METHODOLOGY FOR DEVELOPING
BEST DEMONSTRATED AVAILABLE (BOAT)
TREATMENT STANDARDS
James R. Berlow, 'I
Treatment Technology Section
Robert April
Project Manager
U.S. Environmental Protection Agency
Office of Solid Waste
401 M Street. S. W.
Washington, O.C. 20460
December 1988
-------
*****NOTICE*****
THE COMPLETE SET OF BOAT BACKGROUND DOCUMENTS FOR THE S:--.OND
THIRDS (PROPOSED) HAVE BEEN ASSIGNED THE FOLLOWING EPA/OSW
NUMBERS.
EPA/530-SW-89-017A THROUGH EPA/530-SW-89-017G
EPA/530-SW-89-017I-J
EPA/530-SW-89-017L
EPA/530-SW-89-017M '
PLEASE NOTE THAT THERE ARE TWO NUMBERS THAT DO NOT EXIST
IN THE CONSECUTIVE LISTING:
EPA/530-SW-89-017H or EPA/530-SW-89-017K
*****NQTICE*****
-------
TABLE OF CONTENTS
Section Pace
1. INTRODUCTION 1
2. LEGAL BACKGROUND 2
2.1 General Requirements Under HSWA 2
2.2 Schedule for Developing Restrictions 4
2.3 Variances from the Schedule 6
3. METHODOLOGY FOR ESTABLISHING TREATMENT STANDARDS 8
3.1 Waste Treatability Groups 9
3.2 Determining BOAT for Individual Waste Treatability
Groups 10
3.3 Establishing Numerical Performance Standards on the
Basis of BOAT 14
3.3.1 Evaluating the Adequacy of Existing Data 15
3.3.2 Collecting Additional Performance Data 17
3.3.3 Hazardous Constituents Considered for
Regulation 20
3.3.4 Selecting Constituents for Inclusion in the
Standard 28
3.3.5 Calculation of Standards 30
3.4 Compliance with Performance Standards 32
2.5 BOAT Treatment Standards for "Derived-From" and
"Mixed" Wastes 32
3.5.1 Wastes from Treatments Generating Multiple
Residues 32
3.5.2 Mixtures and Other "Oerived-from" Residues ... 33
3.5.3 Residues from Managing Listed Wastes or
Wastes that Contain Listed Wastes 34
3.6 Other Types of Standards: Technology-Based and
"No Land Di sposal" 36
4 VARIANCES FROM THE BOAT TREATMENT STANOARC 38
5. P AND U WASTES 42
APPENDIX A STATISTICAL METHODS 44
A.I Accuracy Correction of Sampled Data 44
A.2 F Value for Determination for ANOVA Test 45
A.3 Variability Factor 56
i i
-------
LIST OF TABLES
3-1 BOAT Constituent List ............................... 2).
A-l 95th ?ercentile Values for the F Distribution ....... 47
Preceding page blank
-------
*****NOTICE*****
THE COMPLETE SET OF BOAT BACKGROUND DOCUMENTS FOR THE SECOND
THIRDS (PROPOSED) HAVE BEEN ASSIGNED THE FOLLOWING EPA/OSW
NUMBERS.
EPA/530-SW-89-017A THROUGH EPA/530-SW-89-017G
EPA/530-SW-89-017I-J
EPA/530-SW-89-017L
EPA/530-SW-89-017M
PLEASE NOTE THAT THERE ARE TWO NUMBERS THAT DO NOT EXIST
IN THE CONSECUTIVE LISTING:
EPA/530-SW-89-017H or EPA/530-SW-89-017K
*****NOTICE*****
/•a
-------
1. INTRODUCTION
The Hazardous and Solid Waste Amendments of 1984 (HSWA) imposed
substantial new responsibilities on those who handle hazardous wastes,
including stringent new restrictions on the land disposal of hazardous
wastes and associated treatment residuals.
This document summarize? EPA's approach for implementing the land
disposal restrictions program. Section 2 presents the legal authority
under whicn the Agency is basing its regulations. Section 3 describes
the technical methodology EPA uses to define treatment standards for
restricted hazardous waste. Section 4 discusses variances from these
treatment standards, and, finally, Section 5 presents the Agency's
approach to the special problems raised by the P and U listed hazardou
waste categories in 40 CFR /261.33.
-------
2. LEGAL BACKGROUND
2.1 General Requirements Under HSWA
The Hazardous and Solid Waste Amendments of 1984 (HSUA), enacted on
November 8, 1984, amended the Resource Conservation and Recovery Act of
1976 in several significant ways. Among other initiatives, the
amendments require the Environmental Protection Agency (EPA) to
promulgate regulations restricting the land disposal of hazardous wastes
according to a strict and detailed schedule. This effort is generally
referred to as the land disposal restrictions program.
In its enactment of HSUA, Congress stated explicitly that "...to
avoid substantial risk to human health and the environment, reliance on
land disposal should be minimized or eliminated, and land disposal.
particularly landfill and surface impoundment, should be the least
favored method for managing hazardous waste" (RCRA section 1002(b)(7).
codified at 42 U.S.C. 6901(b)(7)). Exceptions to the restrictions are
intended to be minimal: all wastes must be treated unless "it has been
demonstrated to the Administrator, to a reasonable degree of certainty.
that there will be no migration of hazardous constituents from the
disposal unit or injection zone for as long as the wastes remain
hazardous" - - the so-calleO "no-migration" demonstration (RCRA section
3004(d)(l). (e)(l). (g)(5). codified at 42 U.S.C. 6924 (d)(l). (e)(l).
(9)15)).
-------
Factors that must be taken into account when granting any exceptions
to this program reflect the basic rationale of the program itself.
Before it can allow a waste to continue to be disposed of in or on the
land, EPA must consider (RCRA section 3004 (d)(l)(A-C):
1. The long-term uncertainties associated with land disposal,
2. The goal of managing hazardous waste in an appropriate manner in the
first instance; and
3. The persistence, toxicity, mobility, and propensity to bioaccumulate
such hazardous wastes and their hazardous constituents.
Consistent with the comprehensive scope of this program, HSWA's
definition of land disposal is broad. Land disposal includes but is not
limited to "any placement of hazardous waste in a landfill, surface
impoundment, waste pile, injection well, land treatment facility, salt
dome formation, salt bed formation, or underground mine or cave" (RCRA
section 3004(k), codified at 42 U.S.C. 6924(k)). The statute does.
however, set different schedules for restricting various categories of
waste from various types of land disposal (see Section 2.2).
HSWA grants the Agency substantial flexibility in designing treatment
standards to implement the program. The standards on require the use of
specific "methods" (technologies), or they can be stated as numerical
performance standards (i.e., required concentration-based i«ve'is of
treatment), as long as they "substantially diminish the toxicity of 'he
*aste or substantially reduce the likelihood of ^miration of hazaraous
constituent: from the waste so that short-term and long-term tnreats to
human health and the environment are minimized" (RCRA sect'oi 300-1 (T, i; j ;.
-------
codified at 42 'J.S.C. 6924 (m)(l)). In exercising this flexibility. EPA
prefers, wherever possible, to establish numerical performance standards
rather than to require the use of specific treatment methods. The Aqc-ncy
believes that concentration-based treatment standards offer the regulated
community greater flexibility to develop and implement compliance
strategies. Such standards also provide an incentive to develop
innovative technologies.
EPA is not required to establish unique standards for each waste
code. In some instances, variations in physical or chemical
characteristics within a single waste code may require the establishment
of multiple treatment standards for that single code. In other
instances, similarities among wastes may allow the Agency to set a single
treatment standard to cover multiple waste codes. Variances from
standards are also possible in certain instances: if a generator can
demonstrate that the standard promulgated for the generator s *aste
cannot be achieved. EPA can revise the treatment standard for that
particular waste through rulemaking procedure:.
I.2 Schedule for Deve'opinq Restrictions
HSWA set a strict and detailed schedule for establishing treatment
standards, based generally en priorities related to the volume and
intrinsic hazards of different types of wastes. Two groups received
early attention: (lj solvent and dtoxin wastes, tc De regulated /cthm 2*
•nontr.s of HSWA' i passage anc \2, the so-cill-:*: "California LI:*' *as •.•=•:.
to be regulated within 32 months. The ioKent Jioxin *ai'e grouc
-------
identified in HSWA includes those solvent wastes covered under waste
codes FOOl. F002, F003. F004. and FOG1:, as well as the dioxin-contaming
wastes covered under waste codes F020, F021. FOZt. and r023 (RCRA
•
3004(e)) .
The California List Bastes, a grouv of wastes originally '•• s'ed jy
the State of California and adopted intact within HSWA, include liquid
hazardous wastes containing metals, free cyanides. PCBs. uorros'-.es (pH
less than or equal to 2.0). and any liquid or nonliquid hazardous watte
containing halogenated organic compounds (HOCs; above 0.1 percent by
Priorities for all other hazardous waste ''ifted under P,CR or'. NovemDer
3. 1986 (Pr.RA section 3004(9) ( 1 )). This schedule require; a'' '.anc
diSPOia' restr : c.t : 0=1 regulat'ons to be in place by May 3. ;990.
Consistent with the requirements of HSWA. [wa divided all other '-. -;:eo
hazardous wastes into three groups (the 'Thirds'), to be regulated •• i
success i /entr. and d:o.< ir.*, : ••mai standards oron'j' ••.•j
• °. -'-TJ l j*. • on i; .0 •?•;". .ID i
-------
• California List wastes: Final standards promulgated on July 8,
1987.
. "First Third" scheduled wastes: Final standards promulgated on
August 8. 1988.
• "Sacond Third" scheduled wastes: Final standards to be
promulgated on or before June 8, 1989.
