JRfc United States
Environmental Protection
L*l »m Agency
Toxic Weighting Factors Methodology
U.S. Environmental Protection Agency
Office of Water
Engineering and Analysis Division
1200 Pennsylvania Avenue, NW
Washington, D.C. 20460
March 2012
EPA - 820-R-12-005
-------
CONTENTS
Page
1. Introduction 1-1
1.1 Introduction References 1-3
2. Clarification of Key TWF Concepts and Terms 2-1
3. The TWF Equation 3-1
3.1 The TWF Equation References 3-3
4. The Aquatic Life Component of the TWF Equation 4-1
4.1 The Aquatic Life Component of the TWF Equation References 4-2
5. The Human Health Component of the TWF Equation 5-1
5.1 Equation for Calculating the TWF Human Health Value for Noncarcinogens 5-1
5.2 Equation for Calculating the TWF Human Health Value for Carcinogens 5-2
5.3 Data Sources for Calculating the Human Health Component of the TWF
Equation 5-3
5.4 Calculating the TWF Human Health Value for Dioxins 5-4
5.5 Variable Factors Used in the Human Health Component of the TWF Equation 5-6
5.5.1 Relative Source Contribution Factors 5-6
5.5.2 Bioaccumulation and Bioconcentration Factors 5-6
5.6 The Human Health Component of the TWF Equation References 5-7
6. EPA' s Use of TWFs to Support Effluent Guidelines Rulemakings and Program
Planning 6-1
6.1 Use of TWFs in Cost-Effectiveness Analysis 6-1
6.2 Use of TWFs in Effluent Guidelines Program Planning 6-3
6.3 EPA's Use of TWFs to Support Effluent Guidelines Rulemakings and Program
Planning References 6-6
7. Limitations and Appropriate Uses of Toxic Weighting Factors 7-1
7.1 Limitations of Source Data Used to Derive TWFs 7-1
7.2 Limitations of Discharge Data to Which TWFs Are Applied 7-1
7.3 Appropriate Uses of TWFs 7-3
7.4 Limitations and Appropriate Uses of Toxic Weighting Factors References 7-4
i
-------
LIST OF TABLES
Page
Table 3-1. TWF for Copper 3-2
Table 5-1. Toxic Equivalency Factors (TEF), Bioaccumulation Factors (BAF), and TWF Values
for Dioxin Compounds 5-5
Table 6-1. Examples of Effluent Guidelines Cost-Effectiveness Analysis Ranked by Dollars Per
TWPE Removed 6-2
Table 6-2. TWPE Ranking of Industrial Categories Based on 2008 DMR Data 6-4
Table 6-3. TWPE Ranking of Industrial Categories Based on 2008 TRI Data 6-5
li
-------
LIST OF FIGURES
Page
Figure 3-1. TWF Equation 3-1
Figure 5-1. Equation for Calculating the TWF Human Health Value for Noncarcinogens 5-1
Figure 5-2. Equation for Calculating the Human Health Component of TWFs for Carcinogens 5-2
Figure 5-3. Hierarchy of Data Sources for Calculating the Human Health Component of TWFs 5-3
in
-------
1. Introduction
The purpose of this report is to explain the concept, calculation, use, and limitations of Toxic
Weighting Factors (TWFs). The U.S. Environmental Protection Agency's Engineering and Analysis
Division of the Office of Water (EPA) began developing TWFs in the early 1980s to compare the relative
cost effectiveness of treatment technology options for national industrial discharge regulations known as
effluent guidelines. EPA expanded its application of TWFs in 2003 when it began using them in effluent
guidelines program planning to rank industries by their total annual toxic-weighted discharges to identify
those that may warrant additional research (U.S. EPA, 2004).
EPA has calculated TWFs for 1,064 chemicals based on information about the concentrations in
water at which they become harmful to aquatic life, along with information about the levels in fish tissue
at which they become harmful to humans. Information on harmful effects is used to weight the toxicity of
chemicals relative to copper which is a common toxic pollutant in industrial waste streams. By assigning
TWFs to toxic chemicals, EPA can add together varying amounts of different chemicals, each with
different toxicities, to calculate the total toxic-weighted pounds in a waste stream, referred to as Toxic
Weighted Pound Equivalents (TWPE). For example, nickel, which is less toxic than copper, has a TWF of
0.11; whereas acrylonitrile, which is more toxic than copper, has a TWF of 2.3. A facility that annually
discharges 50 pounds of nickel and 10 pounds of acrylonitrile (50 x 0.11 + 10x2.3 = 28.5 TWPE) has a
waste stream that is almost twice as toxic as a facility that discharges 100 pounds of nickel and 2 pounds
of acrylonitrile (100 x 0.11 + 2 x 2.3 = 15.6 TWPE). TWPE-based estimates of the toxicity of waste
streams are solely used by EPA to better understand how treatment technologies, individual facility
discharges, and industry discharges compare to one another, not to assess impacts to aquatic life or human
health.
The equations for calculating TWFs are contained in a set of Excel Worksheets known as the
TWF Database. The TWF Database uses aquatic life values that have already been calculated for various
purposes such as for use in water quality criteria. Human health values are derived in the TWF Database
using information from disparate sources. EPA combines aquatic life and human health effects
information whenever possible when calculating TWFs to help compensate for the incompleteness of data
sources and to simplify TWF calculation and waste stream evaluation. The majority of TWFs, 58 percent
(616), are derived solely from aquatic life values. Five percent of TWFs (54) are derived solely from
human health values. Thirty-seven percent of TWFs (394) are derived from both aquatic life and human
health effects information. If EPA were to separate the aquatic life and human health values during TWF
calculation, there would be 1,010 chemicals with aquatic life-based TWFs (616 + 394 = 1,010) and 448
chemicals with human health-based TWFs (54 + 394 = 448). The parsing of industry rankings by separate
aquatic life and human health categories would defeat the purpose of TWFs, which is to help simplify the
M
-------
interpretation of disparate discharge data. Moreover, there is no clear metric for decision makers to use in
choosing between the results of separate aquatic life and human health-based industry rankings.
