flow"
capacity for treatment processes and some disposal
process (e.g., underground injection), it is unclear,
in the case of landfills particularly, whether
respondents reported annual input rates or total
fixed capacity in response to the question: "What
quantity of hazardous waste could have been dis-
posed of in 1981?" This report assumes, however,
that the data represent annual input capacity for
all processes.
The above points underline the reasons for carefully
interpreting the "capacity" data that follow. Like other quantity
data in this report, important relationships between different
types of management facilities can be observed, even though the
capacity levels may be of questionable accuracy. Sections 8.1 to
8.4 will examine the relationships between different types of TSD
J-Among the facilities listed on EPA's HWDMS file of Part A, the
survey estimated that 4,818 actually managed hazardous waste in
1981. The remaining 3,700 facilities listed on the Part A file
are estimated to have not entered quantities of hazardous waste
into treatment, storage or disposal in 1981, and thus were out-
side the scope of the survey.
212
-------
8. CAPACITY OF TREATMENT, STORAGE AND DISPOSAL FACILITIES
The preceding sections have provided an analysis of
quantities of hazardous waste managed in 1981. This final quan-
titative section of the report focuses attention on the capacity
of these facilities to treat, store and dispose of additional
quantities of hazardous waste. Capacity availability is an
important issue in planning for adequate management of future
quantities of hazardous waste.
Respondents to the national TSD General Questionnaire
were asked what quantities of hazardous waste (a) actually were
and (b) could have been treated, stored or disposed of in 1981.
The quantities reported in the "could have been" category were
intended to represent the facility's annual capacity to treat,
store, and dispose of these wastes. As a result, it is possible
to calculate both the "percent utilization of capacity" and the
"unused annual capacity" as of the beginning of 1982. This
section also analyzes the extent to which hazardous waste man-
agement capacity is "readily available" in that it is available
at facilities that receive over half of their waste from other
firms (i.e., commercial facilities).
The concept of "capacity" requires some careful defini-
tion. The following points of clarification are provided to
qualify the meaning of capacity as the term is used in the
analysis that follows:
• The reported capacity was based only on those
facilities estimated to have managed RCRA-regulated
hazardous waste in 1981. Almost as many more
facilities were "out of scope" (ineligible for the
survey) because they did not manage hazardous
wastes in 1981, even though they were listed in
211
-------
Table 23. Summary of 1981 U.S. "Commercial" and On-Site Hazardous
Waste Management Capacity Utilization
TREATMENT
a) Total Capacity in Bill. Gal.
(MMT)
b) Unused Capacity in Bill. Gal.
(MMT)
c) Cumulative Capacity
Utilization^
d) Average Facility Capacity
Utilization^
STORAGE
a) Total Capacity in Bill. Gal.
(MMT)
b) Unused Capacity in Bill. Gal.
(MMT)
c) Cumulative Capacity
Utilization^
d) Average Facility Capacity
Utilization
DISPOSAL
a) Total Capacity in Bill. Gal.
(MMT)
b) Unused Capacity in Bill. Gal.
(MMT)
c) Cumulative Capacity
Utilization^
d) Average Facility Capacity
Utilization-^
Commercial-'-
Capacity
1.7
(6)
1.1
(4)
35%
37%
1.8
(7)
1.1
(4)
39%
47%
2.4
(9)
1.6
(6)
33%
42%
On-Site
Capacity
203.7
(758)
156.8
(583)
23%
51%
55.2
(205)
19.44
(72)
65%4
57%
38.7
(144)
24.7
(92)
36%
56%
Total U.S.
Capacity
205.4
(764)
157.9
(587)
23%
50%
57.0
(212)
20. 54
(76)
64%4
57%
41.1
(153)
26.3
(98)
36%
53%
MMT = Million metric tonnes.
Note: Footnotes are listed on following page.
214
-------
capacity (commercial, on-site, and regional), and their capacity
utilization rates. These data were imputed based on the esti-
mated treatment, storage, and disposal quantities presented in
Section 7.3, and their national capacity utilization rates.2
8<1 U.S. Capacity Utilization: Commercial and On-Site
Management Facilities
Ample hazardous waste management capacity appears to be
available for 1981 levels of hazardous waste generation. Only 23
percent of treatment, 36 percent of disposal, and, at most, 64
percent3 of storage capacity was utilized across the U.S.
The treatment, storage, and disposal capacity utiliza-
tion rates are presented in the "c" rows of Table 23. This table
illustrates the importance of on-site capacity compared to com-
mercial capacity. "Commercial" facilities, defined here as those
private or public facilities receiving more than half of their
waste from off site, make up less than one percent of total treat-
ment capacity, six percent of disposal capacity, and three percent
of storage capacity. Thus, the U.S. hazardous waste management
capacity and utilization levels are dominated by on-site capacity.
Treatment, storage and disposal capacity utilization rates were
calculated by 100+ C/D, where C = weighted mean of quantities
that "entered," and D - weighted mean of quantities that "could
have been entered."
3This may significantly overstate storage capacity utilization
since the largest facility, which represented 22 percent of the
total storage capacity, did not provide capacity data and was
therefore assumed to have fully utilized its storage capacity in
1981. if this facility were not included, or had it been assumed
to have the same storage capacity utilization rate as the rest of
the storage populations, the U.S. percentage of hazardous waste
storage capacity utilization would be 48 percent.
213
-------
The U.S. capacity and utilization levels are also
strongly influenced by the very large management facilities.
Unused capacity and low utilization rates tend to be concentrated
in the largest facilities. A comparison of the average facility
capacity utilization rate (the "d" rows of Table 23) and the
cumulative capacity utilization rate (the "c" rows) indicates
that the average facility rate is significantly higher for treat-
ment and disposal.^ Low capacity utilization rates at large
facilities caused the cumulative rate to fall significantly below
the average facility rate. The highly skewed distributions of
unused capacity shown in Table 24 accentuate this effect. Large
commercial facilities also seemed to have lower capacity utili-
zation rates than the average commercial facility, but the dif-
ference was not as pronounced as for the on-site facilities,
owing to the greater homogeneity of the commercial population.
In examining the relative scale of commercial versus
on-site hazardous waste management activity, their distributional
differences become obvious. The average commercial treatment
facility had only one-tenth the total capacity of the average
on-site treatment facility. Commercial disposal was only a
quarter as big and storage only half as big as the average on-
site disposal and storage facilities. These comparisons are
heavily influenced, however, by the few, very large, on-site
facilities. On average, commercial management facilities are
larger than the on-site facilities, even though the few very
large facilities tend to be on site. The typical (median) com-
mercial treatment facility was five times larger than the typical
^Ignoring the largest storage facility that was assumed to be at
full capacity, the average facility storage rate was also higher
than the cumulative capacity utilization rate.
216
-------
Notes to Table 23
1"Commercial" facilities are defined in various ways in this and
other reports. Here commercial facilities are defined as those
receiving more than 50 percent of their waste from off site
This includes public as well as private facilities.
Cumulative capacity utilization is the total quantity treated,
stored or disposed of in the U.S., divided by the total T, S or
D capacity. '
3Average facility capacity utilization is a simple average of the
utilization rates computed for the individual facilities.
4These data reflect the assumption that the largest storage facil-
ity (representing 12.4 billion gallons) had no unused capacity
This is an extreme assumption that almost certainly understates
unused capacity and overstates cumulative capacity utilization
215
-------
on-site treatment facility, and commercial disposal and storage
were 3 and 33 times larger respectively.
A final difference between commercial and on-site haz-
ardous waste management capacity is that the commercial capacity
focuses primarily on disposal while the on-site capacity is mostly
for treatment. This can be observed in Table 23.
8.2 Regional Utilization of Existing Capacity
As of January 1, 1982 facilities' average percentage of
capacity that was already being used for the treatment, storage
and disposal of hazardous waste is shown in Figure 35 by region.
Most of these percentages are accurate to + 10 percent or better
at the 95 percent confidence level (Region X, however, with a
much smaller sample, had less accuracy — i.e., + 20%).
With few exceptions, the ten regions of the United
States had average treatment, storage, and disposal capacity
utilization rates of between 45 and 65 percent during 1981. Excep-
tions include: Region VIII with an average facility storage
capacity of 40 percent, and Regions X and VII, where the estimates
were based on limited samples. Estimates for Region X indicate
that all of the disposal capacity was used up, but this estimate
was based on only two reporting facilities, and is very approxi-
mate. Similarly, average treatment capacity utilization for this
region (28%) was based on only six facilities, and the 28 percent
average disposal capacity utilization in Region VII is based on
only five facilities.
218
-------
Table 24
*>r Treatment, Storage
NJ
Unused
Treatment
Capacity
(gallons)
0
12,000
150,000
4,600,000
81,000,000
8,000,000,000
Cum.
of TSD
Facil.*
13
25
50
75
90
100
Unused
o torage
Capacity
(gallons )
0
610
9,800
74,000
840,000
O inn n.on n o ^- 4- +
Cum.
of TSD
Facil. *
20
25
50
75
90
-LUU
Unused
Disposal
Capacity
(gallons )
0
9,800
1,300,000
21,000,000
200,000,000
S b U , 0 0 0 , 0 0 0
Cum.
of TSD
Facil . *
20
25
50
75
90
100
*This percentage
to the left.
of facilities haH I^QQ
* ~~~~
ganons ot unused capacity shown
12-4
-------
8.3 Unused Capacity by Region
The preceding analysis focussed on the proportion of
capacity that has already been used (capacity utilization). This
section focuses attention on remaining capacity, as measured in
billions of gallons. The unused capacity, as of January 1, 1982,
for treating, storing, and disposing of hazardous waste is shown
by region in the maps of Figures 36, 37, and 38 respectively.
Nationally, almost none of this unused capacity (less than 2%)
was located at facilities that received more than half their
hazardous waste from other firms. Therefore, it should be noted
that even though substantial unused treatment, storage, and dis-
posal capacity existed in 1981, very little of this capacity was
available at commercial facilities and public facilities that
service other hazardous waste generators.
In contrast to the earlier analysis of "average facil-
ity capacity utilization rates," the "unused capacity" varies
dramatically by region. The estimates from the survey show that
Region V has almost half oŁ the nation's unused treatment capac-
ity. Regions IV and VI follow, each having about 15 percent of
the nation's unused treatment capacity. However, as indicated
above, most of this unused treatment capacity was located at
captive facilities and may not, in fact, have been available to
generators at large. As was shown in Table 24, much of this
unused capacity was located at a very few large facilities.
Unused storage and disposal capacity was also highly
concentrated in certain regions. Over three-quarters of the
nation's unused storage capacity was located in Regions II and
VI, while nearly 60 percent of the nation's disposal capacity was
located in Region IV. Another quarter of the total disposal
capacity was also located in Region VI.
220
-------
REGION
0
Figure 35
1981 AVERAGE FACILITY TSD
CAPACITY UTILIZATION RATES, BY REGION
Average Percentage of Capacity Utilized in 1981
50%
ill
IV
VI
VII
VIII
IX
100%
TREATMENT
STORAGE
DISPOSAL
TOO
-------
Figure 37
1981 REGIONAL UNUSED STORAGE CAPACITY
(billions of gallons)
8.4
TOTAL U.S. UNUSED STORAGE CAPACITY =
20.5 Billion gallons*(76 Million metric tonnes)
*May be understated as unused capacity for largest facility
(representing 12.4 billion gallons) was unknown.
222
-------
Figure 36
1981 REGIONAL UNUSED TREATMENT CAPACITY
(billions of gallons)
71
10
24
10
12
TOTAL U.S. UNUSED TREATMENT CAPACITY =
158 Billion gallons (587 Million metric tonnes)
-------
Again, readers are cautioned that these figures are
estimates, as of January 1, 1982, and subject to change. In the
past two years it is to be expected that some of this capacity
has been used up, and that some new capacity has been created;
the net effect of these changes in 1.982 and 1983 is beyond the
scope of this present study. Furthermore, some capacity cur-
rently in use for treatment or storage (e.g., in tanks) may be
emptied as stocks are reduced and thereby become available once
again as treatment or storage capacity in future years. Other
types of capacity are more nearly fixed (e.g., landfills), where
capacity decreases steadily with use.
8.4 Summary and Conclusions About Available Capacity
Hazardous waste generation levels during 1981 were
easily accommodated by available treatment, storage, and disposal
capacity across the U.S. In fact, the amount of unused capacity
indicates that capacity was more than adequate on an aggregate
basis.
Changes could have^ occurred in generation rates and
waste management practices since the survey. If capacity becomes
limited, small facilities could be more vulnerable to capacity
constraints since they tended to have higher capacity utilization
rates. Commercial facilities may be able to ease a tight capac-
ity situation, but their quantities represent such a small part
of the overall hazardous waste management system that on-site
facilities would have to accommodate most of the additional demands
From the 1981 survey data, however, it appears unlikely that a
capacity crisis is on the horizon.
224
-------
Figure 38
1981 REGIONAL UNUSED DISPOSAL CAPACITY
(billions of gallons)
TOTAL U.S. UNUSED DISPOSAL CAPACITY =
26.3 Billion gallons (98 Million metric tonnes)
223
-------
-------
PART V (Section 9)
FUTURE STUDIES
-------
The purpose of this section is to provide a brief over-
view of major hazardous waste information collection efforts
currently being conducted or to be conducted in the near future
by EPA. In conjunction with the data obtained through the mail
survey, these information collection activities are designed to
provide EPA with necessary information for development of regula-
tions and standards required under RCRA to assure protection of
human health and the environment in the generation and management
of hazardous waste.
9.i Mail Survey Followup Activities
Several of the information needs left unsatisfied by
the mail survey have been receiving increased attention since the
release of preliminary findings from the mail survey in August of
1983. Most prominent among these are those related to quantities
of hazardous waste affected by the mixture rule and quantities of
hazardous waste managed in processes currently exempt from regul-
ation under RCRA. In coming months, EPA's Office of Solid Waste
will be undertaking a number of efforts to try to resolve these
issues. These efforts will in all likelihood involve two tracks:
• Conducting more detailed, sophisticated analyses
of existing data obtained through the survey ques-
tionnaires, including an examination of the "hard
copy" versions of the questionnaires themselves;
and,
• Designing and implementing followup information
collection devices capable of providing greater
detail in the areas left ambiguous or with
insufficient detail in the survey.
OSW expects these efforts to continue through the cur-
rent calendar year, although specific time frames for completion
have not yet been established.
226
-------
9. NEXT STEPS: FUTURE HAZARDOUS WASTE STUDIES
The National Survey of Hazardous Waste Generators and
Treatment, Storage, and Disposal Facilities regulated under RCRA
during 1981 was designed to meet a variety of EPA's information
needs in the hazardous waste management field. While its scope
was broad, creating more than 6,000 individual statistical data
elements, it was never intended to meet all of the Agency's grow-
ing information needs in this critical issue area. The survey is
seen as providing EPA with baseline data on the nature and scope
of hazardous waste generation and management activities regulated
under RCRA, the first reliable data of this nature to be made
available since enactment of RCRA by Congress in 1976.
The national survey was not designed to provide detailed
information in a number of areas. Examples include: distinguishing
the specific quantities associated with individual constituents
in mixtures of hazardous wastes and mixtures of hazardous waste
and nonhazardous wastes; determining the numbers of sites operating
exempt wastewater treatment systems; and estimating the quantities
of hazardous wastes that are managed in exempt processes.
Followup efforts to this survey would be required to provide the
detailed information necessary in these and other areas.
In addition to questions arising from analysis of the
mail survey data, EPA faces increasing information needs in a
number of issue areas related to the Agency's continuing develop-
ment of the RCRA hazardous waste regulatory program. Detailed
information is now required on specific hazardous wastes and
hazardous waste management activities, including many currently
outside the scope of existing RCRA regulations.
225
-------
9.4 Small Quantity Generators Study
EPA is currently in the middle of a two-year, in-depth
analysis of alternative methods of regulating small quantity
generators of hazardous waste. Small quantity generators,
defined in Section 6, include such small businesses as gasoline
service stations, auto repair shops, dry cleaners, schools, amuse-
ment parks, and others. The study, begun in late 1982, is
designed to evaluate the environmental problems posed by small
quantity generators. It will analyze the types and quantities of
wastes generated by these firms, their current waste management
practices, and various strategies for controlling hazardous waste
generated by these small quantity generators. The study was
prompted, in part, by Congressional proposals to reduce or elim-
inate the small quantity generator exemptions currently in effect
in EPA's RCRA hazardous waste regulatory program.
The study involves an extensive survey of firms
believed to be small quantity generators. Also involved is an
assessment of state experiences in regulating small quantity
generators, through a series of case studies. Survey forms were
mailed to firms in the fall of 1983. Data from the survey are
expected to be compiled by mid-summer of 1984; the state case
studies are expected to be completed by the end of the summer.
EPA expects to issue regulatory proposals by May of 1985, with
final rules promulgated in 1986.
9.5 Survey of Used Oil and Waste-Derived Fuel Material
OSW is conducting a survey to collect economic and
technical information from facilities that handled (collected,
purchased, processed, transported, sold, stored) or burned used
oils or waste-derived fuel materials during 1982 and 1983. The
228
-------
9.2 Updates and Expansions to Mail Survey Data Base
OSW will update the mail survey data base over time as
continuing analyses encounter additional respondent and/or data
processing errors that have avoided detection to date despite the
extensive data editing which has been conducted. In addition,
OSW is continually exploring ways of expanding the data base
through the acquisition of additional data. Existing data sources
that can be targeted to individual sampled generators and facil-
ities will be identified and efforts made to acquire access to
the data contained therein. Finally, as additional data are
obtained through other studies, they will be stored, whenever
possible, in such a way as to allow for comparison with mail
survey data.