• "Third Third" scheduled wastes: Final standards to be
promulgated on or before Hay 8, 1990.
2.3 Variances from the Schedule
Tht land disposal restrictions are effective when promulgated unless
the Mminlstrator grants a national variance and establishes a different
date, not to exceed 2 years beyond the statutory deadline, based on "the
earliest date on which adequate alternative treatment, recovery, or
disposal capacity which protects human health and the environment will be
available" (RCRA section 3004(h) (2),. codified at 42 U.S.C. 6924 (h)(2)).
In addition, if EPA fails to set a treatment standard by the
statutory deadline for any hazardous waste in the First Third or Second
Third of the scheduled wastes, the waste may continue to be disposed of
m 4 landfill or surface Impoundment, but only if the facility is in
compliance with the minimum technological requirements specified in RCRA
section 3004 jo). Furthermore, prior to such disposal, generators must
cerffy to EPA that (1) they have investigated available treatment
capicity, (2) t^sy have determined that disposal in a landfill or surface
impoundment is the only practical alternative to treatment currently
available to the generator, and (3) the waste, if currentl> treated, is
Demg treated to the ful'est extent feasible in the existing treatment
-------
• California List wastes: Final standards promulgated on July 8,
1987.
• "First Third" scheduled wastes: Final standards promulgated on
August 8. 1988.
• "Second Third" scheduled wastes: Final standards to be
promulgated on or before June 8, 1989.
• "Third Third" scheduled wastes: Final standards to be
promulgated on or before May 8, 1990.
2.3 Variances from the Schedule
Thf land disposal restrictions are effective when promulgated unless
the Mmlnlstrator grants a national variance and establishes a different
date, not tr> exceed 2 years beyond the statutory deadline, based on "the
earliest date on which adequate alternative treatment, recovery, or
disposal capacity which protects human health and the environment will be
available" (RCRA section 3004(h)(2).. codified at 42 U.S.C. 6924 (h)(2)).
In addition, if EPA fails to set a treatment standard by the
statutory deadline for any hazardous waste in the First Third or Second
Third of the scheduled wastes, the waste may continue to be disposed of
m * Undfill or surface Impoundment, but only if the facility is in
compliance with the minimum technological requirements specified in RCRA
section 3004 (o) . Furthermore, prior to such disposal, generators must
certify to EPA that (1) they have investigated available treatment
capacity. (2) f.^sy have determined that disposal 1n a landfill or surface
impoundment is the only practical alternative to treatment currently
available to the generator, and (3) the waste, if currentl> treated, is
DBinq treated to the ful'est extent feasible in the existing treatment
-------
system. Such continued land disposal 1s allowed until EPA sets a
standard for the waste 1n question, or until May 8, 1990, whichever is
sooner. If the Agency nas not set a standard by May 8, 1990, the waste
Is automatically prohibited from further land disposal.
After May 8, 1990, the only general variance allowed from the land
disposal restrictions standards will be 1f wastes are disposed of in a
land disposal unit that has m&Oe a successful "no migration"
demonstration. These ••lemon 31 rat ions are based on case-by-case petitions
that must show that there will be no migration of hazardous constituents
from the disposal unit for as long as the waste remains hazardous.
-------
3. METHODOLOGY FOR ESTABLISHING TREATMENT STANDARDS
RCRA section 3004(m) specifies that treatment standards must minimize
long- and short-term threats to human health and the environment arising
from land disposal of hazardous wastes. EPA's general approach for
complying with this requirement was promulgated as part of the
November 7, 1986, rule. It is summarized here.
The legislative history accompanying HSWA states that technical
methods used for treating hazardous waste should be "the best that has
been demonstrated to be achievable," but it note*; that Congress's intent
is "to require utilization of available technology" and not a "process
which contemplates technology-forcing standards" (Vol. 130 Cong. Rec.
S9178 (dally edition, July 25. 1984)). The word "achievable," therefore,
does not require the use of experimental or emerging technologies in
developing treatment standards. Rather, the intent of th.« statute is to
base treatment standards on the best technolog'ss commonly in use and
thus reasonably available *^ Siiy generator.
Accordingly, EPA'y treatment standards, usually stated in terms of
concentrations of hazards constituents in treatment residues (such as
sludges, ashes, or wastewaters), ar» generally based on the performance
of the "best demonstrated a^ailab'e technology," or BOAT. This approach
involves the identification of applicable treatment systems for
individual wastes or for groups .if wastes, determination of v
-------
those that are demonstrated and available, and collection of treatment
data from representative well-designed and well-operated systems to serve
as the basis for numerical performance standards.
The approach ensures that performance standards are achievable In
practice using available technology, but it does not specifically mandate
the use of any particular technology In order to comply with the
standard. Treaters are free to use any method they choose, as long as
thi results are equal to, or better than, use of the model BOAT
/
technology.
3.1 Waste TreaMbllltv Groups
To set standards efficiently, wastes are clustered Into "treatability
groups" that are similar with respect to various parameters that might
affect the success of treatment. These parameters can Include such
factors as physical state, water concentration, presence of nonhazardous
contaminants, organic content, heat content, pH. and so forth. As noted.
waste treatability groups can include multiple waste codes, single waste
codes, or subcategorles of a single waste code. 1n any combination.
The process of establishing and refining treatability groups is a
continuous one within the standard-setting process. Tentative groupings,
such as those presented for the P and I) wastes in Section 5 of this
report, may therefore change before promulgation of the final standards.
The general concept or a treatability group, however, is essential to
identification of BOAT for any given waste code or any subcategories of a
waste code.
-------
3.2 Determining BOAT for Individual Waste Treatabilitv Groups
For any particular waste treatabllHy group, EPA first identifies
applicable technologies either through literature reviews cr on the basis
of information provided by facilities currently treating the waste or
Similar wastes. In some instances, technologies used to separate or
otherwise process chemical or other materials are clearly applicable to
waste treatment and may therefore form the basis of a standard, because
certain wastes are similar to raw materials processed in various
industrial applications.
From among the applicable technologies, EPA then identifies those
that are "demonstrated" for the particular treatabiluy group. To be
considered demonstrated, a technology must be used in a full-scale
operation for treatment of the waste or a similar waste. Where the
Agency does not Identify any facilities treating specific wastes from a
particular group, it may "Tansfer" a finding of demonstrated treatment
by comparing the parameters that affect treatment of the target waste
group to parameters of other waste groups for which demonstrated
treatments are known. For example, on the basis of technical literature
and data collected by the Agency. EPA considers rotary kiln incineration
to be a demonstrated technology for many waste codes containing hazardous
organic constituents, high total organic content, and high filterable
solids content, regardless of whether any facility is currently treatiig
these wastes.
10
-------
The next step is to determine which of the demonstrated technologies
is "best" for the purposes of establishing BOAT. In defining "best," EPA
considers only the effectiveness of treatment—the degree to which
hazardous constituents in the waste are removed or destroyed. Economic
factors are not considered under RCRA.
If only one technology is demonstrated for a particular waste group,
then that technology is automatically "best," even if no acceptable data
are available to measure its performance. If two or more technologies
are available, but acceptable data exist for only one of them, then the
Agency must make a judgment as to whether to develop new data or to use
engineering judgment to determine if the performance of the documented
technology is likely to be equal to, or better than,, the others. If
several technologies are available, each with acceptable performance
data, then the Agency compares the performance of these technologies
using their available daia.
Any such comparisons must be statistically defensible to the extent
that sample sizes and other technical factors permit. First, prior to
performing statistical tests, the Agency must adjust the measured results
to account for the accuracy of the laboratory procedure used to generate
the data (see the detailed discussion in Section 3.3). Second, where
possible it may compare the adjusted performance levels using the
statistical "analysis of variance" (ANOVA) technique to ensure that the
technology selected as "best" does indeed perform statistically better
than the others (see Appendix A-2. "F Value Determination for ANOVA
11
-------
Test"), if the differences among the available data sets are
statistically significant, then two or more technologies can be
considered as be>ng "best demonstrated."
Next, the Agency determines whether the best demonstrated technology
or technologies is "available." "Available" technologies must be both
commercially available and provide "substantial treatment." To be
considered commercially available, the technology may be either a common
technology in universal use (such as neutralization or incineration), or
a proprietary or patented process that can be purchased or licensed from
the proprietor or that is commercially available at a facility offering
use of the technology for a fee.
To be considered as providing "substantial treatment." a technology
must, consistent with the language of HSWA, "substantially diminish the
toxicity" of a waste or "substantially reduce the likelihood of migration
of hazardous constituents" from the waste (section 3004(m)). By
establishing that treatment is "substantial," the Agency both ensures
compliance with statutory objectives and avoids requiring a treatment
•method that provides little or no environmental benefit.
Treatment will always be considered substantial if the residuals from
treatment contain nondetectable levels of the hazardous constituents of
concern, [f concentrations are detectable, then a finding of substantial
treatment must be made on a case-by-case basis, considering the following:
• Number and types of constituents treated.
• Performance (concentration of the constituents in the treatment
residuali). and
12
-------
• Percent of constituents removed.
EPA has used both total constituent concentration and TCLP analyses
of the treated waste as a measure of technology performance and for later
establishment of treatment stand^rds. For organic constituents, EPA
measures performance based on the total constituent concentration found
in the treated waste. This is because technologies exist to destroy
various organic compounds in waste, making the total amount of
constituent left in the treated waste the more logical measure of
*
performance. For all metal constituents, EPA is using total
constituent and/or the TCLP as the basis for evaluating performance.