EPA recognizes that the approach of combining aquatic life and human health values is not
suitable for rigorous applications, such as analyses performed for risk assessments or water quality-based
permitting. EPA acknowledges that the TWF approach of combining aquatic life and human health values
mixes different types of effects data, amplifies uncertainties in the underlying data, and obscures the
relative contribution of aquatic life and human health values in the calculation of TWFs. EPA also
recognizes that the toxic effects of many chemicals are frequently different in humans than they are in
aquatic life. Nevertheless, EPA considers the combination of aquatic life and human health effects
information in the calculation of TWFs to be a reasonable approach for integrating disparate toxic effects
information from different sources for the purpose of informing regulatory decisions about the relative
cost effectiveness of treatment technologies, and for indentifying industries that may warrant additional
analysis.
It should be noted that although the TWF Database contains TWFs for 1,064 chemicals, only a
small subset of chemicals with TWFs are typically assessed by the effluent guidelines program. The
number of chemicals with TWFs regulated by effluent guidelines for any particular industrial category,
and thus assessed in a cost effectiveness analysis, typically range anywhere from one to 20. Effluent
guidelines program planning assessment only examines the approximately 200 chemicals with TWFs that
are annually reported by facilities to be discharged to surface water nationwide. EPA data show that the
total amount and toxic-weighted amount of chemicals discharged to surface water is declining over time
as industries adopt increasingly effective treatment technologies.
This report incorporates comments from an independent peer review panel EPA convened to
examine the TWF methodology (ERG, 2006). The majority of the panel's suggestions concerned
improving the clarity of the explanation of how TWFs are developed. It was also apparent from the
panel's comments that the TWF Methodology did not adequately discuss the limitations of TWFs and
their specific uses by EPA. EPA prepared a detailed response report to address the peer review panel's
comments (U.S. EPA, 2011). Moreover, EPA has endeavored to respond to the panel's suggestions in this
report by more clearly explaining how TWFs are derived from toxic effects data and more precisely
describing EPA's specific uses of TWFs in effluent guidelines regulatory evaluation and planning. EPA
has also revised this report to more thoroughly describe the limitations of the data used to calculate
TWFs, and the limitations of EPA's facility discharge data to which TWFs are applied. EPA will continue
to include a section on TWFs in the Technical Support Document accompanying effluent guidelines
program plans summarizing any revisions to existing TWFs or development of new TWFs.
1-2
-------
EPA has continually developed and refined TWFs during the past three decades as new chemical
toxicity data became available. However, given the current extensive set of chemicals for which TWFs
have been calculated, EPA does not anticipate developing new TWFs in the foreseeable future.
Nevertheless, EPA may decide to revise certain TWFs as new chemical toxicity data become available.
EPA may also decide to revise this report as needed to further clarify the TWF methodology in response
to public comments. EPA will use the peer reviewers' comments to inform any future refinement of
TWFs or revisions to the TWF methodology.
The body of this report is organized into six chapters. The first four chapters discuss TWF
terminology and the equations and data used to calculate TWFs. The chapter explaining the aquatic life
component of TWFs is much shorter than the chapter explaining the human health component. This is
because the TWF Database uses aquatic life values that have already been calculated by other sources,
whereas human health values are calculated in the TWF Database and thus require more detailed
explanation. The last two chapters summarize EPA's use of TWFs and explain their limitations.
1.1 Introduction References
1. ERG. 2006. Eastern Research Group, Inc. Toxic Weighting Factor Development in Support of
CWA 304(m) Planning Process. (June). EPA-HQ-OW-2004-0032-1634.
2. U.S. EPA. 2004. Technical Support Document for the 2004 Effluent Guidelines Program Plan.
EPA 821R-04-014. Washington, DC. (August). EPA-HQ-OW-2003-0074-1346 through 1352.
3. U.S. EPA. 2011. U.S. Environmental Protection Agency. Response to Peer Review Comments on
EPA Report: Toxic Weighting Factor Development in Support of CWA 304(m) Planning Process.
Washington, DC. EPA-HQ-OW-EPA-HQ-OW-2010-0824, DCN 07500.
1-3
-------
2. Clarification of Key TWF Concepts and Terms
The effluent guidelines program only uses TWFs as an indicator of the relative potential of
chemicals to cause harm prior to discharge, and never to assess harmful effects of chemicals in the
environment. This is a critical point to understand about TWFs and, consequently it is emphasized
throughout this report. The toxic effects of chemicals discharged to aquatic environments are highly
variable and site specific. Adequate assessment of the toxic effects of chemicals in aquatic environments
requires site specific information beyond the scope of TWFs.
Toxicity refers to the degree to which a substance can harm an organism. A central concept of
toxicology is that toxic effects are variable depending on the chemical structure of a substance, the dose
an organism receives, and an organism's susceptibility to increasing doses. All chemicals can cause harm
at a certain level. A chemical is considered to be toxic if a small amount of it can cause harm and
relatively non-toxic if only large amounts of it are harmful.
The toxic properties of a waste stream begin to change as soon as it is discharged to surface
water. The potential toxic effects of chemicals vary depending on their concentration in a waste stream,
their dilution once they are discharged, their inherent potential for persistence or degradation in the
environment, and the physical and chemical characteristics of the waterbody to which they are
discharged, such as temperature and pH. After mixing with surface water, toxic chemicals may degrade
through photochemical and microbial processes, or compartmentalize through particulate adsorption,
sediment deposition, or volatilization. Biological alteration in the aquatic environment may decrease or
increase the toxicity of a chemical, such as when certain anaerobic bacteria metabolize inorganic mercury
into methyl mercury, making it biologically available.
Aquatic life effects from toxic chemicals are generally considered to be either acute or chronic.
Acute effects are well defined and occur relatively quickly, such as death within a couple of days.