9-3 Continuing Analysis of Mail Survey Data
While this report summarizes some of the major findings
in the data obtained from the mail survey, it generally limits
itself to population-wide estimates. A wealth of data, more than
6,000 statistical data elements, were obtained through the survey,
however. Separate questionnaires were developed for each process
used to manage hazardous wastes under RCRA. Data relating to
facility design and operating characteristics, costs and prices
associated with waste management operations, and process-specific
capacities all await analysis by OSW, and outside parties ranging
from Congressional committees to individual members of the public.
An extensive description of the computer data base developed
through the survey is presented in Appendix D. Mail survey data
are expected to provide a basis for the development of numerous
reports by EPA and its contractors on specific hazardous waste
management issues over the coming year.
227
-------
Refining. A number of listing proposals resulting from studies
completed to date are currently being finalized. Listing pro-
posals are expected to be issued periodically throughout the
duration of the industry studies project.
9.7 RCRA Biennial Report
The Office of Solid Waste is anticipating a legislative
requirement to submit a report to Congress by March 31, 1985,
describing nationwide hazardous waste generation, treatment,
storage, and disposal during the 1983 calendar year. This report
will be based on data collected pursuant to the RCRA biennial
reporting requirements of Sections 3002 and 3004 of RCRA.
Under the requirements of 40 CFR 262.41, 264.75, and
265.75, hazardous waste handlers located in States that do not
have any form of authorization as of March 1, 1984, are required
to submit a biennial report to EPA by that date. Authorized
States must have substantially equivalent reporting requirements
and are required to submit aggregated data to EPA from their
analogs to the RCRA biennial report.
EPA intends to modify the Hazardous Waste Data Manage-
ment System (HWDMS) to aggregate the data collected from handlers
in unauthorized States through its biennial report. In addition,
EPA will be receiving aggregated data from the authorized States
by September 30, 1984. Aggregated data from both authorized and
unauthorized States will be loaded into a data base during the
late fall of 1984 to permit manipulation of State by State data
to produce a nationwide report on hazardous waste management
during the 1985 calendar year. This report is scheduled to be
completed and transmitted to Congress by March 31, 1985.
230
-------
obtained information will be used for regulatory development
purposes and to support regulatory impact analyses. Information
is being collected in two stages: Track 1, now completed,
generated data on approximately 450 handlers and 100 burners
using two detailed, full length questionnaires. Track 2, ongoing,
will develop more representative data from burners by first
characterizing the universe of potential burners (using a one-
page "short questionnaire"), and then targeting facilities that
burn "waste-derived fuels" for the completion of a more detailed
questionnaire to obtain their 1983 data. Track 2 is expected to
be completed in the spring-summer of 1984.
9.6 Industry Studies
EPA's Office of Solid Waste has been engaged in an
ongoing series of industry studies, which involve detailed evalu-
ations of the wastes generated by specific industries. These
studies, which are expected to continue indefinitely, are
designed to support EPA's decisionmaking process and the
justification for listing and not listing specific industrial
wastes as hazardous wastes, pursuant to requirements contained in
Congressional proposals for amendments to RCRA. The studies
entail engineering analyses of each industry's production
processes, site visits to confirm the engineering analyses, use
of industry-specific questionnaires for data gathering, and
verification and evaluation of the data gathered from all these
sources.
Industries that these studies have focused on to date
include: Dyes and Pigments; mdustrial Organics; Chlorinated
Organics (Chlorinated Aliphates); Pesticides; Plastics and Re-
sins; Organobromines; Rubber Processing Compounds; and Petroleum
229
-------
Figure 39
FUTURE STUDIES
Mail Survey Followup
A. Design, identify gaps, and uncertainties . . Spring 1983
Develop methodology and design data
gathering plan
B. Implement design and report on revised . . . Spring 1985
estimates
Special Analyses by Process Type 1984-85
Ongoing Data Base Update 1984-85
Small Quantity Generators National Survey
Survey data compiled Summer 1984
State case studies completed Fall 1984
Regulatory proposals Summer 1985
Waste as a Fuel
Track 1 completed Fall 1983
Track 2 completed Spring-Summer 1984
Industry Studies Ongoing
Analyses of RCRA Annual Reports
Reports submitted from handlers Winter 1984
State summaries to EPA Fall 1984
EPA summary report Winter 1985
232
-------
These future studies are summarized in Figure 39 on the
following page. More detailed information regarding these studies
can be obtained by contacting the Office of Solid Waste.
231
-------
-------
APPENDICES
-------
code. This was done in order to sample some industries more
heavily than others. The three SIC groups are presented below
The Three SIC Strata
Group 1
Group 2
1.
2.
3.
4.
5.
6.
7.
8.
2491
2812
2816
2861
2892
2992
3351
3356
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
0700-0799
2600-2699
2860-2869
(except 2861)
2879
2911
3310-3319
3341
3471
3479
3711
3714
4953
1.
Group 3
All remaining
SIC codes
not
accounted for
in strata 1
and 2
These strata were sampled at varying rates, depending
on whether they were found on the Notifier file only; also on
the Part A file, but not confirmed as having TSD facilities by
the telephone survey, or on the Part A file and confirmed as
having TSD facilities by the telephone survey. Thus, there were
nine major generator strata. Each of these strata were further
stratified by three other variables: four digit primary SIC
code; within SIC code, by region; within region, by a size
measure (number of employees or a quantity measure depending on
the stratum) when available.
A.2
Sampling Method
Most lists were systematically sampled in order to
achieve increased precision for estimates of facility percentage
characteristics (e.g., what percent of landfills are lined?).
Oversampling of selected strata was not used, except for certain
A-2
-------
APPENDIX A
SAMPLE DESIGN AND STATISTICAL RELIABILITY
This study actually involved 10 different sample designs,
one for each of the following type of hazardous waste handlers:
• Generators;
• Injection Wells;
• Landfills;
• Land Treatment;
• Surface Impoundment;
• Waste Pile;
• Incinerators;
• Storage Containers;
• Storage Tanks; and
• Treatment Tanks.
A brief summary of the sample design features is
presented below. A more detailed description is provided in
Westat's separate report to EPA.
A.I Stratification
To enhance the representativeness of the samples of
TSD facilities, they were stratified by several variables. These
stratification variables were: EPA region, whether hazardous
waste was also generated at the site at which it was being managed,
and, when available, by an approximate measure of size (quantity
of hazardous waste managed).
For the sample of generators, the population of genera-
tors was divided into three sets of strata based on primary SIC
A-l
-------
The value "2" in the equation was a rounding of "1.96," the
value obtained from a normal distribution table associated with
achieving 95 percent confidence. The proportion p was chosen to
take on a value of .5 in order to obtain a conservative estimate
of precision. That is, the confidence limits may be wider than
needed for 95 percent confidence. If the estimate p were .3
rather than .5, the width of the confidence interval would
actually be 92 percent of that presented in this report. Thus,
a data analyst may find it beneficial to compute the actual
confidence limits rather than use the general limits for percentages
presented in this report.
In the cases of the Tank and TSD Questionnaires, an
additional factor must be multiplied times the value of b
expressed above to obtain limits of precision. This is necessary
because the sites sampled to receive these questionnaires were
selected with varying probabilities. The factors are the square
root of the design effects, and allow the limits b1 to be obtained
for tanks on the TSD General Questionnaire. Specifically, the
limits are:
Limits for Tanks and TSD General
Tanks b' = 1.06b
TSD General b1 = 1.39b
where b is obtained as indicated previously.
The bounds obtained for generators were computed based
on a stratified estimate of variance. A description is provided
in Westat's more detailed report.
A-4
-------
generator SIC codes of special regulatory interest to EPA. The
size of the selected sample was designed to be larger than the
expected number of completes by a sufficient margin to allow for
nonresponse and ineligibles. A pre-survey telephone verification
effort was used to more accurately anticipate nonreponse and
ineligibility rates.
A. 3 Precision
The sample design produced estimates with precision
limits of from +5 percent to +11 percent at the 95 percent
confidence level for percentages of facilities with particular
characteristics for each of the various treatment, storage, and
disposal (TSD) facility samples. Variance estimates for specific
questionnaire items were made by half-sample and paired-difference
techniques, since standard statistical programs (e.g., SAS,
SPSS) assume simple random samples and neglect the payoff of the
finite population correction factor and stratification. Particu-
larly for the smaller universes (such as land treatment facilities
and injection wells), the finite population correction adjustment
makes an important contribution to the reduction of the estimated
variance.
The precision bounds for estimated percentages obtained
with 95 percent confidence (from +5% to +11% for the individual
process types, +2.4% for generators, and +3.0% for the TSD General
Questionnaire) were established using the following expression.
In general, the bound b was determined by
where
b - (2) 100 l/T~- n
I/ N
n - the number of respondents to a questionnaire
^ ~ "P J~l Łi /^i O 4- i t-n —* 4- x^. .^3 . , ,' _f_ .-i .
estimated .size for the universe in question
P - the proportion of sites with a particular
characteristic, taken to be 5
q = 1 - p.
A-3
-------
Table A-l. Summary of Statistical Aspects of the TSD and Generator Surveys
Precision at th?? 95%
Confidence Level
Average
Sample Stratified by Weiaht*
Generators SIC Group-4 Digit SIC 6,76
Code-Region-Number of
Employees at Location
l'C;n r*onrj»-;» ') 1 1 (\
Injection Wells Region-Generator- 1.20
Quantity
Landfills Region-Generator- 2.52
Quantity
Land Treatment Region-Generator 1.88
Quantity
1 Surface Impoundments Region-Generator 5.23
^ Quantity
Waste Piles Region-Generator 2.39
Quantity
Incinerators Region-Generator- 1.92
Quantity
Storage Containers Region-Generator- 18.73
Quantity
Storage Tanks Region-Generator- 6.11
Quantity
Treatment Tanks Region-Generator- 5.03
Est i ma ted
Ł; ize
1981
14,098
4ft 1 ft
87
199
70
764
174
240
3,577
1,428
609
Sample Size
(Number of For Foi ^
Respondents "Percentage General
Active in of Facilities" "Quantity"
1981) Estimates*** Estimate
2,084 + 2.4% ± 80%
i Af\~) 4- ~\ n* 4- dQa
73 + 4.8% + 16%
79 + 8.7% + 54%
37 + 11.3% + 32%
145 + 7.5% + 48%(Treated)
+ 64% (Stored)
+ 17 2% (Disposed)
73 + 8.9% + 52%
117** + 6.6% + 28%
191 + 7.2% + 50%
233 + 6.4% + 112%
121 ^ 8.6% + 86%
Quantity
* The weight was essentially constant within a sample except for the tanks, generators, and the TSD general components.
The range of weights for these samples were: tanks, 4.1 to 10.6; generators, 1 to 19.0; and TSD general, 1.1 to 25.0.
For incinerators, there were eight additional respondents who were not active in 1981 but were eligible to complete
the OURS*- ionna i rp>.
the questionnaire.
*** Based on an estimated proportion of .5.
-------
A. 4 Weighting
Base weights were determined as the inverse of probabil-
ities of selection. These weights were adjusted for nonresponse
within broad industry groups for each sample. Thus, differential
cooperation rates by different industries were taken into account.
The base weights for Tanks and the TSD General Questionnaire
reflect the fact that the sampled sites had multiple opportunities
of selection if they appeared on more than one sample frame
list. For example, a site found on the incinerator, landfill,
and waste pile sample frames had three different opportunities
to be selected and receive a TSD General Questionnaire. Finally,
minor weight adjustments were made to reconcile national estimates
of the total number of process facilities obtained from the
individual process type component samples with those obtained
from the TSD General Questionnaire.
Table A-l summarizes the major sampling information
associated with this survey. This includes precision levels for
various estimates, and types of estimates and the average sampling
weight associated with each questionnaire sample. These weights
are essentially constant except for the Tank, Generator, and TSD
General Questionnaires, as noted in the footnote.
A-5
-------
B.I.I Questionnaire Preparation and Pretest
The two original annual survey forms were expanded
into a Generator Questionnaire; a Treatment, Storage and
Disposal General Questionnaire; and eight process-specific
Management Technology Questionnaires to serve Regulatory Impact
Analysis (RIA) data needs. With ongoing interaction with each
RIA group, the questionnaires went through numerous draft
versions. The questionnaires were submitted to OMB for review
and received clearance just before the mailout in September.
In June of 1982, questionnaires were mailed to a
pretest group including associations and other groups concerned
with hazardous waste management. Returned questionnaires were
examined by EPA staff and by Westat questionnaire designers to
determine the efficiency of the instrument in gathering data.
Each of the 10 questionnaires underwent many revisions in the
month of August.
B.I.2 Mailout
The mailout was a complicated time-consuming task
because of the large number of packages to be mailed, because
each TSD facility received a "custom" package with different
components, and because the packages were sent certified mail.
Approximately 15 full-time equivalents were required to complete
the mailout in two weeks. The mailout was completed September
17, 1982. Packages containing Generator Questionnaires were
mailed to a sample of 10,667 generator installations. The sample
was drawn from EPA's computer file of Notification forms, which
had been submitted by firms that indicated they were or would be
B-2
-------
APPENDIX B
FIELD REPORT
B.I Field Period
This survey of hazardous waste handlers was conducted
by Westat, Inc., for the Environmental Protection Agency's
Office of Solid Waste (OSW) in the fall of 1982. It is the most
extensive data collection effort on hazardous waste management
practices to date. The survey was national in scope, covering
the fifty states as well as U.S. Territories. Preparatory work
for the survey actually began in 1981, when Westat began working
with EPA/OSW on the development of annual survey forms for
hazardous waste generators, and for hazardous waste treaters,
storers and disposers. While these forms were being prepared,
the need became apparent for additional information on treatment,
storage and disposal technologies and for more detailed infor-
mation about hazardous waste generators and hazardous waste
treaters, storers and disposers.
Also in preparation for the mail survey, Westat
conducted telephone status verification interviews with approxi-
mately 9,000 treatment, storage and disposal facilities that
filed RCRA Part A Applications stating that they intended to
manage hazardous waste. These telephone interviews were
conducted in two phases during the spring and summer of 1982 in
an attempt to verify the processes used by the facilities to
manage hazardous waste in 1981. The results provided a sampling
frame for the selection of facilities to receive the RIA mail
survey questionnaire, and are summarized in the "Report on the
Telephone Verification Survey of Hazardous Treatment, Storage,
and Disposal Facilities Regulated Under RCRA in 1981," produced
by Westat in November, 1982.
B-l
-------
Two series of individualized followup letters
were sent to TSD's and to generators (see Section
B.2.2.1); and
Technical assistance was provided to respondents
by Westat and EPA telephone hotlines.
Returns were tracked by an automated receipt control
system. Each day, certified cards and returned questionnaires
were keyed into the system. Reports were produced by the system
on a weekly basis. Since the system tracked returns by ques-
tionnaire type as well as by site, we could follow up on cases
where not all questionnaires were returned as well as on cases
where no response was received at all. In addition, the system
was used to produce summary statistics that were submitted
weekly to EPA. Examples of these reports appear as Exhibits B-l
through B-3.
Figure B-l presents the number of responses by month
for TSD facilities. Figure B-2 presents the number of responses,
by month, for generators. As can be seen from these figures,
the followup efforts, which were conducted after the end of
November, yielded a substantial number of responses, especially
for TSD facilities, where high response rates were the most
critical because of small sample sizes for certain sample
strata.
B-4
-------
generators of hazardous waste, but that did not file Part A
Applications with EPA to treat, store, or dispose of hazardous
waste on site. In addition, 2,599 packages containing Treatment,
Storage, and Disposal General Questionnaires and one or more of
nine process-specific management technology component question-
naires, including the Generator Questionnaire, were mailed to
facilities sampled from Westat's Part A/Verification computer
file. (This computer file was developed from the data collected
in the Telephone Verification Survey described in Section B.I,
above.)
Because the packages were sent by certified mail, we
were able to determine when the package was received and who
received it. The letter which accompanied the questionnaires
stated the RCRA regulation that respondents had 45 days from the
date they received the questionnaire in which to reply. By
tracking the certified card date on Westat's Automated Receipt
Control System (discussed below), it was possible to followup
each respondent on the 45th day after the package was received.
The letter also advised respondents that it was possible to
obtain official time extensions. Numerous time extensions were
requested and were granted by EPA (see Section B.4.3).
Based on return rate and sample sizes needed for analy-
sis, decisions to extend the cut-off date for returns extended
the date from November 15 to December 15, 1982 and finally to
(postmarked) February 1, 1983. To increase response rate during
this period, several steps were taken:
• Followup calls were made to TSD's three days
after 45-day limit;
B-3
-------
Exhibit B-2. Example of Automated Receipt Control System Weekly Report — TSD Facility
Return Status by Questionnaire Type
*IA Mail Surv«y R«c«ipt Control Stetui R«port:
TSD Co»pon«nts
12/23/82
i
en
*
facility Stetus
- Applicability not d«t«r«in«d. . . .
- Component not rtrnd •/ raspons*.