When BOAT involves a metals recovery operation, EPA uses both total
concentrations and TCLP to measure performance, because it is important
to establish both the effectiveness of recovery (measure by changes in
total concentration) and the stability of any treated residuals that ma*
be sent to land disposal (TCLP test of the residuals). When BOAT for
metals involves only treatment, that treatment is generally
stabilization, so the appropriate measure of performance of stabilization
is the TCLP test.
EPA's land disposal restrictions for solvent waste codes F001-FG05
(51 FR 40572) use the TCLP value as a measure of performance. At the
time that EPA promulgated the treatment standards for these wastes.
ustful data were not available on total constituent concentrations ir
treated residuals and, as a result, the TCLP data were considered to
be thj best measure of performance.
13
-------
If, for any reason, a demonstrated or available technology cannot be
defined for a particular waste treatablllty group, EPA cannot establish a
treatment standard for that waste group. Wastes in the group would then
be prohibited from continued land disposal, unless managed in accordance
with the exemptions and variances discussed above (especially the
no-migration standard). EPA is committed, however, to establishing new
treatment standards as soon as new or improved treatment processes become
demonstrated and available.
3.3 Establishing Numerical Performance Standards on the Basis of BOAT
Once BOAT is determined for a particular treatability group, EPA
prefers, wherever possible, to define numerical performance •standards in
terms of concentrations of hazardous constituents In the treated waste
and in any •"• '-jals that might be produced. For example, for wastes for
which BOAT is incineration, the Agency may have to define maximum
allowable concentrations of hazardous constituents in associated bottom
ash, scrubber water, and possibly sludges resulting from treatment of
scrubber water.
EPA develops treatment standards using performance data gathered from
r«presentativt facilities. Only data from well-designed and
wel1-operated facilities are accepted as usable--a judgment made on a
case-by-case basis for each set of potentially usable data. Data need
not be generated only by EPA; the Agency may use data submitted by
industry, provided these data are shown to be from a wel1-designed and
-------
well-operated facility and were generated using adequate quality control
and quality assurance procedures. (EPA's policies and procedures for
collecting new oerformance data, where needed, are discussed in
Section 3.4.)
3.3.1 Evaluating the Adequacy of Existing Data
All valid data available to the Agency may be used to establish
BDAT-based performance standards. They may be generated by EPA or its
contractors, by research organizations or universities, or by industry.
Whatever the source, however, all the data underlying all performance
standards must mest explicit standards of quality assurance and quality
control. If the available data for a given technology/waste group
combination are not of adequate quality, then data can be "transferred"
from another standard if they meet certain conditions. These issues are
discussed separately below.
(1) Criteria for accepting existing data. EPA considers a number of
factors in evaluating data sets as the possible basis for BOAT
standards.
1. All data must come from technologies that are BOAT.
2. The facility from which the data were generated must be wel1-designed
and wel1-operated. Adequacy of design can generally be determined
through review of facility specifications; the essential requirement
is that the facility include all processes needed to handle the
hazardous constituents in the target waste group, as well as all
nonhazardous constituents that could affect the system s performance
in treating the hazardous constituents. Adequacy of operation must
be determined based on a review of the operating parameters used
during the sampled test.
3. EPA reviews the adequacy of the quality assurance and duality control
protocols followed in generating the data. If these protocols are
substandard or nonexistent, the data are discarded.
15
-------
4. All candidate data sets must use measures of performance consistent
with those being used to set the standard (TCLP versus total
constituent concentrations).
5. For a data set to be accepted in whole or in part, the data must show
substantial treatment on a constituent-by-constituent basis. Data
must be provided for both untreated and treated concentrations.
Treated concentrations mu.st be lower than untreated concentrations.
Often common sense will suffice to determine whether the degree of
treatment provided is significant, but, if necessary, statistical
tests can be used to arrive at a formal finding of substantial
treatment.
6. Data on concentrations ;,n treated waste must be adjusted for accuracy
using recovery factors specific to the laboratory tests used (see
Appendix A-1).
In situations where the available data show substantial treatment fcr
one class of constituents but not for another, the Agency may conclude
that the standard should be based on a treatment "train" of multiple BOAT
technologies operating as a system. This might be the case, for
instance, in treating wastes that include both organics and metals.
Incineration might show substantial treatment of the organics, but not of
the metals, which might require another form of treatment, such as
stabi1ization.
(2) Transfer of treatment data or standards. In some instances. EPA
is proposing treatment standards that are not based on tests of the waste
in question by the selected BOAT technology. It may do this when it
determines that the constituents present in the subject waste can be
treated to the same performance levels as those observed in other wastes
for which EPA has previously developed treatment data. EPA believes such
transfers may be technically valid in cases where the untested wastes are
16
-------
generated from similar industries or from similar processing steps, or
have similar waste characteristics affecting performance and treatment
selection.
Transfer of treatment standards to similar wastes or to wastes from
similar processing steps requires little formal analysis. It is based on
a detailed comparison of the constituents of concern m the untested
waste to those in the tested waue. If the parameters that affect
treatment performance for these constituents indicate that the untreated
waste is equal to. or easier to treat than the tested waste, tnen the
transfer can be made.
3.3.2 Collecting Additional Performance Data
!f adequate data are not available for use in setting a performance
standard, the Agency collects additional data through a standardized
sampling and analysis procedures at existing facilities that generate or
treat the wastes of concern.
(1) Identification of facilities. EPA uses a number of sources to
identify candidate sources for sampling and analysis. These include
Stanford Research Institute's Directory of Chemical Producers; EPA's
Hazardous Waste Data Management Systems (HWOMS); the 1986 Treatment.
Storage, and Disposal Facility (TSDF) National Screening Survey, and
EPA's Industry Studies Data Base. EPA also contacts trade associations
to solicit help in identifying appropriate facilities.
To the extent possible. EPA prefers to develop dat i from treatment
facilities handling only a single waste, believing that facilities that
routine!/ treat a specific watte have had the beet opportunity to
17
-------
optimize design and operating parameters. It. also prefers to avoid
ambiguities created by the mixing of wastes before and during treatment
If several facilities appear to be equally desirable for sampling. EPA
selects sites strictly on the basis rf which facilities can be most
expeditiously visited and, if justified, sampled.
Wherever possible, the Agency will evaluate treatment technologies
using commercially operated systems. If performance data from properly
designed and operated commercial treatment methods for a particular
waste, or a waste judged to be .-similar, are not available. EPA may use-
data from research facilities, whenever research facility data are used.
EPA will explain in the preamble and background document for the rule in
question why such data were used, and w'V -equest comments on the use of
sucn data.
(2) Engineering site visit. Once a facility is selected. EPA visits
the site to confirm that it is wel1-designed and that the necessary
sampling points are accessible. In general, EPA considers a
wel1-designed facility to be one that contains the unit ooerations or
unit processes necessary to treat the various hazardous constituents of
the selected waste, as well as to control other nonhazardous materials in
the waste that .nay affect treatment performance. During the visit, the
Agency also confirms that the facility appears to be wel1-operated:
actual operation of the treatment system during sampling, however, is the
baiis for determining the proper operation of the treatment unit.
13
-------
(3) Sampling and analysis plan. If. after the engineering •;: te
visit, fPA decides to sample a particular facility, it :n\} then develop
a site-specific Sampling and Analysis Plan (SAP) following the Generic
Quality Assurance Project Plan for the Land Disposal Restriction Program
("BOAT") (EPV503J-SW-87-011). The SAP is based on sampling procedures
discussed with plant personnel during the site visit. It describes *nere
the sampling will take place, how the samples will be taken, the
frequency of sampling, the constituents to be analyzed and the methods or
analysis, the operation parameters to be obtained, jnd specific
laboratory quality control checks on analytical results, "he SAP is sent
to the plant for review and comment.
Facilities wishing to submit data independently for consideration \n
the development of BOAT standards should, to the extent possible, fo'low
the procedures described in EPA/5030-SW-87-G11.
(•«) Sampling visit. While actually sampling a particular fac ••!••./
[PA attempts to collect sufficient samples of the untreated waste and any
solid or liquid residuals so that the inherent variability of *.ne process
car. De properly reflected m the final performance standard. !t also
documents the operating conditions that existed during the waste
treatment period. To the extent practicable (and within safety
constraints). EPA or its contractors collect :he samples themselves ano
follow cha in-of-custody procedures to ensure fhat the integrity of '.he
Oita •; maintained. Any deviations from tne SAP. iuch .1;, those r.hi'.
•n-gnt be necessary because of plant upset;, or :nanges m fa^i'-r.,
j;}•?•• i' ion . ar'e note-J.
19
-------
(5) Onsite Engineering Report. EPA summarizes all its data
collection activities, associated analytical results, and any deviations
from planned procedures in a formal Onsite Engineering Report. After
review by the plant, the report, except for material claimed by the plant
as confidential, is made available to the public.
3.3.3 Hazardous Constituents Considered for Regulation
The list of hazardous constituents for which BOAT performance standards
may be established is known as the BOAT Constituent List. This list,
provided as currently amended in Table 3-1, Is a subset of the
constituents listed in 40 CFR Part 261, Appendix VIII; it also includes
several ignitable constituents used as the basis for listing wastes as
F003 and FOOS. Chemicals are listed in Appendix VIII if they are shown
in scientific studies to have toxic, carcinogenic, mutagenic, or
teratogenic effects on humans or other life forms; for instance, they
include SUCK substances as those identified by the Agency's Carcinogen
Assessment Group as being carcinogenic.
There are five major reasons why not all Appendix VII and VIII
constituents or the Ff)03 and FOOS ignitables are included on the BOAT
Constituent List:
1. A constituent mav be unstable. Based on their chemical structure.
some constituents will either decompose in water or will ionize. For
example, maleic anhydride will form maleic acid when it comes in
contact with water, and copper cyanide will ionize to form copper and
cyanide atoms. EPA may, however, choose to regulate the
decomposition and ionization products.