Chronic effects are more subtle, developing and persisting for long periods over the course of an
organism's life time. Chronic effects are typically measured in terms of physiological decline, such as
reduced growth and reproduction, but they may be also measured or in terms of the lethality of a
chemical.
Human health effects from toxic chemicals are divided into two categories: non-threshold effects,
such as carcinogenicity, and threshold effects, such as acute or chronic toxicity. EPA assumes that there is
no safe level of exposure to carcinogens, and thus there is no effects threshold. The risk of developing
cancer from a certain dose of a chemical for a given duration of exposure is therefore not an effect
threshold, but rather represents an indicator of potential level of harm. Noncarcinogens are generally
assumed to have an identifiable exposure threshold, both for individuals and populations, below which
2-1
-------
effects are not observable. For example, EPA water quality criteria for noncarcinogens represent a type of
toxic effects threshold because they define maximum ambient water chemical concentrations that can
exist before harmful effects begin to occur.
The human health component of TWFs is derived almost entirely from information concerning
fish consumption, which is the most significant pathway of exposure to water borne toxic chemicals.
Certain highly toxic chemicals, such as mercury, may become concentrated at harmful levels in the types
of fish commonly eaten by humans through bioaccumulation, bioconcentration, and biomagnification.
Bioaccumulation refers to the buildup of toxic chemicals in organic tissues from all combined routes of
exposure, including food consumption, water ingestion, and contact with water and sediments.
Bioaccumulation occurs when an organism absorbs a toxic chemical faster than it metabolizes or excretes
it. The longer a toxic chemical persists in organic tissue, the greater the potential for adverse effects, even
if environmental levels of the chemical are not very high. Bioconcentration is a related but more specific
term, referring to uptake and accumulation of a toxic chemical from contact with water alone.
Bioaccumulation typically accounts for significantly more chemical build up in tissues than
bioconcentration. Biomagnification occurs when chemicals become increasingly concentrated up the
levels of a food chain through predation. Chemical bioavailability, biota fat content, and rate of transfer
among food chain levels can affect biomagnifications potential. Chemicals such as dioxins are
particularly prone to biomagnification because they are highly persistent in the environment and
accumulate in fatty tissues.
2-2
-------
3. The TWF Equation
EPA convened a work group in 1981 to develop TWFs for use in evaluating the cost-
effectiveness of treatment technology options for effluent guidelines rulemakings (U.S. EPA, 1988). The
work group decided to simplify TWF calculation by combining information about toxic effects to aquatic
life and human health. The equation is illustrated in Figure 3-1. The 1981 work group also decided to
normalize the aquatic life and human health values to a standard benchmark so that different amounts of
different chemicals could be added together to estimate the total toxicity of a waste stream. The work
group chose the freshwater chronic aquatic life criterion for copper as the benchmark scaling factor
because it is a common and well studied toxic chemical in industrial waste streams.
Where:
5.6
scaling factor (original copper freshwater chronic aquatic-life criterion of 5.6 (ig/L);
AQ =
Aquatic Life Value (|ig/L): and
HH
Human Health Value (|ig/L).
Figure 3-1. TWF Equation
The aquatic life value is taken directly from one of several toxicity databases, depending on the
best source of information available for each chemical. The human health value is calculated for each
chemical using data from multiple sources of best available information depending on whether the
chemical is a carcinogen or noncarcinogen. Data sources for TWFs are discussed in Sections 4 and 5.
The chronic freshwater aquatic life criterion for copper, however, has been revised three times
since it was first published in 1980 due to advances in the scientific understanding of its toxic effects. The
original copper freshwater chronic aquatic life criterion of 5.6 micrograms per liter (|ig/L) was revised to
12.0 |ig/L in 1984, then changed to 9.0 |ig/L in 1998, and revised again in 2007 to a criterion based on
site-specific water quality factors (U.S. EPA, 2007a). As shown in Figure 3-1, EPA has continued to use
the original 5.6 |ig/L value for the TWF benchmark in order to provide consistency in interpretation of
TWF-based analyses from year to year.
The nature of the copper criterion helps to illustrate the complexity of determining the potential
toxic effects of a chemical. Copper frequently occurs naturally in surface water because it is an abundant
trace element in the earth's crust. At low concentrations it is an essential micronutrient for virtually all
plants and animals, but at higher concentrations copper is toxic to aquatic life. To better account for the
waterbody-specific nature of copper toxicity, EPA published revised water quality criteria for copper in
2007 (U.S. EPA, 2007a) based on a chemical speciation model, called the Biotic Ligand Model (BLM)
3-1
-------
(U.S. EPA, 2007b). The BLM calculates the toxicity of copper to aquatic organisms as a function of
certain chemical and physical characteristics of a receiving waterbody. It also simultaneously calculates
ionic concentrations of other chemicals that can either bind with copper and render it biologically
unavailable, or compete with copper for binding sites on aquatic organisms, such as fish gills. Although
the BLM can potentially improve the accuracy of predictions of the ecotoxicity of copper and other
metals, its use requires significant data input, including water temperature along with concentrations of
nine separate chemical constituents: pH, alkalinity, calcium, magnesium, sodium, sulfate, potassium,
chloride, and dissolved organic carbon.
The BLM-based copper criterion provides no national value that can be used in TWF calculations
because it is site specific. Therefore, when calculating the TWF for copper, EPA normalizes the 1998
copper criterion of 9.0 |ig/L to the original 1980 copper criterion of 5.6 |ig/L. The 1998 copper criterion is
used because it is the most recent copper criterion prior to the BLM-based copper criterion. The copper
human health value of 444 |ig/L. derived from the reference dose contained in the 1998 EPA Region 3
Risk-based Concentration Table and noted in Section 5, is also normalized to the original 1980 copper
criterion of 5.6 |ig/L.
Consequently, the TWF for copper is 0.635 rather than 1.0, as illustrated in Table 3-1.