TOTAL IN SAMPLE
IW
114
111
8*
59
25
0
0
4
20
3
1
2
3
1
115
Qu«stion»ir«
LF
202
197
146
70
76
0
0
7
38
6
5
1
5
0
202
LT
79
78
63
24
39
0
1
10
4
4
0
1
1
80
SI
326
325
278
146
130
0
4
33
10
9
1
1
1
327
UP
242
235
189
70
116
0
i
11
25
1C
10
0
7
1
243
Component
IN
264
258
202
128
72
1
1
a
38
10
9
1
6
1
265
CN
420
412
352
301
51
0
0
4
53
3
3
0
8
3
423
TK
842
823
696
491
201
0
4
13
98
16
15
1
19
5
847
GN
550
541
486
390
94
0
2
4
46
5
5
0
9
3
553
1
TG I
+
1
25461
i
24961
i
20981
13961
686|
11
15|
1
461
298!
54|
521
21
501
111
1
1
2557|
1
4
-------
Exhibit B-l.
Centre! Syste. Weekly Keport - TSD Facility
KIA Hail Survey Receipt Control Status Report
ISO Facilities (All components combined)
12/23/82
Facility Status
*
- All components applicable.......
- Sore component* not applicable..
- No components applicable........
TOTAL IN SAMPLE
»
1
i a a
i a 7
ICC
1 5.2
84
49
19
3
1
27
4
4
o
189
2
117
771
? A ft
1 If.
91
4 1
5
•»
1 e
Q
321
3
5 1 Q
1 ? 4
71
? A
j a
257
4
368
1 A Q
65
1 I
370
EPA
5
566
553
466
126
68
10
1 5
1
569
Region
6
401
391
336
325
to
72
8
43
3
1 0
4
405
7
83
82
72
72
40
20
1 2
0
1
8
1
1
1
0
83
8
61
54
45
44
24
15
5
1
0
7
2
2
0
7
0
61
9
241
234
179
173
78
58
37
6
16
33
6
6
0
7
2
243
+ -
1
10 |
4 -
1
59|
1
56|
1
43|
I
«2|
1
18|
61
181
1
1 1
1
11
1
12|
1
01
1
0|
01
1
3|
1
01
1
4__
1
59|
1
«
U.S.
Total
2546
2496
2114
2070
1073
591
400
44
44
296
42
40
2
50
11
2557
to
I
en
-------
Figure B""1
TSD Questionnaire returns by end of each month from
September 17, 1983 mailout. RCRA required return date:
November 15. 2,599 packages mailed.
2,500 -|
2,348
2,277
2,000 -
1,824
1,500-
1,150
1,000-
44*
500 -
82%
$0%
Oct
Nov.
De
Jan.
Mar.
(Final
-------
Exhibit B-3. Example of Automated Receipt Control System Weekly Report — Generation
Installation Return Status by Region
1IA Mail Survey Receipt Control Status Report:
Off-Site Generators
12/23/82
Facility Status
o Applicability not determined....
o Component not rtrnd «/ response.
o Facility closed.....
TOTAL IN SAMPLE
t
1
862
R 1 5
593
224
369
0
0
1 5
194
1 1
13
o
47
5
867
t
2
1081
0*4
639
236
399
2
0
21
280
1 4
1 4
o
127
1 0
1091
3
OKI
638
203
434
1
0
31
197
7
7
o
an
i
954
4
1286
I J J C
2*3
760
0
0
4 8
110
1 ft
1 ft
o
ft 1
1291
EPA
5
34 x i
y » f 7
10X4
ft 5 7
1200
7
0
39
407
i 7
0
I IX
2482
Region
6
1 <0 4
1 47 fl
1 2 fl 5
< a 4
1098
3
0
•> c
let
7
7
n
jx
1509
7
47 c
L X. «.
ion
140
240
0
0
74
i n
476
8
1 1ft
•ii>
j 7*
c a
222
1
0
4
4 fl
•i
339
9
1 9 C 7
t 046
140
AQS
0
0
1 pQ
1 J
1 5
i n
1303
10
140
? Q 1
7 A
214
1
0
c
L L
c.
n
355
t 4
U.S.
Totrl
11 n *o ?
> 1 fi 7
C A 1 1
15
c
? 1 ft
17??
IQ Q
In
c in
10667
-------
B.2 Production Statistics
Production statistics were kept using the automated
(computerized) receipt control system described in Section B.I.2
and below in Section B.2.2. In general, the response rate for
each of the questionnaire samples was between 85 percent and 93
percent. The eligibility rates for the samples ranged between
19 percent and 73 percent.
B. 2 .1 Definitions of Return Status Groups
The return status for each mailed out questionnaire
was recorded on the automated receipt control system. The return
status was initialized (or assigned) with the status code INIT
at the time of mailout, which indicated that the questionnaire
package had been mailed out but no response had been received.
Since the questionnaires were all mailed using certified mail
service, when the certified mail card was returned by the facility
the return status for the facility was updated to CERT, to indicate
positive evidence of contact with the facility. Returned
questionnaires were scan-edited as they were received, and given
return statuses of CMPL, indicating a completed, eligible
response and INEL, indicating a returned ineligible response. A
complete list of return status codes and their definitions is
provided below in Figure B-3.
B.2.2 Response Rate for Generator Questionnaires
Installations sampled from the HWDMS Notifier file
were assumed to be generators of hazardous waste who managed
their wastes off site, since these installations had notified
EPA of their intent to generate hazardous waste but had not
B-10
-------
Figure B-2
Notifier Generator returns by end of each month from
September 1 7, 1983 mailout. RCRA required return date:
November 15. 10,667 packages mailed.
10,000 -
8,000-
6,000-
4,000 -
2,000-
0 -
9,361
8,398
7,762
5,607
53%
73*
8,043
m
*.
*•
Oct- Nov. Dec. Jan.
87.8%
-
Mar.
(Final)
B-9
-------
submitted Part A permit applications to treat, store or dispose
of hazardous waste. Accordingly, these installations were sent
a questionnaire package containing a Generator Questionnaire,
plus the introductory materials and appendices for the survey.
Generators sampled from the Part A/Telephone Verification file
were known (due to their responses in the Telephone Verification
or expected (because they had filed Part A applications) to be
generators of hazardous waste who managed at least some waste on
site. They were sent Treatment, Storage and Disposal General
Questionnaires, as well as Generator Questionnaires. (Some Part
A facilities may also have been sampled to receive one or more
process-specific management technology questionnaires, as well
as the Generator Questionnaire.)
Followup and processing strategies for the Notifier
generators were not identical to those for the Part A/
Verification generators, so the production statistics for the
two groups will be discussed separately first, and then as
combined Generator Sample. Table B-l shows the return status
for Notifier generators, for generators selected from the Part
A/Telephone Verification file, and for all sampled generators.
B.2.2.1 Return Rate for Generators Selected from the
Notifier File
A total of 10,667 installations were sampled from the
Notifier file to receive Generator Questionnaires. Of these,
responses were received from 9,361 installations. Responses for
some installations consisted of a determination of ineligibility
for the survey, based on telephone calls or other evidence (such
as correspondence), rather than a returned questionnaire.
Completed (eligible) responses were received from 1,821
installations, which represents 17.1 percent of the sampled
B-12
-------
Figure B-3 Return Status Codes and Definit-inng
Return
Code
CMPL
INEL
Return Status
Complete
Ineligible
Non-Response Categories:
INIT No contact
NDEL
CERT
PEND
MISS
REFS
CLOS
OTH
Not deliverable
Certified Mail
Time extension
granted
Questionnaire
missing from
respondent's
returned package
Refused
Closed
Other
nonresponse
Definition
Completed, eligible questionnaire received
(facility was engaged in regulated activity
covered by the questionnaire)
Ineligible questionnaire received (facility
was not engaged in regulated activity cov-
ered by the questionnaire)
Questionnaire mailed, but certified mail
card was not returned
Questionnaire mailout package could not be
delivered by the Post Office (repeated
attempts using alternative addresses were
made)
Certified mail card received. As a final
return status, this means there is positive
evidence that facility received the question-
naire, but did not respond.
A time extension beyond the allowed 45 days
was granted. As a final return status, this
means facility did not respond after request-
ing and being granted a time extension.
Facility returned at least one, but not all
of the questionnaires it was sent. This
return status applied only to TSD facili-
ties, which always received two or more
questionnaires.
Facility refused to complete questionnaire.
Facility closed after January 1, 1981.
(Records or personnel were not available.)
(The one Waste Pile Questionnaire and one
TSD Questionnaire assigned to this category
are included, for statistical purposes, in
the CERT nonreponses.)
B-ll
-------
10,667 Notifier file installations. The response rate for Notifier
installations was 87.8 percent.1 Table B-l, Column 1, presents
the return status for the Notifier generator sample.
Nonrespondents from the Notifier generator sample were
followed up twice by mail. The first recontact was a letter
mailed out after the initial 45 days of permitted response time
had elapsed. The second followup contract was a letter mailed
out approximately 90 days after the initial mailing.
The nonresponding Notifier generator cases were reviewed
after the end of the coding period, but before the receipt control
file had been closed out. Using selective telephone calls and
other evidence (such as installation names), a number of
nonresponding Notifier generators were judged to be "ineligibles"
(i.e., non-generators). This "cleansed" the receipt control
file of at least the most obvious "protective filings" (e.g., by
gas stations and shoe stores).
By January of 1983 (in the fourth month of the field
period) a high response rate for Notifier generators had been
obtained. In addition, a higher-than-expected number of eligible
(regulated) generators had responded. Therefore, in mid-January,
1983, the coding operation for this type of generator was closed.
The receipt control counts for Notifier generators therefore
include 111 completed (eligible) Generator Questionnaires and
computational formula used to calculate response rate (R) is
R = E+I/E+I+NR where E is the number of eligible responses, I is
the number of ineligible responses, and NR is the number of
nonrespondents. The definitional formula from which the
computational formula is derived is R = E/(E+NR(E/I+E)).
B-14
-------
Table B-l. Generator Questionnaire return status
Return status for generators
selected from the Notified File
Return status for generators
selected from the Part A/
Telephone Verification File
Return status for all
sampled generators
Mturn_Status Number Percent ^^ ^^
Total responses (9 351) /P7 CM /ri^
vy,JDi; (B/.8) (516) (93.3)
ComPlete 1,821(D 17.1 377(3) 68 -
'"edible 7,540(2) 70.7 139(4) „ ,
Total Non-responses (i,306) (12.2) (37) {6[7)
No contact 57 Q>5 Q
Not deliverable 239 22
Certified card received 848 7 ." 9 13 1"4
Time extension granted 6 01 9
Refused u Q] ± 2 J.4
Closed 145 « 1-4
•H 6 1.1
TDtal Maiied Out: 10,667 100.0 553 100.0
Number
(9,877)
2,198(5)
7,679(6)
(1,343)
57
247
861
8
19
151
11,220
— , _
(1)
(2)
(3)
(4)
(5)
(6
Percent
(88.0)
19.6
64.8
(12.2)
0.5
2.2
7.9
0.1
0.1
1.4
100.0
111
741
. §1
(3d
i
M
U)
-------
reminder letter was mailed out to nonrespondents after the
initial 45 days of permitted response time had elapsed. A
second followup contact consisted of a telephone reminder call
to the facility. Telephone followup began in early December,
approximately 80 days after mailout of the questionnaires. In
addition, for cases where no certified mail cards were returned
and where the questionnaire package was not returned marked
undeliverable by the Post Office, the Post Office was requested
to trace the package to verify the delivery status. (All
deliveries were verified for this sample.)
The field period (during which completed responses
were accepted for coding) for all TSD facilities (including the
Part A/Verification generators) was extended from November 15 to
December 15 of 1982 and finally to February 1, 1983. The Part
A/Verification sample facilities in general, and therefore the
Part A/Verification generators, received more followup attention
than the Notifier generators, because several component subsamples
of the TSD sample were numerically small (even though they were
proportionately large within their strata). In addition, TSD
facilities requested and were granted time extensions well into
January of 1983 and sometimes beyond.
Because the coding operation for TSD facilities remained
open until early February, 1983, almost all of the lateresponding
Part A/Verification generators questionnaires were coded. Of
the 377 complete (eligible) Part A/Verification generators
responding, three (or .08 %) were received too late for coding.
Of the 139 ineligible Part A/Verification generators, two (or
1.4 %) were received after the coding operation closed. (Late
respondents were not factored into population estimates.
However, their small numbers would have little impact on the
result.)
B-16
-------
741 ineligible Generator Questionnaires which were not coded.
The 111 uncoded completed Generator Questionnaires represent 6.1
percent of the 1,821 completed Notifier generators received,
while the 741 uncoded ineligible Generator Questionnaires
represent 9.8 percent of the 7,540 ineligible generators.
B'2-2'2 Return Rate for Generators Selected From the Part
A/Telephone Verification File
A total of 553 facilities were sampled from the Part
A/Telephone Verification file to receive Generator Questionnaires.
(This portion of the generator sample also received Treatment,
Storage and Disposal General Questionnaires, since they had been
identified as TSDF's during the telephone verification and/or
had submitted Part A permit applications.) Complete (eligible)
Generator Questionnaire responses were received from 377 facilities,
which represents 68.2 percent of the total sample. Despite
the telephone verification effort, ineligible Generator Question-
naire responses were received from 139 facilities or 25.1 percent
of the total sample. In all, Generator Questionnaire responses
were received from 516 of the 553 sampled facilities, resulting
in a response rate for the Part A/Verification generators of
93.3 percent.2 Table B-l, Column 2, shows the return status for
the Part A/ Verification generator sample.
Nonrespondents from the Part A/Verification generator
sample were followed up as part of the more extensive followup
efforts directed at TSD facilities. First, an individualized
H PT/F™P at fo™ula urd t0 calculate response rate (R) is
R E+I/E+I+NR where E is the number of eligible responses, I is
the number of ineligible responses, and NR is the number of
Mnn^P? en^'- T^ definiti°nal formula from which the computa-
tional formula is derived is R - E/(E+NR(E/I+E)). ^mputa
B-15
-------
Table B.2. Eligibility Rate Among Respondents From the
Notifier File and the Part A/Verification File
Eligibility Status Number Percent
Generators selected from the
Notifier File:
Total respondents (9,361) (100.0)
Completes (eligibles) 1,821 19.5
Ineligibles 7,540 80.5
Generators selected from the
Part A/Verification File:
Total respondents (516) (100.0)
Completes (eligibles) 377 73.1
Ineligibles 139 26.9
All generators in the
combined sample:
Total respondents (9,877) (100.0)
Completes (eligibles) 2,198 22.3
Ineligibles 7,679 77.7
B-18
-------
B-2-2-3 Return Rate for Combined Notifier and Part- A/
Verification Generator Sample
A total of 11,220 installations and facilities were
sampled from the Notifier and Part A/Verification files to respond
to the Generator Questionnaire, representing hazardous waste
generators nationwide. Responses were received from 9,877
installations and facilities, of which 2,198 were complete
(eligible) responses and 7,679 were ineligible responses. The
response rate for the combined sample of generators is 88 0
percent. 3 Table B_1( Column 3< presenfcs the refcurn ^^ ^
the combined Notifier and Part A/Verification sample.
The response rate for Notifier generators was slightly
lower (87.8%) than the response rate for Part A/Verification
generators (93.3%). The Part A/Verification generators, due to
the fact that they were accompanied by TSD Questionnaires, were
followed up more vigorously than Notifier generators (by telephone
rather than mail only), and the field period was held open longer.
The most apparent difference oetween the Notifier and
Part A/Verification sample response outcomes may be observed in
the Eligibility Rate of the responses from the two groups.4 (See
Table B-2.) Among the 9,361 Notifier file respondents, 1,821
3
nonrespondents, The d^finiH—^i f^~r, . i „ P f number ot
tational for^a is derived Is R' = K
KK = T./T.4-T i!Sed to.calculate the Eligibility Rate (ER) is
the'number ^ne^igJb^re^sesf eU9ible ^^^ ^ * iS
B-17
-------
B. 2. 3 Response Rates for Treatment, Storage and Disposal
Facilities
Facilities sampled from the Part A/Telephone Verification
file were each sent a Treatment, Storage and Disposal General
Questionnaire, as well as at least one process-specific management
technology questionnaire or a Generator Questionnaire. The titles
of the nine questionnaires that could have been included with
the TSD Questionnaire are listed in Figure B-4. (See Appendix A
for a complete description of sampling strategies.) The maximum
number of questionnaires received by any one facility was five.
The response rate and return status of the Generator Questionnaire
component for TSD facilities that received Generator Questionnaires
is discussed in Section B.2.2.3 and in Tables B-l and B-2. In
this section, the response rates for the Treatment, Storage and
Disposal General Questionnaire, as well as for the process-
specific management technology questionnaires, will be discussed.
Table B-3 presents the return status for the TSD
questionnaire and the eight process-specific management technology
questionnaires. Table B-4 presents the Eligibility Rates for the
TSD and technology questionnaires.
B. 2. 3 .1 Response Rate for the Treatment, Storage and Disposal
General Questionnaire
A sample of 2,599 facilities was drawn to receive
Treatment, Storage and Disposal General Questionnaires. Thes-2
facilities were selected from the Part A/Telephone Verification
file, and had indicated during the telephone verification survey
B-20
-------
were eligible, for an eligibility rate of 19.5 percent. In com-
parison, the table indicates that the eligibility rate for Part
A/Verification generators is almost four times greater, at 73.1
percent. (Among 516 Part A/Verification generator respondents,
377 were eligible.) The higher eligibility rate among the Part
A/Verification generators is due primarily to telephone
verification of the universe list prior to sample selection.