20
-------
T«bl« 1-1 BOAT ComtUumt List
BOAT
fmfmrmmrm.
nti.
227.
1.
2.
3.
4.
5.
6.
223.
;.
8.
9.
10.
11.
12.
13.
14.
15.
18.
17.
18.
19.
20.
21.
22.
23.
24
25
26.
27.
28.
29.
224.
225.
226.
30.
227.
31.
214
32
33.
228.
34
CoMtttuvn
Volattte onunici
ACOtOM
Antoinette
Aerate tic
Acrytaftttrlto
Bmim
Brand Ich lorasgthan*
BrOBBtthMM
n-Butyl alcohol
Carton Utrachlortda
Carton dttulfld*
Chlorobwucn*
2-Chloro- 1 . 3-butadtww
Qilorodl brans* thana
Ch loroathano
2-Chlorocthyl vinyl »th»r
Chlorofoni
Ch loroBthww
3-Oiloraprapww
1 .2-OlbraB-3-chloraprap«M
1.2-Oibn»Mth«w
OibnamxsUMiw
trwt*- 1 . 4-0 Ich loro-2-but«n>
Otch lorodlf luoraxthem
1.1-0 ich loro«th*n«
1 . 2-0 ich loroathww
1.1-0 Ich lonwthylm
tr«ns- 1 ,2-Olchloro*th«n«
1 . 2 - 0 1 ch 1 oropropww
trans- 1 . 3-0 ich loropropcna
c » s - 1 . 3 -0 Ich loropropmc
1.4-Dloun*
2-Ethoxyvthtnol
Ithyl
-------
2374«
Ubla 3-1 (ContinuKt)
BOAT
rafaranca
na.
m
35.
37.
38.
230.
39.
40.
41.
w.
43.
44.
45.
46.
47
48.
49.
231.
50.
215.
216.
217
51.
S2.
53.
54
55.
56.
57.
58.
59
218.
60.
61.
62.
>3.
64.
65
66
Conatltuant
VoUtlla oraMitci (continuad)
Nafbyl t totality 1 katona
Htthyl aathacryUta
Ntthacrytonttrt la
Mathylana chlorite
2-HltropropMM
Pyrtdlna
l.l.l.2-Tatr»chloroath«na
1 . 1 .2.2-UtrachloroaUuna
Tat rach loroathana
Toluana
TribroaorMthana
1.1.1-TrichlorMthana
l.l.2-Tr
-------
2374,
r«bl« 3-1 (ConilitiMd)
BOAT
rvfvrwica
no.
67.
U.
M.
70.
71.
72.
73.
74.
75.
78.
11.
78.
79.
80.
81.
a.
232.
83.
84.
85.
86.
87.
88.
8*.
90.
91.
92.
93
94
95.
98.
97
98.
W.
100.
101
102
103.
104.
105.
IOC.
219.
Carat ItuMt
SaHlVoIjtflf flfUntfil frnn* tnnarl)
6t«(2-c«iloro«thaxy)>*th*iw
ais({-chlorwthy 1 )tth»r
Bl»(2-«hlonsi«oprfloyl)«th«r
Sll(2-«thylhsMy l)pMhl Ut«
4-Brianffh«iiyl phwyl tthtr
Butyl tansy 1 pMlwItU
?-Me-Butyl-4.t-dtnitrav»«nel
0-Oiloro*«llln«
Chlorotaiut Ut«
^CJi ioro • crvso 1
2 -Ch loroMphth* l«w
2-Oilereiph««wl
3-Oileropraplanttrt 1«
Chrytan*
ortho-CrMo 1
IMra-CrMa)
Cyc lataMnont
0 « ban* ( « . h ) «nthr tear*
OllMiuo(«.«)pyr«M
OltaMUo(«, l)pyniw
•-Otchlorobwmn*
o-0\chlorotenz*nt
p-0 ten lorotwucn*
3.3>-Olchloratafuldlna
2.4-Olchlorophtnol
2.8~0ich)orophonol
Oicthyl pdth«Uu
1 . 3 ' -0 l«Btho«ybwif Id in*
p • 0 !•• t hy l«i t noixotafUtn*
3 . 3 ' -Chmthy Ibaiu id in*
2.4-OlMthy1p»vnol
OiKBthyt phtfuUtt
01-n-tautyl phthaUtt
l,4-0mitro6snxcn«
4.6-Oinitro-o-crMol
2.4-OlnUroph«nol
2.4-Olnitroto1u«n«
2.(-Omitrotoluw«i
Ol-n-octyl pnth«Ut«
Ol-n -p>*opy InitmtdMin*
Olp»wmy lj«ln«
Oiph«nylnitroi«iin*
CAS no.
111-91-1
111-44-4
39638-32-9
117-81-7
101-55-3
8S-U-7
88-85-7
106-47-8
510-15-8
59-50-7
91-58-7
95-57-8
542-78-7
218-01-9
95-48-7
108-44-5
106-94-1
53-70-3
192-85-4
188-55-9
541-73-1
95-50-1
106-48-7
91-94-1
120-83-2
87-65-0
84-88-2
119-90-4
80-11-7
119-93-7
105-87-9
131-11-3
84-74-2
100-25-4
534-S2-1
51-28-5
121-14-2
606-20-2
117-84-0
621-64-7
122 39-4
86-30-6
-------
aa*r
rtftnmct
no.
107.
100.
10*.
110.
111.
112.
113.
114.
US.
118.
117.
118.
119.
120.
36.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
I3S.
136.
137
130.
139.
140.
141.
142.
220.
143
144.
14i.
146.
Carat ituHit
S«itYoUtll« oromici (continued)
1.2-0tptanylhydr«itiw
FllMTMtlWM
MuOTMM
Hnucti Sorobouww
H»»*ch lorobuUdliwi
touch loracys)apMit«dt«M
HtMchloracUwii*
(touch loroptwn*
(touch loraprapOT*
1 ndmo ( 1 . 2 . 3 -cd ) pyrww
iMWjfrol*
N»th*pyrll«M
3-Nithy Icho l«nthr«n«
4,4'-N>thyl«Mb)t
!?-chloro*ni ltn«)
Nlthyl ••thMtnuiroMt*
lUphthi l«n«
1.4-KiphthaqulnoM
l-IUp»ithyl«iin«
2-IUphtnyl«Hn«
p-NltroMtllM
DltrotMnicn*
4-Nltraphmol
H-Nltrosodl-n-butylaaln*
N-Nltrt>Mdl«thyl*Bin«
N-N I tro«odiaethy iMiin*
»-» 1 trtxowthy Icthy Ijaina
N-NltrtxaKirphol In*
N-NitroMpip«ridln*
N-Nitrotopyrrol idin*
S-Nitra-o-to luidinc
P»nt jch lorotenf •«•
P«flt*ch loro«th«n*
P«it*ch laron i trob«nx*n«
P«nt«ch loroptwno 1
Ph«n*cft In
Ph«n«ntnrtna
Ptvnol
Phthtlic «nny4rid>
7-Plcohn*
Pron«ni(te
Pyrtn*
Reiorcmal
CAS no.
122-66-7
206-44-0
M-73-7
118- 74- 1
87-68-3
77-47-4
67-72-1
70-30-4
1868-71-7
193-39-S
120-S8-I
91-80-5
M-49-5
101-14-4
66-27-3
91-20-3
130-1S-4
134-32-7
91-S9-8
100-01-6
98-9S-3
100-02-7
924-16-3
55-18-5
B2-/5-9
I059I-9S-6
59-89-2
100-75-4
930-55-2
99-65-8
608-93 S
76-01-7
82-68-8
87-86-5
62-J4-2
85-01-8
108-95-2
85-44-9
109-06-8
23950-58-5
129-00-0
108-46-3
-------
T«bl* 3-1 (Continued)
BOAT
r«f«r«nc«
fW
147.
I4a.
149.
t £ A
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
221.
160.
161.
162.
163.
164.
165.
IM
• W i
187.
168.
169
170.
171.
1/2.
171.
1/4.
1/5.
Conttttuant
SaatvaUttte oraMlcs (continued)
Ufral*
1 . 2 . 4 . 5- T«t r«ch lorabMum
2.J.4.l-T«tr«cnlorophMw1
1 7 A-TrtHllMl^^M*MM
liCi^ir i wi iwvBMwHH
2.4.5-TrtcMorapnMol
2.4.S-Trlchlora0«nol
Trtt(2.3-dtbragDprepyl)
ptosphtt*
HiUlt
Antlaony
Arvanic
ttriua
ft^r .. 1 1 1.^
QVry 1 1 IIM
^AJ^Mll^
I^OHIUB
ChnniuB (tot*l)
Chrmlua (huu««l*flt)
i"fMM^A^
Lap|wr
1 B\*tf4
LMQ
Nsrcury
• lr+m \
B 'CUV 1
• » IMM ii^
*9 IBn 1UM
5 1 lv«r
rtMihtw
U »• —4 >.^
TCMU llfll
* 4_..
4 inc
Inonunici oth«r th*n mttili
Cytnid*
Fluor id*
Sulf id*
Orti«nochlortn« Miticidn
Aldrin
-------
2374$
1-1 (Continued)
r«f«r«nc* CoratitiMnt CAS no.
no.
Oloami ind furint
207. touch toracitfaMUo-p-dloi In*
208.
209.
210. NnUchlorodtbMUofurun
211. T«trachl«radlbtn«»-|>-dto«lM
213. 2.3.7,8-T«tr»ctiloTOdib«nw-p-dioxin 1746- 4
-------
2. EPA-approved or verified analytical methods are not available. Many
constituents, such as 1,3,5-trinitronobenzene, are not measured
adequately, or even detected, using EPA's analytical methods as
published in SW-846 Third Edition.