Table 3-1. TWF for Copper
A(|ii;ilic l.il'c Value
(fili/l.)
Milniiiii llcallh Value
(jili/l.)
Application of TW I'
Scaling l-aclor
Toxic Weigh ling
l-.icloi-
Copper
9 (1998 Criterion)
444
5.6/9.0 + 5.6/444
0.635
Although the 1981 cost effectiveness workgroup selected copper as the benchmark, any chemical
could have been used for weighting relative chemical toxicities. Chemicals with toxicities close to the
mean toxicity of a group of chemicals, however, would serve as the most balanced benchmark for scaling
purposes. In the most recent version of the TWF Database, compiled in 2007, there are 258 chemicals
with TWFs greater than that of copper and 805 chemicals with TWFs less than the TWF for copper.
The complete list of TWFs, the values used in the aquatic life and human health components of
TWFs, the values of the constant and variable factors used to calculate TWFs, and all data sources are
contained in the TWF Database. The TWF Database is available in the public docket for the effluent
guidelines planning process and can be accessed through docket number "EPA-HQ-OW-2008-0517" at
www.regulations.gov (U.S. EPA, 2007c). The 2007 TWF Database in the docket is the most current
version as of the date of this report. This report is also available in "EPA-HQ-OW-2008-0517. "Any
3-2
-------
future revised versions of the TWF Database and reports will be posted to the effluent guidelines program
planning docket, which is updated annually and can be found by searching www.regulations.gov.
3.1 The TWF Equation References
1. U.S. EPA. 1988. U.S. Environmental Protection Agency. Office of Water. Ehrensberger, K., and
R. Rico. Cost-effectiveness Analysis for Effluent Guidelines. Washington, DC. (May).
2. U.S. EPA. 2007a. U.S. Environmental Protection Agency. Office of Water. Aquatic Life Ambient
Freshwater Quality Criteria: Copper. EPA-822-R-07-001. Washington, DC. (February). EPA-
HQ-OW-2006-0771-1214. Available online at:
http://water.epa.gOv/scitech/swguidance/standards/criteria/aqlife/pollutants/copper/upload/2009 0
4 27 criteria copper 2007 criteria-full.pdf.
3. U.S. EPA. 2007b. U.S. Environmental Protection Agency. Office of Water. Aquatic Life Ambient
Freshwater Quality Criteria - Copper, 2007 Revision. EPA-822-R-07-001. Washington, DC.
(February). Available online at:
http://water.epa.gov/scitech/swguidance/standards/criteria/aqlife/pollutants/copper/20Q7 index.cf
m.
4. U.S. EPA. 2007c. U.S. Environmental Protection Agency. Office of Water. Toxic Weighting
Factor Database. Washington, DC. (September). EPA-HQ-OW-2008-0517 DCN 04661.
3-3
-------
4. The Aquatic Life Component of the TWF Equation
EPA selects aquatic life effects data to calculate TWFs using a three tiered hierarchy in order of
decreasing data quality:
1. Freshwater chronic aquatic life criteria published by EPA as National Recommended
Water Quality Criteria (NRWQC) (U.S. EPA, 2009);
2. Effects data based on testing of aquatic organisms; and
3. Estimated toxicity effects data derived from various nationally-recognized sources.
Based on this hierarchy, aquatic life values are available for all chemicals for which TWFs have been
developed, with the notable exception of certain dioxin compounds as illustrated in Table 5-1. If aquatic
life effects values are not available for a particular dioxin compound, then the copper-normalized human
health value is used as the TWF for that chemical. Conversely, TWFs for 616 chemicals are based solely
on aquatic life values because human health values are not available.
Aquatic life effects values from freshwater environments, only, is used in TWF calculations.
Saltwater aquatic effects values are not used because they have been calculated for only a small set of
chemicals due to the complexities of determining toxicity effects in marine environments.
Chronic long-term effects values rather than acute short-term effects values are used in most
TWFs because they are more available. For a few TWFs, however, acute values are used because chronic
values are not available. Acute effects are typically expressed as toxicity endpoints. An acute toxicity
endpoint of LD50, for example, is the lethal dose of a chemical that kills 50 percent of test organisms in
specified period, usually within 96 hours or less. Chronic effects are more subtle and develop and persist
for a relatively long period of time, typically one-tenth or more of the life span of an organism. Chronic
effects are measured in terms of physiological decline, such as reduced growth and reproduction, or in
terms of the lethality of a chemical.
Freshwater chronic aquatic life criteria published as NRWQC, available for 72 chemicals, are the
highest quality source data for the aquatic life component of TWFs because they have been meticulously
developed, extensively reviewed, and constitute national guidance for protection of the vast majority of
aquatic freshwater communities in the United States. The Criterion Continuous Concentration (CCC)
values from freshwater chronic aquatic life criteria are used in the aquatic life component of the TWF
equation. The CCC is a chronic value representing the highest in-stream concentration of a toxicant to
which organisms can be exposed indefinitely without causing unacceptable effects. Thorough discussions
of the CCC and other elements of NRWQC are contained in the methodology for developing aquatic life
criteria, titled Guidelines for Deriving Numerical National Water Quality Criteria for the Protection of
Aquatic Organisms and Their Uses (U.S. EPA, 1985).
4-1
-------
Aquatic organism test data, composing the second tier of source data for the aquatic life
component of TWFs, are available for only 46 chemicals. The results of tests on three types of indicator
organisms are used in TWFs: fish, macroinvertebrates, and algae. Chronic effects values from organism
tests indicate concentrations above which long-term effects are likely to occur. Acute toxicity test results,
such as those for short-term lethal concentrations, are used when chronic toxicity information is not
available for a chemical. Organism test data were obtained from two complementary EPA databases: the
Ecotoxicity Database (ECOTOX) (U.S. EPA, 201 la) and the Assessment Tools for Evaluation of Risk
Database (ASTER) (U.S. EPA, 2011b).