A second factor that could account for this higher
observed eligibility rate is the more extensive information
requirements contained in EPA's Part A permit application forms.
Notification forms simply required location and contact
information, an indication of whether hazardous wastes were
generated, treated/stored/disposed, transported, or injected at
the site, and list of hazardous wastes that would be generated.
Part A permit applications, however, required extensive site
descriptions, including, but not limited to, maps and diagrams,
descriptions of the specific processes used to treat/store/dispose
of hazardous wastes at the site, and breakdowns of waste quantities
managed by the processes in which they were to be managed. Given
the greater work load involved, it seems reasonable to expect
that firms would examine their operations more carefully to
determine whether hazardous wastes were actually present before
submitting their applications. Thus, it is not surprising that
a higher percentage of the firms submitting Part A applications
would be observed to actually manage or generate hazardous
wastes than the firms that were only required to submit
notification forms.
B-19
-------
Table B-3. Return Status for TSD and Component Questionnaires
Return Status
TOTAL RESPONSES N
Complete N
Ineligible N
TOTAL N
NON-RESPONSES (%)
No Contact N
(*)
Not Delivered N
(%)
Cert Card Rec'd N
Time Extended N
Quex NR N
Refused N
(%)
Closed N
.
TOTAL MAILED N
Injection !
Wells
103
(89.6)
73
(63.5)
30 \
(26.1)
_
| j
12
(10.4)
1
(0.9)
3
(2.6)
5
(4.3)
0
0
3
(2.6)
0
115
(100. 0)
Landfills
172
(85.1)
79
(39.1)
93
(46.0)
30
(14.9)
0
5
(2.5)
15
(7.4)
3
(1.5)
1
(0.5)
1
(0.5)
5
(2.5)
202
(100.0)
i
Land
Treatment
99
(81. li
37
(30.3)
62
(50.8)
23
(18.9)
16
(13.1)
1
(O.E)
1
(0.8)
1
(0.8)
0
0
4
(3. i)
122
(100. D)
Surface
Impound-
ments
298
(91.1)
145
(44.3)
153
(46.8)
29
(8.9)
1
(0.3)
1
(0.3)
10
(3.1)
2
(0.6)
2
(0.6)
4
; (1.2)
9
(2.8)
! 327
(100.0)
Waste
Piles
215
(88.5)
73
(30.0)
142
(58.4)
28
(11.5)
1
(0.4)
7
(2.9)
8
(3.3)
(0.4)
\ 1
! (0.4)
1
(0.4)
9
(3.7)
i 243
(100.0)
i
Incin- l
erators i
239
(90.2)
129(D
(48.7)
110
(41.5)
26
(9.8)
0
6
(2.3!
5
(1.9)
(0.7)
0
5
(1.9)
8
(3.0)
265
( 100.0)
Con-
tainers
389
(92.0)
195 <2>
(46.1)
194
(45.9)
34
(8.0)
3
(0.7)
8
(1-9)
; 7
•; 1 . 7 )
1
1 (0.2)
0
12
' (2.8)
3
(0.7)
423
(100.0)
TSD
Tanks
General
772 2,353
(91.1) (90.5)
288<4' 1,479(3)
(34.0) (56.9)
484 874
(57.1) (33.6)
75 246
(8.9) , (9.5)
4 23
! (0.5) ' (0.9)
i 19 ; 49
(2.2) l (1.9)
! 31 78
(3.7) , (3.0)
i 0 12
(0.5)
; 1 3
(0.1) , (0.1)
5 ! 30
(0.6) ! (1.2)
15 : 51
(1.8) i (2.0)
847 , 2599
(100.0) (100.0)
(1) Includes 4 unprocessed completed Incinerator Questionnaire responses received after the field period had ended.
(2) Includes 4 unprocessed completed Container Questionnaire responses received after the field period had ended.
(3) Includes 17 unprocessed completed TSD General Questionnaire responses received after the field period had ended.
(4) Includes 5 unprocessed completed Tank Questionnaire responses received after the field period had ended.
ra) Collapsed categories: The counts of Partial Completes were included in the counts of Completes. (There was one Partial
Complete Incinerator Questionnaire response and one Partial Complete TSD General Questionnaire response.) The counts of
"Other" types of nonrespondents were included in Certification Card Received counts. (There was one "Other" Waste Pile
Questionnaire nonresponse and one "Other" TSD General Questionnaire nonresponse.
(b) Anonymous responses: Approximately ten facilities obliterated or removed all survey identifiers from their response
forms, thus making it impossible to log their responses. Only five of these facility responses were useable due to re-
quirements involving weighting. For the purposes of this table, these five responses were added into the counts of Com-
pletes, and subtracted out of the counts of Certified Card Received. There were two useable anonymous Landfill Question-
naire responses, three useable anonymous Land Treatment responses, and five useable anonymous TSD General Responses. The
unuseable anonymous responses were not included among the Complete responses, and are still included in some category of
nonresponse.
B-22
-------
Figure B-4 Titles of the Nine Questionnaires that Were
Mailed with the Treatment, Storage, and Disposal
Questionnaire
1. Hazardous Waste Container Questionnaire
2. Hazardous Waste Generator Questionnaire
3. Hazardous Waste Incinerator Questionnaire
4. Hazardous Waste Landfill Questionnaire
5. Hazardous Waste Land Treatment Questionnaire
6. Hazardous Waste Surface Impoundment
Questionnaire
7. Hazardous Waste Tank Questionnaire
8. Hazardous Waste Underground Injection Well
Questionnaire
9. Hazardous Waste Pile Questionnaire
B-21
-------
that they were TSD facilities during 1981.5 Responses were received
from 2,353 facilities for a response rate of 90.5 percent. Of
the 2,353 responses, almost two thirds (62.9%, or 1,479 facili-
ties) were eligible responses. The remaining 874 responses were
from ineligible (i.e. facilities that did not treat, store or
dispose of hazardous wastes in processes regulated under RCRA
during 1981) facilities. Of the 246 TSD Questionnaires that were
not; returned, 49 represented facilities to whom it was not
possible to mail the questionnaire package; 51 were for facilities
that, reported that they were closed after January 1981, but were
unable to answer the questionnaires; and 78 facilities had not
responded by the end of the survey. Together, these three types
of nonresponse make up nearly two thirds of the total nonresponse
for the TSD Questionnaire. Thirty facilities (or 1.2% of the
total sample) refused to complete the TSD Questionnaire.
B« 2.3.2 Response Rate for the Underground Injection We 1Is
Questionnaire
Of the 115 facilities that were mailed Injection Wells
Questionnaires, responses were received from 103 for a response
ratr of 89.6 percent. Over two thirds of the responses (70.9%,
or 73 questionnaires) were completed questionnaires from facilities
with regulated injection wells. The remaining 30 responses were
ineligible (that, is, from facilities that did not inject
hazardous waste into underground injection wells during 1981).
-'The TSD sample actually comprised the nine technology questionnaire
Sciirsples, The original TSD sample contained 2,557 sampled
facilities, but was increased by a supplementary sample of 42
facilities in the Land Treatment sample.
B-24
-------
Table B-4. Eligibility Rates for the Treatment, Storage, and
Disposal Questionnaire and the Technology Question-
naires
Questionnaire
TSD General
Injection Wells
Landfills
Land Treatment
Surface Impoundments
Waste Piles
Incinerators
Containers
Tanks
Total
Responses
2,353
103
172
99
298
215
239
389
772
Total
Eligible
(Complete )
Responses
1,479
73
79
37
145
73
129
195
288
Eligibility
Rate*
62.9%
70.9%
45.9%
37.4%
48.7%
34.0%
54.0%
50.1%
37.3%
* The Eligibility Rate was calculated by dividing the number of
eligible responses by the total number of responses.
B-23
-------
of hazardous waste through land treatment processes during 1981.
The remaining 62 questionnaires were from facilities that did not
dispose of hazardous waste through land treatment processes
during 1981. Indications of land treatment on Part A permit
applications appear to have been heavily inflated due to applicant
misidentification. Land application was often confused with
other disposal and treatment technologies (such as landfilling,
open burning and surface impounding). The 16 facilities where no
contact was documented accounted for most of the nonresponses to
the Land Treatment Questionnaire. In addition, one questionnaire
package was not deliverable. Two facilities had not responded by
the end of the study period, although one had requested a time
extension. Four of the sampled facilities were closed, and
unable to answer the questionnaires. There were no direct
refusals among the sampled land treatment facilities.
B. 2.3.5 Response Rate for the Surface Impoundment
Questionnaire
Of the 327 facilities that were sent Surface Impoundment
Questionnaires, responses were received from 298, for a response
rate of 91.1 percent. Approximately half (48.7%, or 145
questionnaires) of the responses were from facilities that used
surface impoundments for treatment, storage or disposal of
hazardous waste during 1981. The remaining 153 responses to the
Surface Impoundment Questionnaire were returned by facilities
that did not use surface impoundments for hazardous waste during
1981. There were 29 nonresponding surface impoundment facilities
at the end of the survey period. Only two facilities (less than
1% of the sample) could not be contacted. Fourteen facilities
had not responded by the end of the survey, representing 4.3
percent of the sample. Nine of the surface impoundment facilities
(almost 3%) had closed and were unable to answer the questionnaires
There were four refusals among the surface impoundment facilities.
B-26
-------
Of the twelve nonresponses, four could not be reached by mail
(one was a "no contact", and three were "not delivered"), and
five were simple nonresponses. Three facilities (2.6% of the
sample) refused to answer the Injection Wells Questionnaire.
B. 2 . 3 . 3 Response Rate for the Landfill Questionnaire
Of the 202 facilities that were mailed Landfill
Questionnaires, responses were received from 172, for a response
rate of 85.1 percent. Less than half (about 46%, or 79 responses)
were completed questionnaires from facilities that disposed of
hazardous waste in landfills during 1981. Of the thirty facilities
that did not respond, five were "not delivered" and five were
"closed". Nineteen facilities did not respond during the field
period, and one facility (0.5% of the sample) refused to respond
to the Landfill Questionnaire.
B.2.3.4 Response Rate for the Land Treatment Questionnaire
There were 99 responses to the 122 Land Treatment
Questionnaires that were mailed out. The response rate for the
Land Treatment sample is, therefore, 81.1 percent. This is the
lowest response rate for any of the samples for this survey.
There was a comparatively large proportion of "no contact"
nonresponses (where certified mail cards were not returned) in
the Land Treatment sample (16 facilities or 13.1 percent of the
total number of land treatment facilities sampled). A little
over one third (37.4%, or 37 questionnaires) of the Land Treat-
ment responses were eligible responses from facilities that disposed
B-25
-------
B.2.3.7 Response Rate for the Incinerator Questionnaire
Of the 265 facilities that were sent Incinerator
Questionnaires, responses were received from 239, for a response
rate of 90.2 percent. A little over half (about 54%, or 129
questionnaires) of the responses were received from facilities
that burned hazardous waste in incinerators regulated under RCRA
during 1981. The remaining 46 percent of the Incinerator
Questionnaires were returned by facilities that did not burn
hazardous waste in incinerators regulated under RCRA during 1981.
Eight of the 26 nonresponding incinerator facilities were closed,
and unable to respond. Six of the incinerator questionnaire mail
packages could not be delivered, and seven facilities had received
their packages but did not respond during the field period.
There were five facilities that refused to respond to the
Incinerator Questionnaire, representing 1.9 percent of the
incinerator facility sample.
B. 2 . 3 . 8 Response Rate for the Container Questionnaire
Of the 423 facilities that were sent Container
Questionnaires, responses were received from 389, for a response
rate of 92.0 percent. This is the best response rate among all
of the technology questionnaires, and is also a better response
rate than that of the TSD General Questionnaire (with 90.5%).
Half of the responses (50.1%, or 195 questionnaires) were
received from facilities that did engage in regulated container
storage during 1981, while the other half of the responses were
from facilities that did not store (for 90 days or longer)
hazardous waste in containers during 1981. Respondents' original
misidentifications of their facilities as container storage
facilities seemed to be due largely to the fact that many
respondents were unaware of the exemption from regulation that
B-28
-------
B.2.3.6 Response Rate for the Waste Piles Questionnaires
Of the 243 facilities that were sent Waste Piles
Questionnaires, responses were received from 215, for a response
rate of 88.5 percent. About one third of the responses (34%, or
73 questionnaires) were from facilities that stored hazardous
waste in waste piles during 1981. The remaining 66 percent of
waste pile responses were from facilities that did not store
hazardous waste in piles during 1981. This represents the lowest
eligibility rate among all eight of the waste management technology
questionnaires. Reasons for the low eligibility rate included
protective filing, respondent/applicant misidentification of the
technology and, possibly, respondent use of the technology for
treatment or disposal, rather than for storage. At the time the
questionnaire was developed, waste piles were regarded by the EPA
as only a storage process. Waste piles are now recognized as a
possible disposal process, and are therefore regulated under more
stringent standards (e.g., requiring groundwater monitoring).
However, the questionnaire deals with waste piles only as a
storage process.
Twenty-eight, or 11.5 percent, of the waste piles
facilities were nonrespondents. There were eight facilities
that could not be contacted by mail (one "no contact" and seven
"not delivered"). Ten facilities did not respond during the
survey period (including one facility with a time extension and
one facility that returned other questionnaires, but did not
return the Waste Pile Questionnaire). And nine facilities were
closed and unable to respond. There was one refusal, accounting
for less than one-half percent of the waste piles sample.
B-27
-------
respond. Five facilities refused to complete the Tank Question-
naire, representing less than 1 percent of the combined storage
and treatment tank samples.
B.3 Technical Assistance
Technical assistance was provided to mail survey
respondents in two ways: Westat provided a toll free number
which was staffed by data collection information specialists and
EPA provided assistance through the RCRA Hotline. Technical
assistance was provided extensively throughout the field period.
More than 2,600 telephone calls were placed to the
Westat 800 number and at least as many return calls were placed
from Westat. During the heaviest calling period, September 17
through October 21, as many as five information specialists were
required to staff the telephones. (The Westat toll-free
information service was in operation between 8:00 a.m. and 5:30
p.m., Eastern time, on weekdays, from September 1982 through
January 31, 1983.)
Most of the calls concerned the following issues:
• Whether or not the facility was regulated because
it was a small quantity generator;
• Requests for time extensions;
• How to interpret questions in the questionnaires;
• Requests for replacement questionnaires or
additional information for corporate offices; and
• Combinations of the above concerns.
B-30
-------
applies to generators who store their wastes in containers for
less than 90 days. There were 34 nonresponding facilities in the
container sample. Eleven could not be reached by mail; eight did
not respond within the field period. Three facilities could not
respond because they were closed. Twelve facilities refused to
respond to the Container Questionnaire, representing 2.8 percent
of the container sample.
B.2.3.9 Response Rate for the Tank Questionnaire
Both the sample of treatment tank facilities and the
sample of storage tank facilities were sent an integrated Tank
Questionnaire (with storage and treatment questions combined into
a single Tank Questionnaire). Of the 847 facilities that were
sent Tank Questionnaires, responses were received from 772, for a
response rate of 91.1 percent. However, only slightly more than
one third (37.3%, or 288 questionnaires) of these responses were
from facilities that had RCRA regulated treatment and/or storage
tanks during 1981. Respondents' original misidentifications of
their facilities as storage and/or treatment tank facilities
seemed to be due largely to the fact that many respondents were
unaware of the exemption from regulation that applies to generators
who store their wastes in tanks for less than 90 days (see 40 CFR
262.34), and the exclusion from regulation under RCRA of wastewater
treatment tanks that are regulated under NPDES (see 40 CFR 264.1(6)
Thus, 484 of the responding facilities did not have RCRA regulated
storage or treatment tank facilities on site, although many
reported non-regulated tanks .
There were 75 Tank Questionnaire nonrespondents. Of
these, 23 facilities could not be contacted, and 32 facilities
were contacted (by evidence of the certified mail card) but did
not respond. Fifteen facilities were closed, and unable to
B-29
-------
B.4 .1 Mail Followup Contacts
As indicated in previous sections, each facility that
had not responded at the end of the legally established 45-day
period was sent an individualized followup letter. (See Exhibit
B-4 for an example followup letter.) The letter advised the
facility/installation that its response had not been received,
and reminded the facility/installation of its legal obligation
to respond.
In addition, installations sampled from the Notifier
file that had not responded by early January were sent a second
reminder/enforcement letter.
B. 4 . 2 Followup Telephone_Ca_lIs;
Response rates for specific component (technology)
subsamples of the Part A/Verification sample were particularly
critical because of the relatively small number of facilities
engaged in certain activities, and the resultant small sample
sizes. Therefore, telephone followup, rather than a second mail
followup was employed to obtain the second responses from Part
A/Verification sample nonrespondents. The telephone followup is
a more effective medium than the mail reminder (although more
costly as well), and provided assurance that the reminder reaches
the right person at the facility.
Three hundred and thirty of the Part A/Verification
sample facilities had not responded by December 13, 1982. Between
December 13 and December 17, 1982, followup calls were placed to
314 (or 95%) of these facilities. By the end of the survey
period, there was a total of 94 TSD facilities for which no
final disposition or response had been obtained. (Eighty-two of
B-32
-------
Respondents with questions about the instructions in
the questionnaire or about definitions or interpretations of
questions in the questionnaires, and respondents with procedural
problems (e.g., requiring additional copies of material) were
assisted directly by the Westat data collection specialists.