3. The constituent is a member of a chemical group desiqm ed in
Appendix VIII as "not otherwise specified" (N.Q.S.i. Constituents
listed as N.O.S., such as chlorinated phenols, are a generic group of
some types of chemicals for which a single analytical procedure is
not available. For each N.O.S. group, a representative sample of
those constituents that can be readily analyzed are included in the
BOAT Constituent List.
4. Available analytical procedures are not appropriate for a complex
waste matrix. Some compounds, such as auramine, can be analyzed as a
pure constituent, but the recommended analytical method may not
positively identify it in the presence of other constituents, such as
in a complex waste matrix.
5. Standards for analytical instrument calibration are not commercially
available. For several constituents, such as benz(c)acridine,
commercially available standards of an adequately pure grade are not
available.
The BOAT Constituent List is a continuously growing list that does not
preclude the addition of new constituents as the problems above are
resolved. The initial list was published in EPA's Generic Qua!itv
Assurance Plan (EPA/530-SW-87-011); since then, 18 additional
constituents have been added. Two constituents (fluoride and sulfide)
are not specifically included in Appendix VIII, but are included in the
BOAT list as indicators for certain Appendix VIII constituents, such as
hydrogen fluoride and hydrogen sulfide, which ionize in water.
3.3.4 Selecting Constituents for Inclusion in the Standard
A performance standard for treating a particular waste group will
list acceptable concentrations of BOAT list constituents in treated
residuals. The standard will not necessarily include all BOAT list
28
-------
constituents analyzed in a particular waste stream, and may, in some
instances, include one or more BOAT list constituents that have not been
detected in the waste stream. The rationale for selecting constituents
for inclusion in a standard is as follows.
The constituents considered for regulation in each treatability group
are, in general, those for which available data show statistically
significant reductions in concentrations resulting from treatment. This
process involves the use of the statistical analysis of variance (ANOVA)
test described in Appendix A-2. EPA interprets a statistically
significant reduction in concentration as evidence that the technology
actually "treats" the waste.
There are some instances where EPA may regulate constituents. that are
not detected in the untreated waste but which are detected in the
analyzed residual (ash, sludge, etc.). This may happen, for Instance,
where the presence of other constituents in the untreated waste matrix
interferes with quantification of the constituent of concern. The result
may be a finding of. "not detected," when in fact the constituent is
present in the waste. EPA may also choose to consider a constituent not
found in a sampled untreated waste if it believes that the constituent is
likely to be present in the same hazardous waste generated by another
source. For example, EPA may choose to regulate all conceivable
hazardous solvents that might be used in paint or ink manufacture, even
if its sample daU do not include them all.
29
-------
EPA then reviews the list of candidate constituents to determine
whether any can be excluded from regulation because they /ould be
indirectly controlled by regulation of other constituents. Tor instance,
an incineration regulation might choose to regulate only the least
combustible organic compounds present in the waste, since achievement of
a standard for these compounds would assure achievement of adequate
treatment for the others. This approach (1) reduces analytical cost
burdens on the treater and (2) facilitates implementation of the
compliance and enforcement program.
3.3.5 Calculation of Standards
The final step in setting a performance standard is to define the
maximum acceptable constituent levels in treatment residuals for the
selected BOAT-list constituents for a particular waste treatability
group, based on the performance of the .BOAT technology. This is done by
multiplying the average treatment value observed in the acceptable
available data by a factor known as the "variability factor."
Only data developed at well-designed and well-operated systems are
used to calculate performance standards. Parts or all of the available
data sets may be discarded on a case-by-case basis. For instance, if the
residence time for a waste during a particular test run was substantially
shorter than the planned value, EPA might conclude that the system was
not properly operated during that run and would discard the associated
treatment results in calculating average treatment efficiencies.
30
-------
The variability factor used to calculate performance standards takes
into account that even well-designed and well-operated treatment systems
will experience some fluctuations in performance. These fluctuations may
result from inherent mechanical limitations in treatment control systems,
treatability variations caused by changing Influent loads, unavoidable
variations in procedures for collecting treated samples, or variations in
sample analysis. Setting treatment standards using a variability factor
should therefore not be viewed as a relaxation of section 3Q04(m)'s
requirements, but rather as a response to normal variations in treatment
processes. As a practical matter, facilities will have to be designed to
meet an average level of performance that is more stringent than the
standard in order to ensure continuous compliance with the standard.
EPA calculates the variability factor for each selected constituent
"f concern using the statistical methods described in Appendix A. The
equation is the same as that used for the development of numerous
regulations in the Effluent Guidelines Program under the Clean Water
Act. It sets the standard at the upper 99th percentile value
concentration of the constituent expected in the treatment residual.
using the mean and standard deviation calculated from the acceptable
available data, and assuming that performance varies lognormally.
There is an additional step in the calculation of the treatment
standards in those instances where the ANOVA test shows that more than
one technology achieves a level of performance that represents BOAT, in
such instances. EPA first averages the mean performance /alue for each
technology for each constituent of concern, and then -nultipl-.es tnat
-------
value by the highest variability factor among the technologies
considered. This ensures that all BOAT technologies used as the basis of
the standard will achieve full compliance.
3.4 Compliance with Performance Standards
Compliance with a treatment standard requires only that the treatment
levels specified in the standard for each treatment residual be achieved
prior to land disposal of t.iese residuals. Performance standards do not,
as noted above, require the use of any particular treatment method or
technology. Dilution is prohibited as a means of complying with the
standard: wastes that inherently meet the standard when generated may be
land disposed.
Measurements of compliance with the standard should use the same
procedures as those used to define BOAT, i.e., they can be based either
on total constituent concentrations or on TCIP analyses of the treated
waste.
3.5 BOAT Treatment Standards for "Derived-from" and "Mixed" Wastes
"Oerived-from" wastes are wastes generated in the course of treating
a listed waste. "Mixed" wastes are wastes generated by mixing of
multiple listed wastes, or by mixing of * listed waste with a
nonhazardous waste, in which case the mixture is considered to be
entirely the listed waste. These classes of wastes raise several special
issues.
3.5.1 Wastes from Treatments Generating Multiple Residues
Where the treatment technology or technology on which a performance
standard is based generates residues incidental to treatment, these
32
-------
residues may be hazardous wastes in themselves and may require treatment
prior to ultimate land disposal. The following considerations are
relevant to such derived-from wastes:
1. Al1 of the residues from treating the original listed wastes are
likewise considered »o be the listed waste by virtue of the
derived-from rule contained in 40 CFR 261.3(c)(2). Consequently, all
of the wastes generated in the course of treatment would be
prohibited from land disposal unless they satisfy the treatment
standard, or unless they meet one of the exceptions to the
prohibition on land disposal.
2. The Agency's proposed treatment standards generally contain a
concentration level far wastewaters (less than 1 percent tota[
organic carbon and less than 1 parcent total suspended solids ) and
a concentration level for nonwastewaters (more than 1 percent total
organic carbon and more than 1 percent total suspended solids). The
treatment standards appl> to aT of the wastes generated in treating
the original prohibited waste. Thus, all nonwastewaters generated
from treating these wastes would have to meet ths treatment standard
for nonwastewaters; all wastewaters would have to meet standards for
wastewaters. EPA wishes to make clear that this approach is not
meant to allow partial treatment in order to comply with the
applicable standards.
3. The Agency has not performed tests, in all cases, on every waste that
can result frorr every part of a treatment process or system.
However, EPA's standards are generally based on treatment of the most
concentrated form of the waste. Consequently the Agency believes
that the less concentrated wastes generated in the course of
treatment will also be able to be treated to meet this value.
3.5.2 Mixtures and Other Derived-from Residues
Residues may occur from other types of management than treatment. An
example is contaminated soil, or leachate. that is derived from managing
the waste. In these cases, the mixture is still considered to be the
The term "total suspended solids" (TSS) clarifies EPA s previously
used terminology of "total solids" and "filterable solids." Total
suspended solids is measured \« Method 209c (Total Suspended Solids
Onpd at 103 to 105 C) in Standard Methods for the Examination of
Water and Wastewater (APHA. AWWA. ind WPCF 19°,5).
33
-------
listed waste, either because of the derived- from rule (4C CFR
261.3(c)(2)(i)) or the mixture rule (40 CFR 261 . 3(a) (2) ( H i ) and (w)) or
because the listed waste is contained in the natrix (see, for example,
40 CFR 261.33(d)). The prohibition for the particular listed waste
consequently applies to this type of waste.
EPA believes that the majority of chese residues can meet the
treatment standards for the underlying listed wastes, with the possible
exception of contaminated soil and debris, for which the Agency is
currently investigating whether it is appropriate to establish a separate
treatability subcate^ory. For the most part, these residues will be less
concentrated than the original listed waste and therefore easier to treat
to specified performance levels. In addition, the standards make a
generous allowance for process variability by assuming thit all
treatability values used to establish the standard are lognortnal 1 y
distributed. The waste may also be amenable to a relatively nonvanaDle
form of treatment, such as incineration.
Finally, and perhaps most important, the rules contain a treat ahi ' ' ty
variance that allows petitioners to demonstrate that a particular *aste
cannot be treated to the level specified in the rule (40 CFR 268.4.\(aj.
see ^Iso Section 4) This provides an opportunity to Demonstrate tne
appropriateness of different standards for unusual waste matrices.