Estimates of chronic toxicity, taken from several sources, are used as TWF aquatic life values
when NRWQC or organism test data are not available. The primary source of chronic toxicity estimates is
EPA's Quantitative Structure Activity Relationship (QSAR) model, which derives statistical relationships
between chemical properties and biological effects. QSAR toxicity estimates are maintained in ASTER.
The TWF Database contains a complete list of other less frequently cited sources of toxicity estimates.
4.1 The Aquatic Life Component of the TWF Equation References
1. U.S. EPA. 1985. U.S. Environmental Protection Agency. Office of Research and Development.
Guidelines for Deriving Numerical National Water Quality Criteria for the Protection of Aquatic
Organisms and Their Uses. Washington, DC. Available online at:
http://water.epa.gov/scitech/swguidance/standards/criteria/aqlife/upload/85guidelines.pdf.
2. U.S. EPA. 2009. U.S. Environmental Protection Agency. Office of Water. National
Recommended Water Quality Criteria. Washington, DC. Available online at:
http://water.epa.gov/scitech/swguidance/standards/current/upload/nrwqc-2009.pdf.
3. U.S. EPA. 201 la. U.S. Environmental Protection Agency. Office of Research and Development.
ECOTOX Database Website. Washington, DC. Available online at: http://cfpub.epa.gov/ecotox.
4. U.S, EPA. 201 lb. U.S. Environmental Protection Agency. Office of Research and Development.
Assessment Tools for the Evaluation of Risk (ASTER) Website. Washington, DC. Available
online at: http://www.epa.gov/med/Prods Pubs/aster.htm.
4-2
-------
5. The Human Health Component of the TWF Equation
The derivation of the human health component of the TWF equation is more complex than that of
the aquatic life component, which is simply an effects value taken from a nationally-recognized source.
The human health values are derived from separate equations to estimate effects thresholds for
noncarcinogens and non-threshold cancer risk-based factors for carcinogens. These equations are based
on EPA's Methodology for Deriving Ambient Water Quality Criteria for the Protection of Human Health:
Technical Support Document Volume 1: Risk Assessment (U.S. EPA, 2000). The noncarcinogen and
carcinogen human health values are based on assumptions about fish consumption, which is the most
common pathway for human exposure to toxic chemicals. Both values represent maximum possible toxic
chemical concentrations in fish, based on the best available toxicological science, that are considered to
minimize the long-term exposure risk to the general population.
5.1 Equation for Calculating the TWF Human Health Value for Noncarcinogens
If a chemical is not considered to have the potential to cause cancer in humans, based on EPA's
Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005), the human health effects value is derived
using a threshold concentration for noncancer adverse effects, which is referred to as the oral reference
dose (RfD). More specifically, the RfD is the amount of a noncarcinogen to which a person may be
exposed each day for a lifetime of 70 years without suffering harmful effects. The RfD includes sensitive
subgroups. Usually doses less than the RfD are unlikely to be associated with adverse health effects.
However, RfD uncertainty may span up to an order of magnitude. The equation for calculating the TWF
human health value for noncarcinogens, shown in Figure 5-1, uses the same adult body weight (70 kg),
and fish consumption rate (17.5 grams per day), that are used in the calculation of water quality criteria
(U.S. EPA, 2000). It should be noted that, in accordance with the NRWQC methodology, a different fish
consumption rate of 18.7 grams per day is used for calculating the TWF human health value for mercury.
Mg
TWF Human Health Noncarcinogenic Valu^) =
Li
f~\
Reference Dose -j2- x Relative Soure Contribution (%) x Bodyweight (kg) x |ig to mg Conversion
\ day J
Daily Fish Consumption Estimate x Bioaccumulation or Bioconcentration Factor
RfD (-Zl) x RSC (%) x 70 kg x 1000 ^
y day J v ° mg
0.0175 (-^-Y x BAF or BCF ()
\dayJ Vkg/
aIn accordance with the NRWQC methodology, a different fish consumption rate of 18.7 grams per day is used for calculating
the TWF human health value for mercury.
Figure 5-1. Equation for Calculating the TWF Human Health Value for Noncarcinogens
5-1
-------
5.2 Equation for Calculating the TWF Human Health Value for Carcinogens
EPA does not calculate human health effects thresholds for carcinogens because the Agency
assumes that there is no safe level of exposure. Instead, EPA estimates the lifetime risk of developing
cancer using a cancer potency factor based on a linear relationship between dose and response. The
cancer potency factor represents an upper-bound estimate of the lifetime probability that an individual
will develop cancer from a specific life time dose of a carcinogen. Lifetime cancer risk is expressed in
terms of the likelihood that one person in an equally exposed population of a specific size would develop
cancer if exposed continuously 24 hours per day to a specific concentration of a carcinogen throughout a
lifetime of 70 years.
TWF Human Health Carcinogenic Valu^) =
Li
cancer Risk Level x Relative Soure Contribution (%) x Bodyweight (kg) x |igto mg Conversion
Cancer Potency Factor (kg day
Daily Fish Consumption Estimate x Bioaccumulation or Bioconcentration Factor
10"5 Hg / kg \a / L \
. . x RSC (%) x 70 kg x 1000 0.0175 :- xBAForBCF
nrvRirtru- kBday\ V J 5 mg \dayj \ksj
Cancer Potency Factor \day) Vkg/
aIn accordance with the NRWQC methodology, a different fish consumption rate of 18.7 grams per day is used for calculating
the TWF human health value for mercury.
Figure 5-2. Equation for Calculating the Human Health Component of TWFs for
Carcinogens
The level of cancer risk that is of concern is a matter of personal, community, and regulatory
judgment. In general, EPA considers a lifetime risk of developing cancer below one in a million (10~6) to
be so small as to be negligible, and cancer risk of more than one in ten thousand (10~4) to be sufficiently
large to warrant action, such as the remediation of a Superfund site. The need for action to address a
lifetime cancer risk between 10"6 and 10"4 is evaluated on a case-by-case basis.