Respondents with questions requiring interpretation of the
regulations were referred by the Westat Hotline staff to the EPA
RCRA Hotline. During September and October of 1982, the RCRA
Hotline reported handling more than 1,100 calls that were directly
related to the hazardous waste survey. It is estimated that
survey-related calls increased the total number of calls handled
by the EPA RCRA Hotline staff by as much as 20 percent during
the heaviest calling period.
Most calls to Westat were placed immediately after the
facility received the questionnaire package; the number of calls
began to decline about two weeks after the mailout was completed.
B. 4 Nonresponse Reduction Efforts
Nonresponse reduction efforts included mail and telephone
reminder (followup) contacts, tracing of "lost" mail, and the
time extensions granted to those respondents that requested
them. The nonresponse reduction strategy for the Notifier file
sample (of hazardous waste generators who manage their waste off
site) was slightly less rigorous than the strategy applied to
the Part A/Verification file sample (of telephone verified TSDs
from the Part A file). This was because an adequate (and better
than expected) number of responses had been achieved for the
Notifier file sample fairly early in the field period, and
critical sample sizes in several TSD technology subsamples
required more intensive followup effort.
B-31
-------
Exhibit B-4 (cont'd)
As noted in Dr. Skinner's August 17, 1982, cover
letter, information obtained from this questionnaire will
be used to reevaluate the effectiveness of our existing reg-
ulatory program and to identify situations where regulations
could achieve equal protection at lower cost.
We expect the prompt return of your overdue questionnaire.
If you require additional copies of the questionnaire, or if
you are experiencing difficulties in completing the instrument,
please call our toll-free RIA mail questionnaire assistance
service at: (800) 638-3000. Completed questionnaires or other
written comunications should be directed to:
Mr. George A. Garland
Analysis Branch Chief
Office of Solid Waste (WH-562)
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington , D.C. 20460
Let me remind you once again that your response to this
questionnaire is required by law. You are advised to return
the completed questionnaire to EPA as soon as possible.
Si ncerely,
W. Lamar Miller, Director
Waste Programs Enforcement
Office of Solid Waste and Emergency Response
U.S. Environmental Protection Agency
8-34
-------
Exhibit B-4. Followup Letter
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D C. 20460
Of F ict or
SOLID WASTC AND EMERGENCY RESPONSfc
r.e: EPA I.D. No. [ ]
Dear [ ]
A U.S. Environmental Protection Agency Hazardous Waste
Questionnaire package regarding the site corres pondinq to
the EPA I.D. Number referenced above was mailed to vou at
this address under a cover letter dated August 17, 1982, and
signed by Dr. John Skinner, Acting Director, Office of Solid
Waste.
The questionnaire was sent to you via certified mail
(Certified Mail Control Number [ ]) under the authority
of Section 3007(a) of the Resource Conservation and Recovery
Act. Your response, which is required by law, was due within
45 days of the date the questionnaire package was delivered to
this add res s .
Our records indicate that the questionnaire package was
accepted at this address on [September , 1982]. As such
your completed questionnaire should have been returned to FPA
post-marked no later than [October , 1982.] As of this date
however, the completed questionnaire corresponding to the EPA '
I.D. No. referenced above has not been received by EPA.
The purpose of this letter is to remind you of
your obligation under law to complete the questionnaire
you received for this site and return it to EPA in a timely
fashion. Failure to furnish the requested information could
subject you to civil penalties or other appropriate legal
action under Section 3008 of RCRA. In addition, Federal law
provides for both civil and criminal remedies in case of
knowing or willful falsification, concealment, or covering
* '
B-33
-------
questionnaires for most data items. These field positions (which
were distributed on 130 column records for ease of reading raw
listings) were useful as reference locations for coders, machine
editors, and data entry staff. Questionnaire layouts were
standardized among the ten questionnaires, to the greatest extent
possible, to minimize confusion and to minimize documentation,
software and training requirements.
Code manuals were developed for each of the 10 question-
naire types. These manuals described ~he data to be encoded
from the questionnaire, item by item. Figure B-5 lists the item
characteristics by which each data item was described in the
code manuals. Exhibit B-5 is an exarrple data item description
from the Treatment, Storage and Disposal General Questionnaire.
Figure B-5. Item Characteristics Described in Code Manuals
a. Field position and record number
b. Item name (the name by which the item was called
in all computer programs and other documentation
c. Quotation of the item f' orn the questionnaire
d. List of all code values and their definitions
e. List of reasons for legitimate item nonresponse
(the "inapplicable" definition)
f. List of all missing va Me codes
g. Flags indicating logical relationships between
the item and subsequent: items.
B-36
-------
these facilities simply had not responded. Twelve facilities
had requested and received time extensions but had not responded
by the end of the survey period.)
B.4.3 Time Extensions
Approximately 5,000 facilities/installations asked EPA
for, and were granted, extensions on the due date of their
questionnaires.
For both the Notifier sample and the Part A/Verification
sample, it was necessary to extend the survey period to accommodate
responses from facilities/installations that had received time
extensions.
B.5 Data Preparation
Data preparation began with a development phase involving
questionnaire layout and code manual design. Operational phases
included document handling (including receipt control), coding/
editing, data entry, and machine editing.
B.5.1 Questionnaire and Code Manual Design
Questionnaire layouts were designed for ease of data
preparation/data processing, as well as for ease of respondent
understanding and recording. Many items were designed as "preceded"
questions, that asked the respondent to answer by circling a
code to indicate his/her response. This eliminated the need for
a coder to translate check-marks or other non-code symbols into
coded answers. Computer field positions were printed in the
B-35
-------
B.5.2 Document Handling
Documents were logged in at the time they were received,
on an automated (computerized) receipt control system. (See
Section B.2 for a detailed description of the receipt control
system.) All documents from one facility were handled and filed
under one control number. This made it possible to edit across
questionnaires, when there were two or more questionnaires per
facility.
B.5.3 Coding/Editing
Because there were many more facilities responding
with only Generator Questionnaires, than with Treatment, Storage
and Disposal General Questionnaires and technology questionnaires,
two crews of coders were trained. One crew specialized in coding
just the Generator Questionnaire, and the other crew was trained
incrementally to handle all ten of the questionnaires. The
initial training session, which covered procedural matters as
well as the specific coding of the Generator Questionnaire, was
attended by both crews of coders. The coders' training included
an item-by-item discussion of the coding of the document, practice
coding examples, and group review of the coding of practice
examples. Training materials included code manuals, practice
examples, and a marked up version of the questionnaire that
linked the questionnaire to the code manual and the general
coding instructions. The initial coder training session con-
sisted of approximately twelve hours of training, spread over
two days. Coders assigned to TSD and technology questionnaires
were given an additional two to four hours of training per ques-
tionnaire.
B-38
-------
Exhibit B-5 . Code Manual Data Item Description
Q30
125
Code +'s in Q31 - Q36,
col 126-127, rec 17,
and col 16-33, rec 18
Is this facility located within one mile of a fault that has had dis-
placement within the past 10,000 years (Holocene time)? [CIRCLE ONLY
ONE CODE]
s Inapplicable, coded +'s in Q3B, Q3C and Q3D, col 75-80,
record 01
= Yes
= to
= OK
s Mot ascertained
B-37
-------
data file for the Notifier generators was a simple
one-questionnaire-per-installation file, so its edit program
consisted of a single program. For the Part A/Telephone
Verification TSD facility file, however, the complete record for
a TSD facility consisted of a TSD questionnaire plus at least
one (and up to four) technology questionnaires. The machine
edit program for this file used the TSD Questionnaire as a base
and provided for the co-editing of each additional technology
questionnaire. (That is, the ten programs were linked together
to make up one edit program.) This made it possible to do at
least some computerized interquestionnaire checking.
Machine editors were selected from the trained stock
of coders available from the two questionnaire coding crews.
The training consisted of procedural instructions, and a
walkthrough using an example edit problem.
The machine edit programs provided a list of test
errors for each edited case, as well as a listing of each case
in error. Each of the errors was checked, and often the hard
copy of the case was reviewed. Updates to the data files were
written on update sheets, key entered and run against the data
file to produce a new master file. Then the edit cycle was
rerun to make sure that the update corrections had been made
correctly. Because of the complexity of some of the technology
data files, it was necessary to rerun edit cycles several times:
updates to some fields tended to unexpectedly impact consistencies
with other fields.
After the final machine edit cycles, frequency
distributions for all items of the data files were reviewed by
supervisors to spot problems not captured by the machine edit
programs.
B-40
-------
Coders were trained to edit questionnaire responses
for consistency and completeness as they were coding them. Coders
flagged any problems they discovered during coding, and referred
the problem questionnaires to the coding supervisors. Some
problems required the development of new codes — such as when
different units of measure than those specified in the question-
naire were specified for quantity questions. Other problems
required that the respondent be called to verify a response or
provide missing information (a process called "data retrieval").
In some instances, decisions could be made based on the evidence
available, by the Project Officer or by other EPA staff. Decisions,
both general and case-specific, were recorded in a Decision Log,
for future reference.
All coding was 100 percent sight verified prior to
being sent for data entry.
B.5.4 Data Entry
Data was entered ("keypunched") by highly trained data
entry operators, using a key-to-tape entry system. This key-to-
tape system is computer driven and provides a formated entry
keying program that minimizes many types of data entry errors.
All data entry was 100 percent key verified by a different
operator from the entry operator.
B.5.5 Machine Editing
Machine editing is a means of data quality control
that uses a computer program to test item ranges, skip patterns,
and logical consistencies in a data file. Such a machine edit
program was prepared for each of the ten questionnaires. The
B-39
-------
B. 7 Confidentiality
Respondents were informed, in the cover letter to the
questionnaire package, that they could assert a claim of
business confidentiality as provided in Title 40 of the Code of
Federal Regulations, Part 2, Subpart B, and according to the
procedures set forth in Section 40 CFR 2.203 (b). A total of 93
Part A/Telephone Verification facilities and 36 Notifier generator
installations (or a total of 129 sites, overall) chose to request
business confidentiality.
The EPA Office of Solid Waste received all incoming
survey mail and screened all responses for requests for CBI
prior to forwarding survey responses to Westat for coding. All
incoming responses with CBI claims were held by the Project
Officer at the EPA Office of Solid Waste in a locked file. At
the end of the survey, after deleting all identifying informa-
tion and assigning pseudo-identifiers to the CBI responses, the
questionnaires were coded and entered on the data file. The EPA
Project Officer holds the link list of actual and pseudo-
identifiers in a locked file at the EPA Office of Solid Waste.
In addition, Westat entered into agreements with two
corporations concerning confidentiality. And one trade association
acted as a go-between, funnelling anonymous responses from its
members. (It was not possible to use many of the responses from
this organization, since many of the questionnaires could not be
assigned correct sample weights.)
All Westat staff involved in this project signed
standard Westat confidentiality agreements. In addition, all
staff members were required to sign EPA confidentiality
agreements.
B-42
-------
B.6 Data Retrieval
Data retrieval is the term used to refer recontacting
respondents for the purpose of verifying or clarifying responses
to completed questionnaires or interviews. For this study,
recontact of respondents generally took the form of a telephone
call, though occasionally it was necessary to mail a list of
questions to a respondent.
The need for a questionnaire to have data retrieval
could have been determined at several stages during the data
preparation process (and indeed, some data retrieval work was
done during the analysis stage.) Initially, a list of key
"must-be-answered" questions was provided by EPA staff. Missing
key items (or inconsistent key items) were often flagged by
coders during the coding/editing process. In addition, some
questionnaires required data retrieval in order to determine the
hazardous waste management or generation status of the facility.
Data retrieval staff consisted initially of Westat's
Hotline personnel. As the need for a larger data retrieval
staff grew, the hotline staff trained others to assist in the
retrieval calls. The data retrieval staff ultimately included
Westat Hotline personnel, the receipt control supervisor, coding
supervisors, coders, telephone interviewers, and EPA staff.
During the data preparation period, a total of 1,005
installations from the Notifier generartor file (or 11% of the
9,361 respondents) received data retrieval calls. And a total
of 762 (or 32%) of the 2,353 Part A/Telephone Verification file
respondents received data retrieval calls. In additon to these
calls, analysts from EPA and from Westat have made additional
data retrieval calls during the analysis period.
B-41
-------
-------
APPENDIX C
SELECTED REGULATIONS IMPLEMENTED
UNDER THE RESOURCE CONSERVATION AND
RECOVERY ACT OF 1976, AS AMENDED
C-l
-------
-------
S Ł iE»
Monday
November 17, 1980
Part X
Environmental
Protection Agency
Hazardous Waste Management System:
Suspension of Rules and Proposal of
Special Standards for Wastewater
Treatment Tanks and Neutralization
Tanks
C-3
-------
Federal Register / Vol. 45, No. 223 / Monday. November 17. 1980 / Rules and Regulations 76075
in this action, the Agency believes that
an effective date six months after
promulgation would defeat the purposes
of these amendments. Consequently, the
Agency is making these amendments
effective on November 19,1980.
Dated: November 10,1980.
DouglM M. CoatU.
Administrator.
Title 40 of the Code of Federal
Regulations is amended as follows:
PART 260—HAZARDOUS WASTE
MANAGEMENT SYSTEM: GENERAL
1. Add the following definitions to
5 2e0.10{a).
§ 2*0.10 Definitions.
(a) ' * '
(15a) "Elementary neutralization unit
means a device which:
(i) Is used for neutralizing wastes
which are hazardous wastes only
because they exhibit the corrosivity
characteristic defined in § 281.22 of this
Chapter, or are listed in Subpart D of
Part 281 of this Chapter only for this
reason; and,
(ii) Meets the definition of tank.
container, transport vehicle, or vessel in
§ 260.10 of this Chapter.
(76a) "Wastewater treatment unit"
means a device which:
(i) Is part of a wastewater treatment
facility which is subject to regulation
under either Section 402 or Section
307(b) of the Clean Water Act; and
(ii) Receives and treats or stores an
influent wastewater which is a
hazardous waste as defined in § 281.3 of
this chapter, or generates and
accumulates a wastewater treatment
sludge which is a hazardous waste as
defined in § 261.3 of this chapter, or
treats or stores a wastewater treatment
sludge which is a hazardous waste as
defined in § 261.3 of this chapter and
(iii) Meets the definition of tank in
§ 280.10 of this chapter.
PART 264—STANDARDS FOR
OWNERS AND OPERATORS OF
HAZARDOUS WASTE TREATMENT,
STORAGE, AND DISPOSAL
FACILITIES
2. Add the following paragraph to
§ 264.1(g):
§ 264.1 Purpose, scope and applicability.
(g) * * *
(6) The owner or operator of an
elementary neutralization unit or a
wastewater treatment unit as defined in
§ 260.10 of this chapter.
PART 265—INTERIM STATUS
STANDARDS FOR OWNERS AND
OPERATORS OF HAZARDOUS WASTE
TREATMENT. STORAGE, AND
DISPOSAL FACILITIES
3. Add the following paragraph to
§ 285.1(c):
§ 26S.1 PurpoM, scop* and applicability.
(c) * * "
(10) The owner or operator of an
elementary neutralization unit or a
wastewater treatment unit as defined in
§ 260.10 of this chapter.
PART 122—EPA ADMINISTERED
PERMIT PROGRAMS: THE NATIONAL
POLLUTANT DISCHARGE
ELIMINATION SYSTEM; THE
HAZARDOUS WASTE PERMIT
PROGRAM; AND THE UNDERGROUND
CONTROL PROGRAM
4. Add the following definitions to
§ 122.3.
§ 122.3 Definitions.
"Elementary neutralization unit"
means a device which:
(a) Is used for neutralizing wastes
which are hazardous wastes only
because they exhibit the corrosivity
characteristic, defined in § 261.22 of this
chapter, or are listed in Subpart D of
Part 261 of this Chapter only for this
reason; and.
(b) Meets the definition of tank.
container, transport vehicle, or vessel in
§ Z60.10 of this Chapter.
"Waste wafer treatment unit" means a
device which:
(a) Is part of a wastewater treatment
facility which is subject to regulation
under either Section 402 or Section
307(b) of the Clean Water Act; and
(b) Receives and treats or stores an
influent wastewater which is a
hazardous waste as defined in § 261.3 of
this chapter, or generates and
accumulates a wastewater treatment
sludge which is a hazardous waste as
defined in § -161.3 of this chapter, or
treats or stores a wastewater treatment
sludge which is a hazardous waste as
defined in § 261.3 of this chapter and
(c) Meets the definition of Unk in
§ 260.10 of Lh".s chapter.
5. Add the following paragraph to
§ 122-21(d)(21:
§ 122.21 Purpose and scope of Subpart B.
(d) • ' *
(2) * ' '
(vi) Owners and operators of
elementary neutralization units or
wastewater treatment units as defined
in 40 CFR 260.10.