3.5.3 Residues from Managing Listed Wastes or Wastes that Contain
L ' sr.so Wanes
' from managing scheduled wastes (tne First, 'jtjcnno. jna 7r. •. r-
Tnir-o «asi?>) the listed California LI it wastes, ano .pen1, ;j' .«n!. ano
-------
dloxin-containlng wastes are all considered to be subject to the
prohibitions for the underlying hazardous waste. Residues from managing
California List wastes likewise are subject to the California List
prohibitions when the residues themselves exhibit a characteristic of
hazardous waste. This determination stems directly from the derived-from
rule in 40 CFR 261.3(c)(2) or, in some cases, from the fact that the
waste is mixed with, or otherwise contains, the listed waste. The
underlying principle stated in all of these provisions is that listed
wastes remain listed until delisted.
The Agency's historic practice in processing delisting petitions that
address mixing residuals has been to consider them to be the listed
waste, and to require that delisting petitioners address all constituents
for which the derived-from w*ste (or other mixed waste) was listed. The
language in 40 CFR 260.22(b) states that mixtures or derived-from
residues r.\r\ be delisted provided a delisting petitioner makes a
demonstration identical to that which a delisting petitioner would make
for the underlying waste. Consequently, tnese residues are treated as
the underlying listed waste for delisting purposes. The statute likewise
takes this position, indicating that soil and debris that are
contaminated with listed spent solvents or dioxin wastes are subject to
the prohibition for these wastes, even though these wastes are not the
originally generated waste, but rather are a residual from its management
(RCRA section 3004(e)(3)). It is EPA's view that all such resioues are
covered by the existing pronibitions and treatment standards for the
35
-------
listed hazardous waste that these residues contain and from which, they
are derived.
3.6 Qther Types of Standards: Technology-Based and "No Land Disposal"
In some circumstances it is not possible to develop
concentration-based performance standards. The Agency then has two
options: to set a standard based on a specific treatment method or to set
a standard of "no land disposal."
If no acceptable data are available for a particular waste or waste
group using the BOAT technology Identified for that group, the Agency may
decide to establish a technology-based standard rather than a performance
s?andard. This may happen when it is not technically possible because of
interferences to measure the constituents of concern in the specific
waste matrix Involved .
A standard establishing a standard of "no land disposal" for a waste
groLp simply prohibits any further land disposal of that group. It sets
no performance standards for treatment. "No land disposal" is
appropriate when:
1. The waste group in question is no longer generated.
2. None of the waste currently generated is land disposed.
3. The technology exists for total recycling of wastes in the group.
For any wastes having a proposed treatment standard of "No Land
Dispo'-.l", EPA solicits comments on the potential for disposal of that
particular waste. EPA is especially concerned with such standards
because, once promulgated, these standards make it illegal to land
dispose these wastes. Should it be revealed after promulgation of the
36
-------
"No Land Disposal Based on No Generation" treatment standard that these
wastes are being generated and land disposed, the generator may apply for
a variance from the treatment standard. The variance petition should
clearly Indicate that the waste 1s being generated contrary to EPA's
original assessment, and should present treatment data to be used tc
establish a new treatment standard (40 CFR 268.44). However, during the
period the variance is being processed, the waste may not be land
disposed, notwithstanding the inaccuracy of the original assessment that
the waste was not being generated. Should commenters provide information
that one or more of the premises used to determine the "No Land Disposal"
standard are not valid, the treatment standard may not be finalized and
land disposal of the waste is usually subject to the "soft hammer"
provisions. Prior to May 8, 1990, EPA Intends to develop and propose
treatment standards for these wastes. If no specific comments are
received refuting the validity of the basis for the "No Land Disposal"
standard, EPA generally proceeds with the promulgation of the standard as
proposed.
The "No Land Disposal" standard does not imply that the waste is so
extremely hazardous that it can not be safely land disposed or handled;
rather, it means that there is no need to land dispose the waste because
alternative forms of management exist.
37
-------
4. VARIANCES FROM THE BOAT TREATMENT STANDARD
EPA recognizes that unique wastes may exist that cannot be treated to
the level specified as the treatment standard. A particular waste may,
for example, be significantly different from wastes considered in
establishing treatabillty groups because it might be mixed with other
waste streams through spills or other Inadvertent mixing. This might
alter Its treatabillty such that it cannot meet the applicable treatment
standard. In such a case, a generator or owner/operator may submit a
petition to the Administrator requesting a variance from the treatment
standard.
Variance petitions must demonstrate that the treatment standard
established for a given waste cannot be met. This demonstration can be
made by showing that attempts to treat the waste by available
technologies were not successful, or by performing appropriate analyses
of the waste that document that its characteristics affecting treatment
performance are such that it cannot be treated to specified levels.
Variances will not be granted based solely on a showing that adequate
•
BOAT treatment capacity is unavailable. The Agency will consider
granting generic petitions provided that representative data are
submitted to support a variance for each facility covered by the petition,
Such demonstrations can be made according to the provisions in Part
268.5 of RCRA for case-by-case extensions of the effective date of
promulgated standards.
38
-------
Petitioners should submit at least one copy to:
The Administrator
U.S. Environmental Protection Agency
401 M Street. S.W.
Washington, OC 20460
An additional copy marked "Treatability Variance" should be submitted
to:
Chief, Waste Treatment Branch
Office of Solid Waste (WH-565)
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, OC 20460
Petitions containing confidential information should be sent with
only the Inner envelope marked "Treatability Variance" and "Confidential
Business Information" and with the contents marked in accordance with the
requirements of 40 tF-R 2 (41 FR 36902, September 1, 1976, amended by 43
FR 4000).
The petition should contain the following information:
1. The petitioner's name and address.
2. A statement of the petitioner's interest in the proposed action.
3. The name, address, and EPA identification number of the facility
generating the waste, and the name and telephone number of the plant
contact.
4. The process(es) and feed materials generating the waste and an
assessment of whether such process(es) or feed materials may produce
a waste that is not covered by the demonstration.
5. A description of the waste sufficient for comparison with the waste
considered by the Agency in developing BOAT, and an estimate of the
average and maximum monthly and annual quantities of waste covered
by tne demonstration. (NOTE: the petitioner should consult the
appropriate BOAT background document to determine the
characteristics of the wastes considered in developing treatment
standards.)
39
-------
6. If the waste has been treated, a description of the system used for
treating the waste, Including the process design and operating
conditions. The petition should Include the reasons why the
treatment standards are not achievable and/or why the petitioner
believes that the standards are based on Inappropriate technology
for treating the waste. (NOTE: the petitioner should refer to the
appropriate BOAT background document as guidance for determining the
design and operating parameters that the Agency used in developing
treatment standards.
7. A description of the alternative treatment systems examined by the
petitioner (If any); a description of the treatment system deemed
appropriate by the petitioner for the waste 1n question; and, as
appropriate, the concentrations in the treatment residual or extract
of the treatment residual (i.e., using the TCLP, where appropriate,
for stabilized metals) that can be achieved by applying such
treatment to the waste.
8. A description of those parameters affecting treatment selection and
waste characteristics that affect performance, including results of
all analyses. (See Section 3 for a discussion of waste
characteristics affecting performance that the Agency has identified
for the technology representing BOAT.)
9. The dates of the sampling and testing.
10. A description of the methodologies and equipment used to obta'n
representative samples.
11. A description of the sample handling and preparation techniques,
including techniques used for extraction, containerization, and
preservation of the samples.
12. A description of the analytical procedures used, including QA/QC
methods.
After receiving a petition for a variance, the Administrator may
request additional information or waste samples that may be required to
evaluate and process the petition. In addition, all petitioners must
certify that the information provided to the Agency is accurate under 40
CFR 263.4 ;b).
40
-------
In determining whether a variance will be granted, the Agency will
first look at the design and operation of the treatment system being
used. If 1t determines that the technology and operation are consistent
with BOAT, it will evaluate the waste to determine if the waste matrix
and/or its physical parameters are such that the BOAT treatment standards
reflect treatment of this waste. Essentially, this latter analysis will
concern the parameters affecting treatment selection and waste
characteristics affecting treatment performance.
In cases where BOAT 1s based on more than one technology, the
petitioner will need to demonstrate that the treatment standard cannot be
met using any of the technologies, or that none of the technologies is
appropriate for treatment of the waste. After the Agency has made a
determination on the petition, the Agency's findings will be published in
the Federal Register, followed by a 30-day period for public comment.
After review ot the public comments, EPA will publish its final
determination in the Federal Register as an amendment to the treatment
standards in 40 CFR 268, Subpart D.
41
-------
5. P AND U WASTES
P and U category hazardous wastes Include discarded commercial
chemical products, off-specification species, container residues, and
spills residues associated with any of these; each is listed in terms of
a single chemical constituent. EPA defines P wastes as "acute
hazardous" wastes and U wastes as "toxic" wastes.
P and U wastes pose special problems for the development of treatment
standards for four reasons:
1. Although P and U wastes are listed as single chemical constituents,
their composition may vary substantially, either because the waste
may be an off-specification product or because it may become mixed
with other substances—hazardous and nonhazardous--during a spill.
2. There are 230 P and U wastes In all. It is therefore rare to find
acceptable treatment data on specific listed categories.
3. EPA does not have analytical methods that are approved by the Office
of Solid Waste for many P and U wastes.
4. Some P and U constitutents may either react or degrade in the
presence of water or other leaching solutions, preventing the direct
measurement of the P or U constituent in treatment residues. EPA
therefore considers the possibility of using reaction or degradation
products to develop treatment performance standards for these wastes,
but in some instances appropriate indicator compounds may not exist.
EPA's general approach to developing standards for P and U wastes is
to assign them to treatability groups and to transfer aata anO standards
from similar wastes. Distinctions between "acute hazardous" (P) wastes
and "toxic" (U) wastes generally have no bearing on their treatability.
co they do not contribute to the development of the treatability groups
identified in this section.