When TWFs were first developed in 1981, EPA decided to use a cancer level of 10"5 for
calculating the human health value of TWFs for carcinogens to be consistent with the 10"5 cancer risk
level used in the calculation of NRWQC. However, a more protective risk level of 10"6 began to be used
in the calculation of water quality criteria when the NRWQC methodology was revised in 2000 (U.S.
EPA, 2000). In order to maintain data consistency for comparing TWF-based analytical results from year
to year, EPA decided to continue using 10"5 in TWF calculations as shown in Figure 5-2.
5-2
-------
The equation for calculating the TWF human health value for carcinogens uses the same adult
body weight (70 kg), and fish consumption rate (17.5 grams per day), that are used in the TWF human
health value for noncarcinogens.
5.3 Data Sources for Calculating the Human Health Component of the TWF Equation
Since there is no single comprehensive source of toxicity information, EPA draws data from
multiple sources to calculate TWF human health values. As illustrated in Figure 5-3, RfD and cancer
potency factors are taken from a hierarchy of sources based on criteria outlined in Section 2.4.6 of the
Technical Support Document for Water Quality-Based Toxics Control (U.S. EPA, 1991).
Tier I
EPA National Recommended Water Quality Criteria, published after 2000 (EPA, 2007)
EPA Integrated Risk Information System (IRIS) Oral RIDs or Oral Cancer Potency Factors (EPA, 201 la)
Tier II
EPA's Health Effects Assessment Summary Tables (HEAST)
EPA's Region III Risk-Based Concentration Table (RCB)
California Environmental Protection Agency health assessments and toxicity criteria (CalEPA)
Tier III
Criteria from EPA National Recommended Water Quality Criteria Documents , prior to 2000
Agency for Toxic Substances and Disease Registry (ATSDR) profiles
Figure 5-3. Hierarchy of Data Sources for Calculating the Human Health Component of
TWFs
EPA's Integrated Risk Information System (IRIS) and NRWQC published after 2000 are
preferred sources of RfDs and cancer potency factors for calculating TWF human health values. IRIS
contains quantitative and qualitative information on effects that may result from exposure to
environmental contaminants (U.S. EPA, 201 la). IRIS information is used to support risk assessments,
policy development, and regulatory activities. IRIS has varying levels of data for most of the chemicals
typically assessed by the effluent guidelines program. NRWQC are developed through rulemaking
process.
If RfDs or cancer potency factors are not available in IRIS or NRWQC developed after 2000,
then other sources are checked beginning with the Health Effects Assessment Summary Tables (HEAST).
The HEAST tables are compiled from toxicity studies by EPA's National Center for Environmental
Assessment (NCEA) for analysis of Superfund and hazardous waste sites. If RfDs or cancer potency
factors are not available in HEAST, then EPA Region 3's Risk-based Concentration Table (RBC) is
checked. The RBC is similar to HEAST and also used to support analysis of Superfund and hazardous
5-3
-------
waste sites EPA, (U.S. EPA, 201 lb). If RfDs or cancer potency factors are not available in HEAST or
RBC, California Environmental Protection Agency's Office of Environmental Health Hazard Assessment
(CalEPA) health assessments and toxicity criteria are checked. If none of these data sources have RfDs or
cancer potency factors for a chemical, then NRWQC developed prior to 2000 are checked. If still no data
are available, then EPA searches the Agency for Toxic Substances and Disease Registry (ATSDR)
toxicological profiles (ASTDR, 2011).
If a chemical has both noncarcinogenic and carcinogenic effects, and if both the RfD and cancer
potency factor are from the same tier in the data hierarchy, then the TWF is calculated using the cancer
potency factor. For example, if IRIS contains both an RfD and cancer slope factor for a chemical, then the
TWF is calculated using the cancer slope factor. If, however, the RfD and cancer potency factor for a
chemical come from sources on different tiers in the data hierarchy, then the value from the higher tier
data source is used. For example, if an RfD for a chemical is available from IRIS but the cancer potency
factor is only available from the EPA Region 3 RBC table, then the TWF is calculated using the RfD.
5.4 Calculating the TWF Human Health Value for Dioxins
The calculation of the TWF human health value for dioxin compounds differs slightly from that
of other carcinogens. The term dioxin refers to a set of chlorinated compounds that share similar chemical
structures and toxic properties. Dioxin compounds, which are among the chemicals in the TWF Database
with the highest TWF values, have widely ranging toxicities and potential to increase the risk of cancer.
Dioxin compounds include seven poly chlorinated dibenzodioxins (PCDDs) and 10 polychlorinated
dibenzofurans (PCDFs). There are TWF values for all of these dioxin compounds in the TWF Database.
PCDDs and PCDFs are unintentional byproducts of most forms of combustion and several
industrial chemical processes. Although release of dioxins to the environment has dropped sharply during
the past two decades - by 90 percent or more - they are still a concern because of their long persistence in
the environment and high toxicity. Reservoirs of 2,3,7,8-tetrachlorodibenzo-p-dioxin, for instance, which
has the second highest TWF after radium-228, can persist for decades in soils and sediments before
eventually entering the human food chain. Dioxin compounds, moreover, especially PCDDs, have high
octanol to water partition coefficient values, which means that they easily bioaccumulate in the fatty
tissues of fish and biomagnify up the food chain.
Because dioxin compounds have widely ranging toxicities and typically occur together in
mixtures, the World Health Organization has developed Toxic Equivalency Factors (TEF) to assess their
toxic potency relative to 2,3,7,8-tetrachlorodibenzo-p-dioxin. As shown in Table 5-1, the cancer potency
factor of each dioxin compound is adjusted by a TEF relative to the cancer potency factor of 2,3,7,8-
tetrachlorodibenzo-p-dioxin (U.S. EPA, 1985).