(Sees. 1006, 2002(a), 3004. 3005, Solid Waste
Disposal Act, as amended by the Resource
Conservation and Recovery Act of 1976, as
amended (42 U.S.C. 8905, 6912(a). 8924 and
6925))
|FR Doc. U-1M1S FlUd 11-14-40: 8:49 «m|
BtUJNQ COOC (SM-2A-M
-------
76074 Federal^
, 4, No. - ^»«**~V.™ '**•"*>***--
40 CFR Parts 122, 260, 264 and 265
IWH-FRL 1670-31
among
obtain
0«ne« and Op.ra.or.
status sta
these
operators
^'facilities that
the Clean
: Environmental Protection
hazardous waste re» and
122. 2M -and 2*3 *
Parts 122. 2M -an
operators of 11) "««
. -P o«»nerate ana »w»* -* ". , .
waste or generalc sludge which is a
wastewater treauij; additi0n. owners
hazardous waste.x" s ^ neutralize
and operators ot iac t.iv Ky virtue
BUM wf hazardous soieiy uj
waste tnai »s u «ubiect to these
of its corrosivity are su ' . ents.
u*»"»•* , ctoratte requircinw""*
"M^r^i^-tetter
necessity of regulatmg these f_;v__
. ^ _«.f ^rtfi
Suspending the currem ~Ł—;ewater
insofar as they appiy neutralization
Sff^^iSSr'
•^srJSSijar*'
permit
treatment an
[ewav«i»
totalization
treatmenv a». detai\ed in the
^"^r^°^
--tr^r^atert-tmen.
y- w r
II Amendments
==ssi»"
of these
(1) Section iand operators
amended to add owners , erllary
ot W"feW,Un SSSTtoW of persons
neutraliiauon units ,
are
22=23?=""
promulgated in .^^g enacted in
This suspenSlon is Dcmg tor8 0{
order to relieve '""^ ^enlary
,»'0.^265 appUcab.e ,0 »». ^ WWj^l^S^
™{Sta*^-cSSSS3SK3 gSS^-S?
C-5
-------
56592 Federal Register / Vol. 48, No. 221 / Tuesday, November 17, 1981 / Rules and Regulations
ENVIRONMENTAL PROTECTION
AGENCY
40 CFR Part 265
[SWH-FRL 1960-5]
Interim Status Standards for Owners
and Operators of Hazardous Waste
Treatment, Storage, and Disposal
Facilities
AGENCY: Environmental Protection
Agency.
ACTION: Interim final rule and interim
final amendments to rules and request
for comments.
SUMMARY: The Environmental Protection
Agency (EPA) has issued standards
applicable to owners and operators of
hazardous waste management facilities
as required by the Resource
Conservation and Recovery Act
(RCRA). One of these standards bans
the disposal of most containerized liquid
hazardous waste in landfills, effective
November 19,1981. As a result of
reconsideration of this restriction. EPA
is today promulgating an interim final
rule to allow the disposal of small
containers of liquid and solid hazardous
waste m landfills provided that the
wastes are placed in overpacked drums
(lab packs) in the manner specified in
today's rule. The purpose of today's rule
is to provide an environmentally sound
disposal option for generators of small
containers of hazardous wastes, such as
laboratories.
DATES: Interim final rule and interim
final amendments effective November
17, 1981.
COMMENT DATE: The Agency will accept
comments on this rule and amendments
until January 18, 1982.
ADDRESSES: Comments should be
addressed to Deneen M. Shrader.
Docket Clerk. Office of Solid Waste.
(WH-562), U.S. Environmental
Protection Agency. 401 M Street, S.W.,
Washington, D.C 20460. telephone (202)
755-9173. Comments on today's interim
final rule and amendments should
identify the regulatory docket as
follows: "Section 3004—Lab packs."
FOR FURTHER INFORMATION CONTACT:
The RCRA hazardous waste hotline, toll
free at(800)424-9346 (544-1404 in
Washington, D.C.). For technical
information contact Kenneth Shuster.
Program Manager, Land Disposal
Branch. Office of Solid Waste (WH-
564). U.S. Environmental Protection
Agency. 401 M Street. S.W..
Washington. D.C. 20460. telephone (202)
755-9125.
SUPPLEMENTARY INFORMATION:
I. Introduction
On May 19. 1980, EPA promulgated
hazardous waste regulations in 40 CFR
Parts 260-265 (45 FR 33066 et seq.} which
established, in conjunction with earlier
regulations promulgated on February 28,
1980 (45 FR 12721 et seq.), the principal
elements of the hazardous waste
management program under Subtitle C
of the Resource Conservation and
Recovery Act of 1978, as amended (42
U.S.C 8921, et seq.). Sincelhat time, the
Agency has received numerous requests
to promulgate regulations tailored to the
special problems involved in the
management of smaller quantities of
different hazardous wastes. In
particular, some commenters have
stated that some of the interim status
hazardous waste standards for landfills
are geared towards large, homogeneous
waste streams but are inappropriate for
generators, such as laboratories, who
produce smaller quantities of many
different wastes. For reasons discussed
in Sections II and III of this preamble,
many of these commenters have
requested that the Agency allow these
smaller quantities of waste to be
disposed of in landfills when packaged
in "lab packs."
Laboratory wastes are commonly
collected in small containers ranging in
size from an ampule to 5 gallon pails.
These containers are surrounded by
some type of absorbent material such as
vermiculite and overpacked in-large
drums (usually 55 gallon) prior to
disposal in a secure landfill. The entire
package is commonly called a lab pack.
Although the term lab pack is
generally used to refer to a method of
disposing of laboratory wastes, today's
rule is not limited to the disposal of such
wastes. The disposal option authorized
by today's rule may be utilized by any
type of generator. It is designed to
accommodate generators who produce
smaller quantities of many different
wastes.
Today's amendments are designed to
relax two separate prohibitions against
the landfilling of lab packs which would
otherwise have become effective on
November 19.1981. Section 265.312
allows the burial of containerized liquid
ignitable waste in landfills until
November 19.1981. After that date,
liquid ignitable waste may not be placed
in landfills. Section 265.314 prohibits,
after November 19,1981. the burial of
containerized liquid hazardous wastes
except very small containers, such as an
ampule, or containers designed to hold
liquids for a use other than storage, such
as a battery or capacitor. (See 45 FR
33213 (May 19, 1980) and 45 FR 33502
(June 29.1981) for explanations of these
prohibitions.) The Agency has received
numerous requests to allow lab packs
containing liquid and liquid ignitable
hazardous wastes to be disposed of in
secure landfills after November 19, 1981.
the effective date of the prohibitions.
The disposal of hazardous wastes in
lab packs is a common practice for
many small volume generators (not
necessarily small quantity generators as
defined in 40 CFR 261.5) including.
particularly, commercial research
laboratories, school laboratories, and
large Governmental laboratones. This
represents a general trend away from
previous improper disposal methods for
these types of wastes, such as mixing
these wastes in dumpsters with
municipal waste or pouring the wastes
down the drain.
Preliminarily, it should be noted that
many high school, college and
university, or other small laboratories
may be small quantity generators and.
therefore, need not comply with the full
RCRA hazardous waste management
regulations provided that the wastes are
managed in accordance with § 261.5(g).
If generators are small quantity
generators as defined in 40 CFR 261.5,
their wastes, including those placed in
lab packs, are not subject to the RCRA
regulations contained in Parts 262
through 267 and Pans 122 through 124. or
to the notification requirements of
section 3010 of RCRA. provided that the
generator complies with § 281.5(g).
Hazardous wastes subject to the
reduced requirements of § 261.5 may be
mixed with non-hazardous wastes and
remain subject to these reduced
requirements, even though the resultant
mixture exceeds the quantity limitations
identified in § 261.5. unless the mixture
meets any of the characteristics of
hazardous waste identified in Subpart C
of Part 281.
Several commenters representing
laboratories have stated that although
they qualify for the small quantity
generator exemption, they would prefei
to dispose of their hazardous wastes at
a RCRA-permitted or interim status
hazardous waste landfill. By allowing
the disposal of lab packs in hazardous
waste landfills, the Agency is providing
a practical disposal option for these
generators, as well as for the generators
who do not qualify for the small
quantity generator exemption.
II. Summary of Comments
Most of the comments that the Agency
has received on the subject of lab packs
have been in responses to the February
20, 1981 amendment to 40 CFR 265.312.
which concerns the disposal of ignitable
wastes in landfills. These commenters
Ci
— i
-------
stated that disposal of lab packs in
secure landfills is environmentally
sound, provided that certain packaging
and pretreatment conditions are
followed. The commenters. in general.
requested that small containers—
ampule to 5-gallon pails—should be
allowed to be disposed of in lab packs
in landfills. One commenter specifically
requested that small containers (one
gallon and smaller, approved for DOT
shipment) be permanently allowed to be
landfilled since these non-leaking, small
containers, in cartons and palletized, do
not pose a substantial risk to human
health and the environment However,
the commenter further stated the EPA
could require that small containers be
placed in 55-gallon steel drums with the
voids packed with absorbent materials
before landfilling.
The comroenters stated that the
techniques for handling lab packs
prevent the potential for escape of
liquids. Additionally, they stated that
the quantity of such waste is small and
will not burden landfills that are
capable of handling chemical waste.
Even if the bottles or cans break or leak.
the packing will absorb the liquids.
Commenters also stated that isolating
materials that may be incompatible is
very important (i.e., incompatible
materials should not be placed in the
same lab pack), since chemicals must
not be allowed to react to cause fires or
other hazards. Further, one commenter
provided a list of substances that he felt
should not be allowed to be lab-packed
for disposal in landfills because, even m
small quantities, these substances
present too great a hazard for land
disposal.
III. Discussion of the Problem
Many thousands of generators
currently generate a variety of
hazardous wastes in smaller quantities.
Most of these generators are
laboratories, including chemistry and
biology laboratories in junior and senior
high schools, colleges and universities.
hospitals and clinics, Governmental
agencies with laboratories, large and
small research firms, and chemical.
pharmaceutical and other manufacturing
firms.
Although the number of generators
fitting this description is not known, the
15th edition of Industrial Research
Laboratories of the United States
contains information on 10,028 research
and development facilities belonging to
6,947 organizations engaged in
fundamental and applied research,
including development of products and
processes. Most of the facilities are
owned and operated by industrial firm*
but some foundations and cooperatively
• Vol. 46. No. 221 / Tuesday. November 17. 1981 / Rules and Regulations
^^^^^^^^^^^^^^^'"""""^""^''""^^"^"^^''""•••••'^•••••••••••••••••i
56593
supported units are also covered as
well as university laboratories having
research facilities separate from
university control. The American
Chemical Society's Directory of College
Chemistry Faculties (which covers rwo-
and four-year colleges and universities)
lists approximately 3,200 college
departments of chemistry, bio-
chemistry, chemical engineering, or
medical-pharmaceutical chemistry, each
of which can be expected to have at
least one laboratory.
The Agency has received several
examples indicating the magnitude of
laboratory waste generation. One large
university stated that it has more than
2.000 laboratories, each or which
generates a wide variety of waste
chemicals in small quantities. One
company that picks up small quantities
of laboratory wastes from generators
and then packs and transports the
wastes in lab packs for disposal
commented that it handled over 25,000
different chemicals in approximately
500.000 small containers in 1980. The
containers varied generally from
ampules of a few grams to 5-gallon pails.
One research laboratory stated that it
typically generates-well over a thousand
such small containers (several milliliters
up to about one gallon in size) for
disposal each month.
The availability of commercial
treatment options for small quantities of
hazardous-waste is greatly limited. A
typical laboratory produces small
quantities of many different wastes. The
variety and quantity of compounds
discarded are often unpredictable. Often
the specific waste characteristics are
unknown and the cost to characterize
such wastes is prohibitive. Commercial
treatment facilities (e.g., incinerators
and solvent recovery operations)
typically accept only reasonably sized
lots of well-characterized liquid wastes
delivered in a form which makes them
readily suitable for treatment Diverse
laboratory wastes in small containers
are not considered to be readily suitable
for treatment by operators of these
facilities.
Because in many cases the contents of
each small container of laboratory- or
hazardous -waste cannot be precisely
defined, commercial waste handlers are
reluctant to incinerate them. Proper
incineration requires analysis of waste
feeds for identification and designation
of principal organic hazardous
constituents, a very difficult task with
respect to diverse drummed wastes.
IV. Solutions
Based on the lack of available
treatment or disposal options for
laboratory wastes and on the Agency's
conclusion that landfill disposal of small
containers of hazardous wastes in
overpacked drums is environmentally
sound, the Agency has decided to allow
lab packs to be disposed of in hazardous
waste landfills.
The Agency believes that the disposal
of lab packs in landfills is an
environmentally sound practice.
Although the drums in which the
laboratory wastes are overpacked will
eventually degrade, the Agency believes
that by having, at a minimum, sufficient
absorbent material in each drum to
completely absorb all of the liquid
content of the inside containers, lab
packs will not contribute substantial
volumes of liquids to landfill leachate.
Today's requirement that the outside
container be full (i.e., absorbent material
to the top of the drum with no void
space), will assure that no breakage or
rupture of the inside containers will
occur during handling and placement.
One disposal alternative, other than
disposal in lab packs, is to mix liquid
wastes with an absorbent material
before placement in a drum, or to pour
liquid wastes directly into drums with
sufficient absorbent material to solidify
the liquid wastes. Provided that the
liquids are sufficiently absorbed or
solidified to remove free liquids, full
drums of such treated wastes are
already allowed to be landfifled under
the regulations, even after the § 265.314
ban on containerized liquids in landfills
takes effect. This method differs from
packaging in lab packs in that liquid
wastes are absorbed prior to disposal
rather than contained in inside
containers. The effectiveness of the
absorption is therefore observable.
While the option of mixing before
disposal may be viable for some
generators, based on the chemical
handling procedures of many
laboratories, disposal in overpacked
inside containers may be much more
practical and often safer for small
quantities of wastes.
V. DOT and EPA Coordination
The Department of Transportation
(DOT) has issued regulations governing
the transport of hazardous materials at
49 CFR Parts 171-179. Those regulations
specify packaging requirements
applicable to the transport of hazardous
materials in commerce within the United
States. However, the DOT regulations
do not caver all hazardous wastes and
are not applicable to all lab packs (e.g..
lab packs disposed of on-site).
It should be noted that EPA has
previously adopted certain DOT
regulations in its Standards Applicable
to Generators of Hazardous Waste (40
C-9
-------
56594 Federal Register / Vol. 46. No. 221. / Tuesday. November 17. 1981 / Rules and Regulations
CFR Part 262). Pursuant to § 262.30. a
generator who transports hazardous
waste oc offers hazardous waste for
transport off-site, must package the
waste in accordance with applicable
DOT regulations on packaging under 49
CFR Parts 173,178, and 179. Therefore.
any generator transporting lab packs for
off-site disposal is already required to
conform with all applicable DOT
requirements for packaging.
The objective of the DOT regulations
is to insure the safe transport of
hazardous materials. EPA'3 concern in
promulgating today's regulation is to
insure the safe disposal of hazardous
wastes. To the extent possible. EPA has
adopted DOT specifications for the
packaging of lab packs for disposal.
However, because the objective of the
DOT regulations varies somewhat from
the purpose of today's rule, in some
cases the requirements of § 285.316 are
different, or stricter than the DOT
requirements. However, the Agency has
attempted to ensure consistency with
the requirements of DOT and to avoid
the imposition of conflicting
requirements wherever possible.
Today's rule applies certain DOT
specifications to some situations which
are outside of DOTs jurisdiction and
thus are not directly covered by the
DOT regulations (e.g., lab packs being
disposed of on-site). On the other hand.
generators or transporters who are
already covered by the DOT regulations
must still comply with ail applicable
sections of those regulations. Thus lab
packs offered for transportation may, as
in the past, be subject to additional DOT
requirements such as weight and
container size limitations. In addition.
DOT prohibits the shipment of corrosive
liquids in metal outside drums or barrels
(see 49 CFR 173.25) unless an exemption
is obtained in accordance with 49 CFR
Part 107 Subpart B. Since EPA is
requiring metal outside containers for
purposes of disposal (§ 265.316(b)),
persons subject to the DOT regulations
wishing to dispose of corrosive liquids
in lab packs must first obtain an
exemption from DOT.
VI. Content of thd Regulation
To achieve the objectives discussed
above, today's regulation adds a new
section to Part 265 (§ 265.316) and makes
conforming amendments to § § 265.312
and 265.314. In accordance with today's
regulation, wastes to be disposed of in
lab packs must be packaged in sealed
inside containers. The inside containers
must be of a design and constructed of a
material that will not react dangerously
with, be decomposed by, or be ignited
by. the waste held therein. In addition.
the inside containers must be of the size
and type specified in the DOT
hazardous materials regulations (49 CFR
Parts 173,178 and 179), if those
regulations specify a particular inside
container for that waste. The
requirement of using DOT-specified
inside containers for purposes of
packaging wastes for disposal in lab
packs is applicable whether or not the
lab pack will be regulated by DOT for
purposes of transportation. The reason
that EPA is adopting DOTs
specifications for inside containers is
that EPA seeks to achieve the same
objective that DOT has defined in its
regulations, namely that the inside
containers safely and effectively hold a
material without leakage. Based on the
fact that EPA seeks to achieve the same
objective, the Agency has decided to
employ the DOT specifications for
inside containers.
The DOT hazardous materials
regulations do not specify inside
containers for all hazardous wastes,
however. Therefore, for any waste not
addressed in the DOT regulations,
inside containers must meet only the
general performance standard (i.e., be of
a design and constructed of a material
that will not react dangerously with, be
decomposed by, or be ignited by, the
waste held therein).
In addition to the requirement that
the inside containers be non-leaking, the
Agency has also included a requirement
in § 265.316(a) that all inside containers
be tightly and securely sealed. This
requirement is intended to help insure
that no waste leaks from the inside
containers before the lab pack is placed
in the landfill.
Section 285.316(d) prohibits the
placement of incompatible wastes in the
same outside container. The purpose of
this restriction is to prevent any
potentially dangerous reaction between
wastes packaged in the same lab pack.