42
-------
As for other wastes, treatment for P and U wastes may be stated in
terms of numerical performance standards, but, to respond to the problems
listed above, the Agency may sometimes specify method:, of treatment (such
as incineration) rather than levels of treatment performance. EPA also
believes that recycling may be feasible for some of the P and U wastes:
some off-specification products might be required to undergo further
onsite processing; others might require treatment prior to recycling.
EPA's strategy is to define P and U treatability groups based on
similarities in elemental composition (e.g., carbon, halogens, and
metals) and the presence of key functional groups (e.g., phenolics,
esters, and amines) within the structure of the individual P or U
constituent. The Agency also considers physical and chemical factors
that are known to affect the selection of treatment alternatives and to
affect the performance of treatment—examples include volatility and
solubi1ity.
43
-------
APPENDIX A
STATISTICAL METHODS
A-l. Accuracy Correction of Sampled Data
To calculate treatment standards, it is first necessary to adjust
laboratory results for accuracy, based on the laboratory test's "recovery
*
value" for each constituent it analyzes. The recovery value measures
the amount of constituent recovered after "spiking"--the addition to the
waste sample of a known amount of constituent. The recovery value is
equal to the amount of constituent recovered after spiking, minus the
initial concentration in the sample, divided by the amount recovered.
Once the recovery value is determined, the following procedures are
used to select the appropriate percent recovery value to adjust the
analytical data:
1. If duplicate spike recovery values are available for the constituent
of interest, the data are adjusted by the lowest available percent
recovery value — the value that will yield the most conservative
estimate of treatment achieved. (If a spike recovery value of less
than 20 percent is reported for a specific constituent, however, the
data cannot be used to set a national treatment standard and are
discarded.)
2. If data are not available for a specific constituent, but are
available for an isomer, then the spike recovery data are transferred
from the isomer and the data are adjusted using the percent recovery
selected, according the procedure described in (1) above.
It may also be necessary to estimate recovery yalues in order to
perform the ANOVA test discussed in Section 3.2 to determine whi
demonstrated technologies are "best."
-------
3. If data are not available for a specific constituent, but are
available for a similar class of constituents, then spike recovery
values for this class of constituents are transferred. All spike
recovery values greater than or equal to 20 percent for a spiked
sample are averaged, and the constituent concentration is adjusted by
the average recovery value. If spiked recovery data are available
for more than one sample, the average is calculated for each sample
and the data are adjusted by the lowest average value.
4. [f spike recovery data are not available for the waste matrix, then
spike recovery values are transferred from a waste that the Agency
believes is a similar matrix. For instance, if the data are for an
ash resulting from incineration, then data from other incinerator
ashes could be used. This is not an exact analysis, but it is
considered the best practical approach. In assessing the recovery
data to be transferred, the procedures outline in (1), (2), and (3)
above are followed.
The analytical procedures employed to generate the data used to
ca'cv'late each treatment standard for tested wastes are provided in
Appendix B cf the background document prepared for that waste. This
appendix will also document any alternatives or equivalent procedures
and/or equipment allowed by EPA's SU-846, Third Edition (November 1986).
NOTE: The Agency will use the methods and procedures presented in
Appendix B of each background document to enforce the treatment
standards. Facilities should therefore use these procedures in assessing
the performance of their treatment systems.
A.2 F Value Determination for ANQVA Test
As noted earlier in Section 1.2. EPA is using the statistical method
known as analysis of variance (ANOVA) to determine the level of
performance that represents "best" treatment where more than one
technology is demonstrated. This method provides a measure of the
differences between data sets.
45
-------
If the Agency found that the levels of performance for one or more
technologies are not statistically different (i.e., the data sets are
homogeneous), EPA would average the long-term performance values achieved
by each technology and then multiply this value by the largest
variability factor associated with any of the acceptable technologies.
If EPA found that one technology performs significantly better (i.e., the
data sets are not homogeneous), the "best" technology would be the
technology that achieves the best level of performance, i.e., the
technology with the lowest mean value.
To determine whether any or all of the treatment performance data
sets are homogeneous using the analysis of variance method, it is
necessary to compare a calculated "F value" to what is known as a
"critical value." (See Table A-l.) These critical values are available
in most statistics texts (see, for example, Statistical Concepts and
Methods by Bhattacharyya and Johnson, 1977, John Wiley Publications.
New York).
Where the F value is less than the critical value, all treatment data
sets are homogeneous. If the F value exceeds the critical value, it is
necessary to perform a "pair wise F" test to determine whether any of the
sets are homogeneous. The "pair wise F" test must be done for all of the
various combinations of data sets using the same method and equation as
trie general F test.
-------
CRITICAL VALUES
[ R»P
Reproduced trom
•v«ll«M« Copy.
nt
n>
»Jth PIllCINmi VAIUIS FOt
THI F DISTH1IUT10N
• dicr««s of frtNdom for numtrxtor
* dtyr««s of Irwdom for dtnominator
• JI)
N-k
1^1 1
1 1C1.4
: is.31
; 10.13
4 7.71
S 4.81
C 3J»
7 349
8 341
!» U2
10 446
1 4.84
2 4.73
3 4.07
4 4.80
3 4.34
4 i 4.;9
- 4.;3
I 4.41
9 448
t) 443
22 UO
1 441
21 443
21 440
30 4.17
40 4.08
30 4.03
40 OO
70 198
10 19«
:oo 194
'.JO Ul
2:0 ', :.i9
.47 :<
LSI
1.70
1.44
L73
1.22
L03 '
140 '
184
130
140
t *<* <
^i*»»
124
US
113
'_3?
1.34
*_30
.94
.31
.14
.12
.71
.74
.44
.40
44
>43
41
.48 '
.44
.*2 :
.31 '.
•*•» •
00 •
s:.a 23,.;
>.«9 IS. JO
U3i i.:;
s.4c i.i:
1.40 «.:<
:.7i ;.r
143 u:
131 u:
:.78 i::
139 13,
» • I * . ^
to « J •. \*
* ^€ * *•
•K^J •^^W
128 241
113 u;
« . i« * »«•
* ^» •) 41
•. • ' ^ * .
»:: '.ji
.)i :.r.
.34 -..u
.90 14;
.14 ..'1
-JO 1.73
.'1 '..»
.72 1.13
.59 l.C
49 141
42 1. «4
.a 149
.43 1.23
.42 142
.23 143
.24 '. 4T
.;; '.^.j
43 :.::
44 ;.;o
able
95th Percentile Values 'or :r,e F Di
-------
The F value is caio-ated as follows:
(i) All data are natural logtransformed.
(ii) The sum of the data points for each data set is computed (T )
(iii) The statist';.?1 parameter known as the sum of the squares
between data sets (SSB) is computed:
]2
J_
J
SSB
f
k
r
V7
i
1
!
k ]<
J. 'J
N
where:
k • number of treatment technologies
n, - number of data points for technology i
N • number of data points for oil technologies
T, • sum of natural logtnnsformed data points for each technology
(iv) The sum of the squares within data sets (SSW) is computed:
SSW
••here:
' .J
T
2 i
<, , • The natural logtransformed observations (j) for treatment
technology (i).
(/) The degrees of freedom corresponding to SSB and SSW are
calculated. For SSB. '.he oegree of freedom i •; gi.en by «. • '. . For- i>S
tne degree of freedom is given by N-*.
;/i) Using the above parameter. 'he F value is calcvlatej a:
• T I '• ow s:
-------
where:
MSB • SSB/(k-l) and
MSW - SSW/(N-k).
A computational table summarizing the above parameters is shown below.
Computational Table for the F Value
Source
Between
Within
Degrees of
freedom
k-1
N-k
Sum of
squares
SSB
SSW
Mean square
MSB • SSB/k-1
MSW - SSW/N-k
F
MS3/MSW
Below are three examples of the ANOVA calculation. The first two
represent treatment by different technologies that achieve statistically
similar treatment; the last example represents a case where one
technology achieves significantly better treatment than the other
technology.
-------
1790Q
1
N*thyl«n> Chlorid*
m »triooinq Biolocicil
effluent li>(«fflu«nt) [ln(«fflittnt)]2 Influent Effluent
[ln(*ffliMnt)]
1SSO.OO
1290.00
1 MO 00
5100.00
KM. 00
4800.00
1760.00
2400.00
4800.00
12100.00
10.00
10.00
10.00
12.00
10.00
10.00
10.00
10.00
10.00
10.00
2.30
2.30
2 30
2.48
2.30
2.30
2 30
2.30
2.30
2 30
S.29
S.29
5.29
6. IS
S.29
S.29
S.29
S.29
S.29
S.?9
1960.00
2S88.00
1817 00
1640.00
3907 . 00
10.00 2.30
10.00 2.30
10.00 2 30
26.00 3 26
10.00 2.30
S.29
S.29
S 29
10.63
S 29
Su»:
23.18
S3 76
12 48
31 .79
S ue.
10 10
10
3669
10 ?
2.32
2378
13.2
2 49
>t«nd«rd Deviation:
J32U.6/ &3
.06
923.0-i
/.IS
43
V*r ubi I ity f jctor
1
2 48
MtOVA CalcuUt icni:
2
SS8
t r
H'-B • '.SS/lk 1 )
' i2
-------
l/90g
1 (Continued)
NSB/MSU
k
n •
of tro«ta*flt technologist
at d*L« point* for tecrwjlogj i
of natural lagtraraforad data ixiinti for til technologic*
T * tim of loglra/uforned data points for etch technology
i
1 • the nal logtrwtsforad abeenMit lant (j) for tre«UMnt technology (\)
n - 10, n - S. N • IS. k • t. I • ?3 18. T • l?.4«. I • 15. M, I • 1770 21
r . i3;.3i r - iss.zs
SSB
537 31 155
ft
MMeH^M^HW •^^•^H
10 5
• 0.10
SSU - (53 76 • 31 79)
WS8 • 0.10/1 • 0.10
NSW > 0.77/13 • 0.06
r . HI • ..67
0.06
537 31 155.25
»
••^i^^^i^^a^ ^^^H^
10 5
0.77
MUVA T«ble
Source
0*q
of
<>$
NS
v«lu«
Within(W) 13
0 10
0.77
0.10
0.06
1.67
The critical value of the F teit at th» 0.05 significance" level is 4 67 Since
the F value 11 lets than the critical value, the Mans are not significantly
different (i.e.. they are hceogeneous).