5-4
-------
Table 5-1. Toxic Equivalency Factors (TEF), Bioaccumulation Factors (BAF), and TWF Values for
Dioxin Compounds
Dioxin
III
i i:k-
Adjlisk'd
('¦nicer
Poleno
l";icloi-
IJAI-"
T\\ 1
A(|ii;ilic
Life Value
(pii/l.)
T\\ 1 lliiiiiiin
llciillh \ iilue
T\\ 1
Tetrachlorodibenzo-p-dioxin, 2,3,7,8-
1
160,000
31,160
5,600,000
697,984,000
703,584,000
Pentachlorodibenzo-p-dioxin, 1,2,3,7,8-
1
160,000
30,934
NA
692,928,000
692,928,000
Pentachlorodibenzofuran, 2,3,4,7,8-
0.5
80,000
49,760
NA
557,312,000
557,312,000
Hexachlorodibenzofuran, 2,3,4,6,7,8-
0.1
16,000
22,859
NA
51,204,160
51,204,160
Hexachlorodibenzofuran, 1,2,3,7,8,9-
0.1
16,000
21,120
NA
47,308,800
47,308,800
Tetrachlorodibenzofuran, 2,3,7,8-
0.1
16,000
19,549
29,474
43,790,080
43,819,554
Hexachlorodibenzo-p-dioxin,
1,2,3,4,7,8-
0.1
16,000
10,490
NA
23,498,240
23,498,240
Hexachlorodibenzofuran, 1,2,3,6,7,8-
0.1
16,000
6,299
NA
14,109,440
14,109,440
Hexachlorodibenzo-p-dioxin,
1,2,3,7,8,9-
0.1
16,000
4,730
NA
10,595,840
10,595,840
Hexachlorodibenzo-p-dioxin,
1,2,3,6,7,8-
0.1
16,000
4,266
NA
9,556,480
9,556,480
Hexachlorodibenzofuran, 1,2,3,4,7,8-
0.1
16,000
2,571
NA
7,632,640
5,760,000
Pentachlorodibenzofuran, 1,2,3,7,8-
0.05
8,000
6,815
NA
5,760,000
7,632,640
Heptachlorodibenzofuran, 1,2,3,4,7,8,9-
0.01
1,600
13,545
NA
3,033,984
3,033,984
Heptachlorodibenzo-p-dioxin,
1,2,3,4,6,7,8-
0.01
1,600
1,835
NA
411,136
411,136
Heptachlorodibenzofuran, 1,2,3,4,6,7,8-
0.01
1,600
383
NA
85,760
85,760
Octachlorodibenzofuran
0.0001
16
590
5,600
986
2,021
Octachlorodibenzo-p-dioxin
0.0001
16
440
700
1,321
6,586
Methylmercury
1,225,029
NA
23,494
23,494
Source: Toxic Equivalency Factors (TEFs) for PCBs, PCDDs, PCDFs for Humans and Wildlife (Van den Berg,
1998).
a - The aquatic life and human health component values in this table have been normalized to copper. There are no
TWF aquatic life values for most dioxins.
b - Methymercury data are shown for comparison. Currently there is no cancer potency factor for methylmercury,
nor is there an aquatic life value.
Given the high toxicity of dioxins and their high potential for biomagnification in aquatic food
chains, it is worth noting that only a small number of facilities discharge small amounts of dioxins to
surface water. In 2008, the most recent year for which discharge monitoring data were available for this
report, only six facilities submitted DMRs showing the discharge of dioxin and dioxin-like compounds.
Four facilities in the pulp and paper point source category reported a total annual discharge of 0.000208
pounds of dioxins (145,000 TWPE). Two facilities in the chlorine and chlorinated hydrocarbons point
source category reported a total annual discharge of 0.000619 pounds of dioxins (436,000 TWPE) (U.S.
EPA, 2010).
5-5
-------
5.5
Variable Factors Used in the Human Health Component of the TWF Equation
Two variable factors representing sources of exposure and aquatic food chain bioaccumulation
are used in the human health component of the TWF equation to adjust the otherwise constant factors
representing RfD, cancer potency factor, body weight, and fish consumption rate: 1) relative source
contribution factors and, 2) bioaccumulation and bioconcentration factors.
5.5.1 Relative Source Contribution Factors
The relative source contribution (RSC) factor is used in the calculation of the human health
component of TWFs to apportion exposure pathways of a chemical among drinking water, fish
consumption, skin contact, or inhalation. The RSC is applied to both the RfD for noncarcinogens and the
cancer potency factor for carcinogens. The RSC approach is documented in the Methodology for Deriving
Ambient Water Quality Criteria for the Protection of Human Health (U.S. EPA, 2000). EPA, however,
has published RSCs for only 28 chemicals. The RSCs were calculated as part of the development of
NRWQC. For TWF human health calculations for all other chemicals, fish consumption is assumed to be
the only exposure pathway, and thus an RSC of 100 percent is applied. Consequently, the drinking water
exposure pathway is assumed to be zero percent for most TWFs. EPA considers this assumption to be
reasonable for the purpose of screening-level analysis, considering that most Americans drink treated
water, or tested well water, and thus generally are not exposed to unsafe levels of toxic chemicals with
TWFs through drinking water.
5.5.2 Bioaccumulation and Bioconcentration Factors
Toxic chemicals have a range of physical characteristics which determine how readily they
partition from an aqueous phase into an organic phase in the surface water environment. In fish,
chemicals typically partition into the organic phase in the gills. An aquatic bioaccumulation factor (BAF)
is the ratio of a toxic chemical in an organism to its concentration in the ambient water, considering
exposure from all potential pathways. An aquatic bioconcentration factor (BCF) is the ratio of a toxic
chemical in an organism to its concentration in the ambient water, considering direct contact with the
water as the only exposure pathway. Bioaccumulation typically accounts for significantly more chemical
build up in tissues than bioconcentration, especially for highly toxic and highly persistent chemicals like
dioxin compounds. National-level BAFs are not available, however, because they are complex to develop.