The DOT hazardous materials
regulations contain a similar provision.
Those regulations state that the offering
of packages of hazardous materials in
the same packaging, freight container, or
overpack, with other hazardous
materials, the mixture of contents of
which would be liable to cause a
dangerous evolution of heat or gas or
produce corrosive materials, is
forbidden except as specified (see 49
CFR 173.21). EPA has included a similar
provision, however, because not all
hazardous wastes and thus not all lab
packs will be covered by the DOT
regulations
In addition to the prohibition against
co-packaging incompatible wastes
contained in § 265.316, it should be
noted that $ 265.313 already prohibits
the placement of incompatible wastes or
incompatible wastes and materials in
the same landfill cell unless § 285.17(b)
is complied with. Section 285.17(b)
states that: the mixture or commingling
of incompatible wastes or incompatible
wastes and materials must be conducted
so that it does not: (1) Generate extreme
heat or pressure, fire or explosion, or
violent reaction: (2) Produce
uncontrolled toxic.mists, fumes, dusts.
or gases in sufficient quantities to
threaten human health; (3) Produce
uncontrolled flammable fumes or gases
in sufficient quantities to pose a risk of
fire or explosions; (4) Damage the
structural integrity of the device or
facility containing the waste; or (5)
Through other like means threaten
human health or the environment.
Section 265.313 is. of course, applicable
to the placement of lab packs in
landfills.
Section 285.316(b) deals with the
outside container and the type of
absorbent material required. EPA is
requiring that the inside containers be
overpacked in DOT specification open-
head metal drums no larger than 110
gallons in capacity and surrounded by,
at a minimum, a sufficient quantity of
absorbent material to completely absorb
all of the liquid contents of the inside
containers. DOT specifications for
containers are contained in 49 CFR Parts
178 and 179.
All lab packs must be in DOT
specification outside drums, whether or
not the wastes contained in the lab pack
are covered by the DOT regulations. The
reason for this is that these drums have
already been determined by DOT to be
sturdy enough to safely hold hazardous
materials. The 110-gallon capacity
limitation coincides with the maximum
size DOT specification container. In
addition, this capacity limitation is
designed to ensure that lab packs will
be used for their intended purpose, i.e.,
the disposal of smaller quantities of
many different wastes.
Commenters have stated that many
off-site landfill operators will accept
containerized wastes only in 55-gallon
drums. Comments are specifically
solicited on whether a capacity
limitation for outside containers is
appropriate and if so, what this
limitation should be. Based on the
volume and content of comments
received on this issue, the Agency will
consider amending the 110-gallon
limitation.
In many cases, the DOT regulations
allow a variety of acceptable packaging
options including metal, fiberboard,
plastic or wooden containers. However,
for purposes of disposal, EPA is
C-10
-------
Federal Register / Vol. 46. No. 221 / Tuesday. November 17. 1981 / Rules and Regulations
56595
requiring that all outside containers be
metal. The need for metal drums is due
to the nature of disposal. Allowing fiber
or wooden containers to be used as an
outside container would increase the
risk of breaking or rupturing the inside
containers because fiber or wooden
containers are more likely to be
ruptured or crushed during handling and
after placement in a landfill than are
metal drums. The drums must be of the
open head variety to allow the proper
placement of the inside containers and
absorbent
The inside containers must be
overpacked and surrounded, at a
minimum, by a sufficient quantity of
absorbent material to completely absorb
all the liquid contents of the inside
containers. In addition, the outside
container must be full after packing with
the inside containers and absorbent
material to prevent breakage of inside
containers. The absorbent material used
must not be capable of reacting
dangerously with, being decomposed byr
or being ignited by the contents of the
inside containers in accordance with
5 26S.l7(b). The Agency has not
specified the type of absorbent that
must be used in a lab pack. However.
based on comments received, it appears
that vermiculite and fuller's earth are
commonly used because of their price.
availability, and the fact that they will
not react dangerously with most wastes.
The Agency has not specified a
maximum limit on the size of the inside
containers except where the DOT
regulations impose a specific
requirement. However, the total amount
of liquid which may be placed in the lab
pack will be Limited by the amount of
absorbent material required. Of course.
the higher the absorptive capacity of the
absorbent material used for
overpacking. the more liquid the lab
pack may contain.
VII. Ban on Certain Reactive Wastes
Section 265.312 bans the disposal of
reactive waste in landfills unless the
waste is treated or rendered non-
reactive prior to or immediately after
placement in the landfill. However, as a
result of comments received, the Agency
recognizes that cyanide- or sulfide-
bearing wastes, which are deemed
reactive because they meet the
characteristic of reactivity set forth in 40
CFR 261.23{a)(5). may be safely
landfilled in lab packs provided they are
properly handled so as to avoid contact
with incompatible wastes, as required
by § 285.316(d).
By definition cyanide- and sulfide-
bearing wastes are those which will
generate toxic gases, vapors, or fumes
when exposed to acidic or basic
conditions characterized by a pH
between 2 and 12.5. All other reactive
wastes will explode or release toxic
gases, vapors, or fumes, when they are
at standard pressure and temperature:
when they are mixed with or exposed to
water: when they are subject to a strong
initiating force; or when they are heated
under confinement, or else are DOT-
forbidden. Class A. or Class B
explosives. While it is possible to isolate
cyanide- and sulfide-bearing wastes in a
lab pack from wastes or conditions that
would cause them to generate toxic
gases, vapors, or fumes, it is much more
difficult to protect other reactive wastes
from conditions which would cause
them to explode or otherwise
dangerously react, even when packaged
in a lab pack. Therefore, today's
regulation contains a ban on the landfill
disposal of reactive wastes, other than
cyanide- and sulfide-bearing wastes, in
lab packs unless the waste is rendered
non-reactive prior to packaging.
It should be noted that some wastes,
such as oxidizers. may meet a
•characteristic of reactivity as well as the
characteristic of ignitabiliry. Although,
pursuant to today's rule, ignitable
wastes may be landfilled in lab packs,
any ignitable waste that also meets a
characteristic of reactivity other than
§ 261.23(a)(5). may not be disposed of in
a lab pack unless it is treated or
rendered non-reactive prior to
packaging.
VIII. Effective Date
Section 3010(b) of RCRA provides that
EPA's hazardous waste regulations and
revisions to the regulations take effect
six months after promulgation. The
purpose of this requirement is to allow
persons handling hazardous wastes
sufficient lead time to prepare and to
comply with major new regulatory
requirements. Today's amendments are
designed to reduce burdens imposed by
existing regulations. Therefore, an
effective date of six months after
promulgation would be contrary to the
purpose of section 3010(b). For this
reason, this rule and amendments take
effect immediately.
IX. Interim Final Rule and Amendments
and Request for Comment
EPA is promulgating today's rule and
amendments as interim final and is
providing a 60-day comment period. The
Agency believes that the public should
have an opportunity to comment on the
rule and amendments and. indeed, has
specifically requested comments.
However, the Agency believes that the
rule and amendments should be put into
effect during the comment period. To do
otherwise would be contrary to the
public interest by causing the regulated
community to comply with requirements
which this rule and amendments are
designed to change. Therefore, the
Agency finds that there is a "good
cause" to allow today's rule and
amendment to take effect prior to notice
and public participation under Section
553(b) of the Administrative Procedures
Act.
X. Regulatory Impact
Under Executive Order 12291. EPA
must judge whether a regulation is
"major" and therefore subject to the
requirement of a Regulatory Impact
Analysis. This interim final regulation is
not major since its effect is to reduce the
overall costsrand economic impact of
EPA's hazardous waste management
regulations. This reduction is achieved
by allowing the landfill disposal in lab
packs of certain hazardous wastes
which would otherwise be banned from
landfills. This being the case, the
present rule and amendments are not a
major regulation and no Regulatory
Impact Analysis need be conducted.
This amendment was submitted to the
Office of Management and Budget
(OMB) for review as required by
Executive Order 12291.
XL Regulatory Flexibility Act
Pursuant to the Regulatory Flexibility
Act. 5 U.S.C. 601 et seq., whenever an
agency is required to publish a general
notice of rulemaking for any proposed or
final rule: it must prepare and make
available for public comment a
regulatory flexibility analysis which
describes the impact of the rule on small
entities (i.e.. small businesses, small
organizations, and small governmental
jurisdictions). The Administrator may
instead certify, however, that the rule
will not have a significant economic
impact on a substantial number of small
entities.
This amendment will generally have
no adverse economic impact on small
entities in that it merely provides
another disposal option to entities
already subject to regulation under
RCRA. Accordingly, I hereby certify that
this final regulation will not have a
significant economic impact on a
substantial number of small entities.
This regulation therefore does not
require a regulatory flexibility analysis.
Oil
-------
56598 Federal Register / Vol. 46, No. 221 / Tuesday, November 17, 1981 / Rules and Regulations
Dated: November 12.1981.
Ann* M. Gorsuch,
Administrator.
PART 265—INTERIM STATUS
STANDARDS FOR OWNERS AND
OPERATORS OF HAZARDOUS WASTE
TREATMENT, STORAGE, AND
DISPOSAL FACILITIES
For the reasons set out in the
preamble, 40 CFR Part 265 is amended
as follows:
1. The authority citation for Part 285
reads as follows:
Authority: Sees. 1008, 2002(a). and 3004.
Solid Wa«te Disposal Act as amended by the
Resource Conservation and Recovery Act of
1976, as amended (42 U.S.C 6905. 6912(a).
and 6924).
2. Section 285.312 is amended by
revising paragraph (a) to read as
follows:
§ 265.312 Special requirements for
ignitable or reactive wast*.
(a) Except as provided in paragraphs
(b) and (c) of this section and in
§ 265.316, ignitable or reactive waste
must not be placed in a landfill, unless
the waste is treated, rendered, or mixed
before or immediately after placement in
the landfill so that:
(1) The resulting waste, mixture, or
dissolution of material no longer meets
the definition of ignitable or reactive
waste under § 281.21 or 261.23 of this
chapter, and
(2) Section 285.17(b) is complied with.
• • * • *
3. Section 265.314 is amended by
revising paragraph (b)(2) and by adding
paragraph (b)(3) to read as follows:
§ 265.314 Special requirements for liquid
wast*.
(bp • •
(2) The container is very small such
as an ampule; or
(3) The container is disposed of in
accordance with § 285.316.
4. A new § 265.316 is added to read as
follows:
§ 265.316 Disposal of small containers of
hazardous wast* in overpack*d drums (lab
packs).
Small containers of hazardous waste
in overpacked drums (lab packs) may be
placed in a landfull if the following
requirements are met
(a) Hazardous waste must be
packaged in non-leaking inside
containers. The inside containers must
be of a design and constructed of a
material that will not react dangerously
with, be decomposed by, or be ignited
by the waste held therein. Inside
containers must be tightly and securely
sealed. The inside containers must be of
the size and type specified in the
Department of Transportation (DOT)
hazardous materials regulations (49 CFR
Parts 173.178 and 179), if those
regulations specify a particular inside
container for the waste.
(b) The inside containers must be
overpacked in an open head DOT-
specification metal shipping container
(49 CFR Parts 178 and 179) of no more
than 418-liter (110 gallon) capacity and
surrounded by, at a minimum, a
sufficient quantity of absorbent material
to completely absorb all of the liquid
contents of the inside containers. The
metal outer container must be full after
packing with inside containers and
absorbent material.
(c) The absorbent material used must
not be capable of reacting dangerously
with, being decomposed by, or being
ignited by the contents of the inside
containers, in accordance with
S 265.17(b).
(d) Incompatible wastes, as defined in
§ 260.10(a) of this chapter, must not be
placed in the same outside container.
(e) Reactive waste, other than
cyanide- or sulfide-bearing waste as
defined in § 261.23(a)(5) of this chapter.
must be treated or rendered non-
reactive prior to packaging in
accordance with paragraphs (a) through
(d)-of this section. Cyanide- and sulfide-
bearing reactive waste may be packaged
in accordance with paragraphs (a)
through (d) of this section without first
being treated or rendered non-reactive.
IFK Doc 81-33192 Filed 11-16-*!: 8:43 am|
S4UJNO COOC CS60-30-M
C-12
-------
APPENDIX D
SUMMARY OF MAJOR DATA ELEMENTS
CONTAINED IN COMPUTER DATA BASE FROM
THE NATIONAL SURVEY OF RCRA-REGULATED
GENERATORS AND MANAGEMENT FACILITIES
D-l
-------
I. HAZARDOUS WASTE GENERATOR QUESTIONNAIRE
Facility Identification (1-3)
1-2. Facility EPA I.D., name and address, and contact person
3. Primary and first three secondary SIC codes.
Waste Generation Activities, Quantities and Character-
istics(4-10)
4. 1981 hazardous waste activities: generate, treat,
store, dispose, transport or recycle.
5. Hazardous waste generation activities: waste generated
before, during and/or after 1981, waste generated but
exempt from RCRA for small quantity, farming or
recycling, or waste delisted.
6. Reasons for filing RCRA Notification.
7. Total quantity of hazardous waste generated.
8. Quantity generated that will be used, reused, recycled
or reclaimed.
9. Reasons for handling waste as hazardous.
10. EPA waste codes for all hazardous wastes generated.
Hazardous Waste Shipped Off Site (11-18)
11-12. Total quantity hazardous waste shipped off site for
TSD.
13-15. Quantity and cost of transporting, treating, storing
or disposing of wastes shipped off site to facilities
owned by other firms.
16. Percentage of total waste shipped off site and sent to
other states.
17-18. Identify transporters and find destinations, by
quantity, of hazardous waste shipped off site.
D-4
-------
Hazardous Waste Use, Reuse, Recycling or Reclamation
(UR3) (19-25)
19-21. Facility generated waste that was used, reused, recycled
or reclaimed before, during or after 1981.
22. Total quantity of waste generated that was used, reused,
or recycled; quantity recycled on site; quantity shipped
off site to same company and to other companies.
23. For the five wastes generated in largest quantities
and shipped off site for recycling: identify quantities
shipped and destinations; storage methods and duration
of storage prior to shipment.
D-5
-------
II. HAZARDOUS WASTE TREATMENT, STORAGE AND DISPOSAL
GENERAL QUESTIONNAIRE
Description of Faci1ity and Waste Processing Technologies
n^9_)_ " " ~~
1-2. Facility EPA I.D., name and address, and contact person.
3. 1981 hazardous waste activities: generate, treat,
store, dispose, transport or recycle.
4. Year waste management operations began.
5. Total area of facility (sq. ft.) and size of areas
containing each waste management process.
6. Number of employees, hourly wage and person-hour per
week, by occupational categories, by waste management
process.
7. Primary and first three secondary SIC codes.
8-9. Facility owned and/or operated privately or by
government.
Description and Quantity of Wastes Managed at Facility
'(10-15)
10. Total quantity of waste (hazardous and nonhazardous)
managed on site in 1981.
11. Potential waste (hazardous and nonhazardous) management
capacity during 1981.
12. Total quantity of hazardous waste managed on site in
1981.
13. Potential hazardous waste management capacity during
1981.
14. Percentage of hazardous waste transported to TSD
operation, by transportation method.
15. Quantities of wastes (hazardous and nonhazardous) and
hazardous waste that were or could have been disposed
of, treated or stored at the facility in 1981.
D-6
-------
Description and Capacity of TSD Technologies, Sources
of Hazardous Waste (16-24)
16. Existence and operational status of each type of waste
processing technology at the facility in 1981.
17. For each of the 10 largest volumes of hazardous wastes
managed at the facility in 1981, the quantity of TSD
by each waste processing technology.
18-24. Did facility receive hazardous waste for TSD from
off site source in 1981?
If yes, total quantity, quantity by transportation
method, quantity from other firms, quantity from RCRA-
exempted small generators, SIC codes of five largest
off-site generators, off-site hazardous waste analysis
methods used.
Groundwater Monitoring (25-29)
25. Methods used to prevent groundwater contamination.
26. Does facility have groundwater monitoring we11s?
27. Presence and number of hydraulically up- and down-
gradient groundwater monitoring wells.
28. Depth, cost, use and uses of groundwater monitoring
we11s.
29. Are there geologic/hydrogeologic studies of this
facility?
Site Geography (30-44)
30-31. Is facility within one mile of a fault active during
past 20,000 years? If so, distance in feet.
33-35. Type, intensity and damage caused by seismic ground
motion, if any, experienced by facility.
36. Design or locational criteria used by facility to
mitigate ground motion induced damage.
37. Is facility located in a floodplain?
38-39. Type of floodplain and frequency of flooding.
D-7
-------
40-42. If facility has been flooded, magnitude of most severe
flood and hazardous waste released, if any.
43-44. Types of flood protection used by facility, and
reasons for use.
Hazardous Waste Use, Reuse, Recycling or Reclamation
(UR3) (45-51)
45-47. Did or will facility generate or receive hazardous
waste for UR3 before, during or after 1981?
48. Total quantity of hazardous waste UR3 in 1981; quantity
on site, off site (same company), off site (another
company).
49. For five principal hazardous wastes generated or
received by facility that were shipped off site for
UR3 in 1981: description and quantity of waste, EPA
I.D. of facilities that received waste, on-site
storage methods used and number of days waste was
stored prior to shipment.
50-51. For five principal hazardous wastes that facility
used, reused, recycled, or reclaimed on site in 1981:
description and quantity of waste, methods used for
UR3, storage methods used and number of days waste was
stored prior to UR3. For waste UR3 in manner
constituting disposal, describe process.