Hote All calculations »cr» rounded to l«o dec i«a I places Result! *ay differ
depending upon the nu*b«r of dec ioa I places used in eacn >tep of the calculationt
-------
17**
Tr ich lortMthy )«••
ill
Influent
(•0/1)
16U.OO
SJOO.OO
SOOO.OO
1720.00
ISM 00
10300.00
210.00
1800.00
204.00
160.00
M ItrlOOifta
Effluent
(«/»
10.00
10.00
10.00
10.00
10.00
10.00
10 00
27 50
as. oo
10.00
B lp loo teal treatntnt
ln(«f fluent ) (!n(«f fluent)] Influent
2.30
2.30
2 30
2.30
2.30
2.30
2.30
3.3P
4.44
2.30
S.29
S 29
S 29
S.N
S.29
S.29
S.29
10.89
19. 7i
S.29
wn
200.00
224.00
134.00
ISO. 00
444.00
1U.OO
182.00
Cfflucnt In(cfflutnt) [ln|«ff lu«nt )]
(^/»
10.00
10.00
10.00
10.00
16. 25
10.00
10.00
2.30
2.30
2.30
2 30
2.79
2.30
2 JO
S.29
S.29
S.29
5 29
7.78
S.29
S 29
Sdople Silt:
10 10
2160
19.
Standard Deviation:
3209.6 23 7
V«ridbi I ity >4Clor
3.70
28.14
10
2.61
71
?2 92
220
120. 5
10.89
J.38
I.S3
11. S9
2.37
19
39 S2
AMOVA GJ Iculat i
SS8
S5W • I S T.'
"T58 • SS8/(k-l)
-------
I790g
Cxacvle 2 (Continued)
F • NS8/MSU
•her*:
k « mjebar of tre*tennt technologies
n • nuetoer of data points for technology i
i
» • mater of d*ta point* for all technologic*
I • ua of natural loglran»fone«d data points for Men technology
i
I • the natural logtrcnafonHd observation* (j) for treatment technology (i)
|J
N • 10, II • /. N - I?, k • 2. t • 26.14, T • 16.S9, I • 42.73. T • I82S 8S. I - 683.10.
T' - 275.23
. fW3 30 T75 23 } 1825.85
SS8 • • - • 0.25
10 7 17
, 683 30 275 ?J
SSW • (77 92 » 19,42) - | _ » _ | • 4.79
10
XSB • 0.2V I ' 0.25
NSW • « n/ \ 5 • 0 32
- .0.78
0 32
ANOVA Table
Degree* of
Source freedo* SS NS T value
Vtthtn(V)
1
15
0.25
4 79
0 ?S
0 32
0 78
I he crilmai it lue of the F lest *l the 0.05 jignificjnce level it 4 V4 Since
the f value n leu than the critical value, the ne*n* are not t iqn if icant ly
different (> t . they are nonoqeneoui)
tote All calculation*
-------
Kw«l«d
iludm follcmad by carbon tdtorotion
8 IQ logical tr««tm»nt
.jiMucnt Effluent ln(«f fluent) [\n(»t f lutnt)]2 Influent
(•4/1) («4/U (»fl/U
7200.00 80.00 438 19.18 9206.00
6500.00 70.00 4.25 18.06 16846.00
6075.00 35.00 3.56 12.67 49775.00
3040.00 10.00 2.30 S.29 14731.00
3159.00
6756.00
3040.00
Sun:
14.49 55.20
S«4>U Size:
444 - 7
Mean:
S703 49 3.62 • 14759
Effluent
Uq/D
1083.00
709. SO
460.00
142.00
603 00
153.00
17.00
.
7
452 5
ln(«f fluent)
6.99
6 56
6.13
4 96
6.40
5.03
2.83
38.90
7
5 56
ln((«ff lucnt)]2
48.88
43.03
37 58
24.60
40 96
25 30
8.01
228.34
-
Suncurd 0«i 141 ion:
1835.4 32 24
jri*biIity Factor:
95
16311.86
/.OO
379 04
lb./9
1 .42
AHOVA CikuUt ions:
SSB
k
l-l
MSB - SS8/(k-l)
MSW
1 n,
t r. «', ,
i-i j-i ''J
F
-------
1790g
Example 3 (Continued)
•tare.
* * nuvter of treatment technologies
n • nueter of data points for technology i
N - ntav&ar of data points for all technologies
T • SUM of natural log transformed data points for each technology
i
X • the natural loqtraraforad observations (j) for treatment technology (i)
|J
N . 4. N • 7. N . 11. k - 2. T - 14.49. T - 38.90. T • S3.39. T*« 2850.49. T* - 209.96
T? - 1S13.2\
t-yna aft 1^1171 \ 7A^n AQ
• 9.S2
SSW . (SS.20 * 228.341 - . ' .14.88
USB • 9.S2/1 • 9 S2
NSU • 14 ua/9 • 1.6b
F - 9 M/l.65 • S.77
ANOVA Table
Ocgrvet of
Source freeotoe
&etM«n(B) 1
Withm(U) 9
SS MS F value
9 W 9 S3 S.77
1489 1.6S
The critical value of the F test at the 0 OS nqnif tcance level n S 12 Since
trie f value is larger than the critical value, the meant are nqnif \cant ly
different (in., they are heleroqeneou*). Activated J lutlqe followed by carbon
adsorption u 'best" in this e»anp le because the mean of the long-ten* performance
value. ie. the effluent concentration, n lover
Hole » i I CJ leu Lit ions «er« roonrted to t«o decimal places Deiults nay differ rlependinq
upon the .iutt>er of Jec i«a I place* uied in each step of the calculations
-------
A.3. Variabil ity Factor
VF • Mean
where:
VF • estimate of the daily maximum variability factor determined
from a sample population of daily data.
Cgg • Estimate of performance values for which 99 percent of the
daily observations will be below. Cgg is calculated using
the following equation: Cgg • Exp(y + 2.33 Sy) where y and
Sy are the mean and standard deviation, respectively, of the
logtransformed data.
Mean - Average of the individual performance values.
EPA is establishing this figure as an instantaneous maximum because
the Agency believes that on a day-to-day basis the waste should meet the
applicable treatment standards. In addition, establishing this
requirement makes it easier to check compliance on a single day. The
99th percentile is appropriate because it accounts for almost all process
v a r i a b i 1 11 y .
In several cases, aJJ. the results from analysis of the residuals from
BOAT treatment are found at concentrations less than the detection
limit. In such cases, all the actual concentration values are considered
to be unknown and hence, cannot be used to estimate the variability
factor of the analytical results. The following ii a description of
EPA's approach for calculating the variability factor for cases in wmcn
all concentrations below the detection limit.
It nas been postulated that a lognormal oiitribut ion adequately
describes the variation among concentrations. Agency data snows ttiar. ;ne
56
-------
treatment residual concentrations »re often distributed approximately
lognormally. Therefore, the logncrmal mode! has been used routinely in
EPA's development of numarous r»(juUt.ior.s in the Effluent Guidelines
Program and is being used in tht- BOAT program. The variability factor
(VF) was defined as the ratio of the 99th percentile (C ) of the
lognormal distribution to its arithmetic mean (Mean).
VF • C99 (!)
Mean
The relationship between the parameters of the lognormal distribution
and the parameters of the normal distribution created by taking the
natural logarithms of the lognormally distributed concentrations can be
found in most mathematical statistics texts (see, for example,
Distribution in Statistics -Volume 1 by Johnson and Kotz, 1970). The mean
of the lognormal distribution can be expressed in terms of the
mean (M) and standard deviation (j) of the normal distribution as
fol1ows:
C9g - Exp U * 2.33U) (2)
Mean • Exp (^ 1- O.S.i ) (3)
By substituting (2) and (3) in (I) the variability factor can then be
expressed in terms of o as follows:
VF - Exp (2.33 o • 0.5,,2) (4)
For residuals with concentrations that are not all below the
detection limit, the 99th percentile and the mean can be estimated from
tne actual analytical data and. accordingly, the variability factor1 (VF)
57
-------
can be estimated using equation (1). For residuals with concentrations
that are below the detection limit, the above equations can be used in
conjunction with the following assumptions to develop a variability
factor.
• Assumption 1: The actual concentrations follow a lognormal
distribution. The upper limit (UL) is equal to the detection
limit. The lower limit (LL) is assumed to be equal to one-tenth
of the detection limit. This assumption is based on the fact that
data from wel1-designed and wel1-operated treatment systems
generally fall within one order of magnitude.
• Assumption 2: The natural logarithms of the concentrations have
a normal distribution with an upper limit equal to In (UL) and a
lower limit equal to In (LL).
• Assumption 3: The standard deviation (j) of the normal
distribution is approximated by:
a - [(In (UL) - In (LL)] / [(2)(2.33)] - [ln(UL/LL)] / 4.66 (5)
(Note that when LL • (0.1)(UL) as in Assumption 1, then j •
(InlO) / 4.66 • 0.494.)
Substitution of the a value from equation (5) into equation (4)
yields the variability factor, VF.
VF . 2.8
58
------- |