Financial Assurance (52-63)
52. Instrument or method used to cover facility closure
and/or post-closure costs.
53. Annual administrative charges for maintaining financial
assurance.
54-55. Value of collateral, if any, facility owner has put up
for financial assurance.
56-57. When, if ever, did facility obtain liability coverage
for third-party damages from sudden or nonsudden ~
hazardous waste releases?
58. Was facility upgrade, modification or assessment
necessary to obtain liability insurance?
59. How many facilities does insurance cover?
D-8
-------
60. Number and type or liability policies held, amount of
coverage, annual cost and amount of deductible for
policies.
61-62. For how many years prior to the year policy was obtained,
if any, does policy cover accidents resulting from
sudden or nonsudden hazardous waste releases?
63. Does policy cover legal defense costs?
D-9
-------
III. HAZARDOUS WASTE LANDFILL QUESTIONNAIRE
1. Does facility have an active hazardous waste landfill?
Landfill Capacity and Utilization (2-6)
2-3. Design life and remaining operational life of
hazardous waste landfill.
4. Total design capacity, capacity used by 1981 and
capacity available after 1981 for hazardous waste
landfill.
5. Quantity of waste (hazardous and nonhazardous) and
hazardous waste that was or could have been disposed
of in landfill in 1981.
6. Additional capacity, if any, currently being added to
landfill.
Landfill Liner Systems (7-15)
7. Is landfill lined?
8. Composition, permeability and thickness of each layer
of landfill liner.
9. Date liner was installed.
10. Cost of installed liner system, by liner type.
11-12. Area and configuration of liner.
13. Source of synthetic liner materials.
14. Liner/waste compatibility information available.
15. Does landfill have liner leak detection system?
Leachate Collection System (16-32)
16. Does landfill have leachate collection system?
17-32. Description and capacity of leachate collection system;
treatment, storage and disposal methods for collected
leachate.
D-10
-------
Physical Characteristics of the Site (33-76)
33. Is any part of landfill operated as an area fill?
34-42. Description and number of lifts, if any, completed or
planned at the facility.
43. Has any part of the landfill been operated as a trench
fill or cell fill?
44-54. Description of cells, trenches and working cover used
in landfill; size, materials, waste/fill ratio.
55. Has final cover been applied to any part of landfill?
56. Composition, permeability and thickness of each layer
of final cover.
57. Date of cover installation.
58. Cost of installed cover systems, by type of cover.
59-60. Area of cover and source of synthetic cover materials.
61. Slope of ground surface after landfill closure.
62-66. Description and specifications of run-on control system.
67-71. Description and specifications of run-off control
system.
72-75. Description and specification of gas-venting system.
76. Does landfill have gas-to-energy recovery system?
Containerized Waste in Landfills (77-90)
77. Did facility dispose of containerized waste (hazardous
or non-hazardous) in 1981?
78-80. Number of containers, quantity of waste and containerized
liquid waste or waste with free liquids disposed of in
1981.
81. For five containerized liquid hazardous waste
landfilled in greatest quantity prior to 1981: description,
liquid content, and physical characteristics of the
waste; SIC codes of generators; number of containers
landfilled in 1981.
D-ll
-------
82-85. Number of containers and quantity of containerized
waste, if any, that, is currently accepted or disposed
of in landfill.
86. How is presence of liquid in containers determined?
87. How is presence of liquid in waste determined?
88-89. How does facility handle containers containing liquid
wastes?
90. Quantity of bulk liquid landfilled in 1981.
Cost and Pricing (91-111)
91-95. Corporate financial ratios, facility's hazardous waste
revenues and costs.
96-102. Ownership status, quantity, cost and value of land and
equipment used for hazardous waste landfill.
103-104. Facility costs for operation and maintenance of landfill.
105-109. Pricing policy and total sales for commercial landfilling
110-111. Facility expenses and depreciation, profit or loss if
landfill sold.
On-site Landfilling (112-113)
112-113. On-site hazardous waste generation and landfilling.
D-12
-------
IV. HAZARDOUS WASTE SURFACE IMPOUNDMENT QUESTIONNAIRE
1. Facility has active hazardous waste TSD surface
impoundment.
Surface Impoundment Capacity and Utilization (2-7)
2-5. Number, utilization and capacity of surface impoundments
for TSD of hazardous and nonhazardous waste.
6-7. Expansion of surface impoundment capacity.
Impoundment Design and Construction (8)
8. For each of facility's six largest active hazardous
waste surface impoundments: age; remaining life;
construction cost; utilization; remaining capacity;
configuration; liner; leachate management; and run-on
protection.
General Operating Practices (9-16)
9. Waste codes and descriptions of hazardous waste TSD in
six largest surface impoundments.
10-15. Presence of sludges and liquid wastes in surface
impoundments.
16. For each of six largest hazardous waste surface
impoundments: extent and cost of sludge accumulation,
dredging and drainage; handling of sludge; description
of sludge; cost of sludge TSD; effects of sludge on
TSD and wastewater treatment activities.
Cost and Pricing (17-28)
17-20. Ownership status, quantity, cost and value of land
used for surface impoundments.
21-23. Quantities, and prices charged for commercial surface
impoundment TSD activities, over time.
24-25. Acceptance and prices charged for bulk liquid TSD in
surface impoundments.
26-27. Pricing policy relative to waste characteristics and
quantity.
D-13
-------
On-site Generation and TSD in Surface Impoundments
(28-30)
28-29. Proportion of total hazardous waste generated on site.
30. Reasons for having on-slte surface impoundment capacity
D-14
-------
V. HAZARDOUS WASTE PILE QUESTIONNAIRE
1. Does facility have active hazardous waste piles?
Waste Pile Capacity and Utilization (2-5)
2-5. Number, utilization, capacity and life of hazardous
waste piles.
Waste Pile Descriptions (6-24)
6-8. Number of piles indoors and outdoors.
9-12. Characteristics of run-on and run-off control systems.
13. For smallest, average and 10 largest hazardous waste
piles: height; base area and characteristics;
capacity; leak and leachate detection and collection.
14-i6. Waste/base compatibility.
17-18. Below-grade hazardous waste piles.
19. Slope of waste pile site.
20-21. Inspection of construction bases.
22-24. Removal of wastes, closure and covering of pile.c,.
Costs and Pricing (25-47)
25-29. Corporate financial ratios, facility's hazardous waste
revenues and costs.
30-31. Fees and receipts for commercial hazardous waste pile
storage.
32-33. Corporate rates of return on equity and rates.
34. Prices charged for commercial waste pile storage, over
time.
35-36. Pricing policy, by waste characteristics and quantity.
37-43. Ownership status, quantity, descriptions, cost and
value of land, buildings and equipment used for waste
pile operations.
D-15
-------
44-45. Annual costs for operating and maintaining hazardous
waste piles, by type of cost.
46. Facility depreciation and expenses for fiscal year.
47. Facility profit or loss if waste pile operations sold
On-Site Hazardous Waste Pile Storage (48-49)
48-49. On-site hazardous waste generation and reasons for
on-site storage in waste piles.
D-16
-------
VI. HAZARDOUS WASTE INCINERATOR QUESTIONNAIRE
1. Does facility have incinerators that have or will be
used to incinerate hazardous waste?
Incinerator Status (2-3)
2. Number of incinerators active, under construction or
out of service.
3. Plans for incinerator expansion.
Waste Quantities and Characterisitics (4-7)
4. For each active hazardous waste incinerator: quantity,
average feed rate and heating value of hazardous and
nonhazardous wastes burned.
5. Number of customers served and number of waste streams
burned by incinerator facility.
6. For each type of residual remaining after hazardous
waste incineration: quantity, percent hazardous,
percent treated on site and percent shipped offsite.
7. For five largest quantities of waste streams incinerated
waste code and description; quantity incinerated; feed
rate; heating value and physical form; storage and
pretreatment; and hazardous organic constituents.
Incinerator Design and Operating Characteristics (8-10)
8. For each active or under-construction incinerator
unit: incinerator type and manufacturer7; age and life;
utilization and capacity in hours, Btu's and feed
rate; largest quantity waste streams; actual feed rate
for hazardous and nonhazardous waste; waste feed and
residue removal methods; flue gas volume and air
pollution control; emergency procedures; and energy
recovery.
9. For each combustion chamber in two largest units:
volume and air flow rate and control; chamber
temperature and residence time; refractory type and
thickness; and auxiliary fuel type and use.
0- For the stacks of the two largest units: material;
height; diameter; and gas exit temperature and
velocity.
D-17
-------
Fuels, Materials and Maintenance (11)
11. For the two largest incinerator units: types and totals
of supplemental fuels used; daily supplemental fuel
use; electrical power used; chemicals used in scrubbers;
and type and frequency of routine maintenance.
Capital Investment Costs (12-15)
12-13. Types of costs of original incinerator units and major
modifications.
14. Cost and age of items and equipment related to
incinerator operation.
15. Depreciation, interest and debt amortization for items
12-14.
Operating and Maintenance (16-17)
16. Operating and maintenance costs, by type.
17. Cost savings due to energy or by-product recovery.
Pricing Policies (18-21)
18-19. Range of prices charged for hazardous waste incineration,
by type of waste.
20. Reasons for price differential.
21. Gross sales of incinerator services and recovered
energy.
Corporate Financial Data (22-25)
22-25. Corporate financial ratios.
D-18
-------
VII. HAZARDOUS WASTE LAND TREATMENT QUESTIONNAIRE
1. Does facility have any hazardous waste land treatment
areas?
Land Treatment Site Design Capacity (2-7)
2-3. Area of land treatment units, by type and status.
4-5. Design capacity and operation life of land treatment
areas.
6-7. Utilization and capacity of hazardous waste land
treatment areas.
Site Design (8-13)
8-13. Location, run-on, run-off and erosion controls of land
treatment area.
Soil Characteristics (14-19)
14-18. Soil characteristics and area drainage.
19. Presence and concentration of EPA listed hazardous
constituents.
Waste Characteristics (20-22)
20. Types and sources of wastes treated.
21. For the 10 hazardous wastes treated in largest
quantities: waste code and description; quantity treated;
moisture content and pH; hazardous constituents; and
pretreatment.
22. Results of land treatment on hazardous constituents.
Site Operation (23-27)
23-27. Methods and timing of applications of hazardous wastes
to land treatment areas.
Monitoring (28-40)
28-30. Run-off collection and analysis.
31-37. Soil core and soil pore water sampling and analysis.
D-19
-------
38-40. Frequency and parameters of air monitoring.
Closure Program (41-42)
41-42. Timing and frequency of activities planned for land
treatment area after final application of hazardous
waste.
Costs and Pricing (43-65)
43-47. Corporate financial ratios, facility's waste
management revenues and costs.
48-49. Total sales receipts for commercial land treatment.
50-51. Corporate rates of return on sales and equity.
52-54. Pricing policies and prices changed, over time, for
commercial land treatment.
55-61. Ownership status, quantity, cost and value of land,
buildings and equipment used for land treatment
operation.
62-63. Facility costs for operation and maintenance of land
treatment area.
64-65. Facility expenses and depreciation, profit or loss if
land treatment operation sold.
On-site Land Treatment (66-67)
66-67. On-site hazardous waste generation and land treatment
D-20
-------
VIII. HAZARDOUS WASTE TANK QUESTIONNAIRE
1. Facility has tanks used for accumulation (<90 days),
storage and/or treatment of hazardous wastes.
Tank Capacity and Utilization (2-20)
2. Number of tanks at facility.
3. Facility generated hazardous waste.
4. Wastes accumulated under RCRA 90-day rule.
5-7. Number of tanks used for hazardous waste storage.
8-9. Quantity of hazardous waste in storage tanks at
beginning and end of 1981.
10. Average duration of hazardous waste storage.
11-15. Number, capacity and utilization of tanks operating
under RCRA wastewater treatment tank exemption.
16-20. Number, capacity and utilization of tanks used for
treating hazardous wastes (not under wastewater
treatment exemption).
Tank Description Sheets (21:Tl-29)
21. Description of each hazardous waste treatment and/or
storage tank at facility (excluding 90-day
accumulation tanks).
T1-T4. Tank type, capacity, average and maximum utilization.
T5-T6. Age and life of tank.
T7. Characteristics of wastes stored or treated in tank.
T8-T11. Tank location, configuration (open or closed-topped,
above or below ground) and air emission control.
T12-T14. Frequency of inspection and other integrity monitoring
T15-T17. Construction material for tank and liner.
T18-T22. Presence and capacity of secondary containment and
safety systems.
D-21
-------
T23-T27. Wastewater treatment processes used (if any) and
characteristics of wastewater streams.
T28. Tank received hazardous wastewater generated by other
facilities.
T29. Cost and age of ta.nk, ancillary equipment and secondary
containment for tank and ancillary equipment.
Corporate Financial Information (22-25)
22-25. Corporate financial ratios and facility solid waste
management revenues.
Costs and Pricing for Storage (26-33)
26. Cost of land, buildings and equipment related to
hazardous waste storage tanks.
27. Cost of operating and maintaining storage tanks.
28-32. Prices and pricing policies for commercial tank
storage of hazardous waste, and total annual receipts.
33. Gain or loss if tank waste storage operation sold.
Costs and Pricing for Treatment Services (34-41)
34. Cost of land, buildings and equipment related to
hazardous waste tank treatment.
35. Cost of operating and maintaining treatment tanks.
36-40. Prices and pricing policies for commercial tank
treatment of hazardous waste, and total annual receipts
41. Gain or loss if tank waste treatment operation sold.
D-22
-------
IX. HAZARDOUS WASTE CONTAINER QUESTIONNAIRE
1. Facility uses containers for storing or accumulating
hazardous waste.
Storage and Accumulation in Containers (2-10)
2- Number of container storage or accumulation areas.
3. Does facility generate hazardous waste?
4-5. Number of hazardous waste container accumulation areas
(stored < 90 days).
6-7. Number of hazardous waste storage areas.
8~9- Quantity of containerized waste in storage at beginning
and end of 1981.
10. Duration of hazardous waste storage in containers.
Container Storage Area Description Sheet (11:C1-C12)
11. Description of each hazardous waste container storage
area at facility (excepting those operating under the
90-day accumulation rule).
C1-C3. Characteristics, typical and maximum quantities of
hazardous wastes kept in container storage areas.
C4-C8. Description and cost of container storage area base
material, run-off control and secondary containment.
CIO. Distance between this container area and nearest off-
site structure.
Cll. Number, cost and types of containers used in this
area.
Corporate Financial Information (12-15)
12-15. Corporate financial ratios and facility solid waste
management revenues.
Costs and Pricing for Storage Services (16-23)
16. Cost of land, buildings and equipment related to
hazardous waste storage containers.
17. Cost of operating and maintaining storage containers.
D-23
-------
18-22. Prices and pricing policies for commercial container
storage of hazardous wastes, and annual receipts.
23. Gain or loss if container storage operation sold.
D-24
-------
X. HAZARDOUS WASTE UNDERGROUND INJECTION WELL QUESTIONNAIRE
1. Does facility have* active hazardous waste disposal
injection well?
Well Capacity and Utilization (2-4)
2-3. Number, type and characteristics of injection wells
and their sites.
4. Responses to the following questions for wells that
(a) dispose below deepest groundwater, (b) dispose
into groundwater, (c) dispose above groundwater, or
(d) operate in another manner: quantity of hazardous
waste injected; capacity; average and maximum
injection rate; number of days of operation; and 10
greatest volume waistes.
Intermediate Storage and Treatment (5-9)
5-7. Presence, methods and capacity of on-site intermediate
waste storage.
8-9. Presence and methods of on-site waste treatment prior
to injection.
Costs and Pricing (10-26)
10-12. Corporate assets/liabilities, debt/equity and waste
revenues/total revenues ratios.
13. Average unit cost ,.o inject hazardous wastes.
14-17. Ownership status i.nd cost of land, buildings and
equipment related to hazardous waste injection wells.
18-19. Annual injection well operation and maintenance costs.
20-22. Source and amount of sales receipts for commercial
injection well activities.
23-24. Prices and pricing policies for commercial hazardous
waste injection well activities.
25-26. Facility depreciation and total expenses, net loss or
gain if injection well operation sold.
D-25
-------
Qn-site Well Injection (27-28)
27-28. On-site hazardous waste generation and deep well
injection.
94 9^ For five wastes generated in largest quantities and
24~25' recycled on site! methods and guantities of use reuse
recycling~oTTic-lamation; types and duration of storage
prior to recycling; description of recycling methods,
if any, that constituted disposal.
34
»•
(26-45)
26-28 Number and design capacity of tanks operating under
system exemption.
29-31 Number of storage areas and average guairtitY of waste
sTofed under the _9^-daj^ccumulation exemption.
32-33. Location and supervision of waste accumulation areas.
Percentage of waste accumulated under the 90-day rule
ainers.
in
containers or other containers.
35-36. Number and edacity of tank! used for 90-day accumulation
37-38. Number and capacity of containers areas used for 90-day
accumulation .
40. Was 90-day accumulated waste shipped off_site for
TSD or recycling?
41-42. Normal and optimum size for waste shipments.
47 44 Number of days to generate optimum sized waste shipment;
number of times it was necessary to ship smaller than
optimum quantities to meet 90-day accumulator exemption
requirements .
45-46 For waste exchange: Chemical Abstract Registry numbers
for e^cTTh^^rdo^ waste stream generated.
D-26
TVU.S. GOVERNMENT PRINTING OFFICE : 1984 O - 442-790
------- |