BCFs, on the other hand, are readily available, and EPA uses them to calculate most TWFs with
exception of the 17 dioxin compounds and methylmercury as shown in Table 5-1.
EPA estimates BAFs for dioxins because of their extremely high potential to bioaccumulate. EPA
derives BAFs for dioxins by combining estimates of the percentage of fish tissue that is fat with estimates
of the tendency of a dioxin to be absorbed in fat. These estimates were developed by the Great Lakes
5-6
-------
Water Quality Initiative (GLWQI) (U.S. EPA, 1995). EPA applied the GLWQI estimates to equations for
BAF calculation published in EP A's Methodology for Deriving Ambient Water Quality Criteria for the
Protection of Human Health (U.S. EPA, 2000).
EPA estimates the BAF for methylmercury differently than it does for dioxins. Unlike other
NRWQC, the water quality criterion for methymercury is based on fish tissue concentration rather than on
ambient water concentration. For the purpose of calculating methymercury TWFs, EPA converts the fish
tissue criterion of 0.3 milligrams methylmercury per kilogram in fish tissue to a national BAF estimate of
1,225,029 using EPA published methodology (EPA, 2001).
The TWF Database contains the complete list of sources for BCFs used in TWF calculations.
Most BCFs come from the following sources:
"Estimating Bioconcentration Potential from Octanol/Water Partition Coefficients"
(Veith and Kosian, 1983);
Assessment Tools for Evaluation of Risk (ASTER) Database (U.S. EPA, 1999);
"Evaluation of Waterborne Exposure Pathways to Paragraph 4(c) Pollutants" (Little,
1983); and
Handbook of Chemical Property Estimation Methods - Environmental Behavior of
Organic Compounds (Lyman et al., 1981).
5.6 The Human Health Component of the TWF Equation References
1. ASTDR. 2011. Agency for Toxic Substances and Disease Registry. ASTDR Website. Available
online at: http://www.atsdr.cdc.gov/.
2. Little, Arthur D. 1983. "Evaluation of Waterborne Exposure Pathways to Paragraph 4(c)
Pollutants." Draft Report. (April 28).
3. Lyman WJ, WF Reehl, and DH Rosenblatt. 1981. Handbook of Chemical Property Estimation
Methods - Environmental Behavior of Organic Compounds. New York, NY: McGraw-Hill Book
Company.
4. U.S. EPA. 1985. U.S. Environmental Protection Agency. Office of Health and Environmental
Assessment. Health Assessment Document for Polychlorinated Dibenzo-p-dioxins. Washington,
DC. EPA-600-8-84-814F.
5. U.S. EPA. 1991. U.S. Environmental Protection Agency. Office of Water. Technical Support
Document for Water Ouality-Based Toxics Control. Washington, DC. (March). EPA-HQ-OW-
2004-0032-0935
6. U.S. EPA. 1995. U.S. Environmental Protection Agency. Great Lakes Water Quality Initiative
Technical Support Document for the Procedure to Determine Bioaccumulation Factors. EPA-
820-B-95-005. (March).
7. U.S. EPA. 1999. U.S. Environmental Protection Agency. Assessment Tools for Evaluation of Risk
(ASTER) Database. Duluth, MN: U.S. EPA, Environmental Research Laboratory.
5-7
-------
8. U.S. EPA. 2000. U.S. Environmental Protection Agency. Office of Water. Methodology for
Deriving Ambient Water Quality Criteria for the Protection of Human Health: Technical Support
Document Volume 1: Risk Assessment. EPA-822-B-00-005. Washington, DC. (October).
9. U.S. EPA. 2001. U.S. Environmental Protection Agency. 66 FR 1344. Water Quality Criteria:
Notice of Availability of Water Quality Criterion for the Protection of Human Health:
Methylmercury, (January 8).
9. U.S. EPA. 2005. U.S. Environmental Protection Agency. Guidelines for Carcinogen Risk
Assessment. EPA/630/P-03/001F. Washington, DC. (March). Available online at:
http://www.epa.gov/raf/publications/pdfs/CANCER GUIDELINES FINAL 3-25-05.pdf.
10. U.S. EPA. 2007. U.S. Environmental Protection Agency. Water Quality Standards Handbook:
Second Edition. EPA-823-B-94-005. Washington, DC. (June). Available online at:
http://water.epa.gov/scitech/swguidance/standards/handbook/
11. U.S. EPA. 2010. U.S. Environmental Protection Agency. Technical Support Document for the
2010 Effluent Guidelines Program Plan. EPA 820R-10-021. Washington, DC. (December). EPA-
HQ-OW-2008-0517 DCN 07320.
12. U.S. EPA. 201 la. U.S. Environmental Protection Agency. Integrated Risk Information System
(IRIS) Website. Available online at: http://www.epa.gov/IRIS/.
13. U.S. EPA. 201 lb. U.S. Environmental Protection Agency. Mid-Atlantic Risk Assessment -
Human Health Risk Assessment Website. Available online at:
http://www.epa.gov/reg3hwmd/risk/human/index.htm.
14. Van den Berg, et al., 1998. Toxic Equivalency Factors (TEFs) for PCBs, PCDDs, PCDFs for
Humans and Wildlife. Environ. Health Perspect. 106:775-792.
15. Veith, GD and P. Kosian. 1983. "Estimating Bioconcentration Potential from Octanol/Water
Partition Coefficients." In D. Mackay et al. (Eds.) "Physical behavior of PCBs in the Great
Lakes". Ann Arbor Scie. Publ. Ann Arbor, ML269-282.
5-8
-------
6. EPA's Use of TWFs to Support Effluent Guidelines Rulemakings and Program
Planning
EPA's effluent guidelines program uses TWFs for two specific purposes. Since the early 1980s,
EPA has used TWFs to help inform regulatory decision making about the relative cost effectiveness of
technology options to treat effluent containing toxic chemicals. In 2003, EPA began using TWFs in
effluent guidelines program planning to rank industries by their total annual toxic-weighted discharges to
identify major sources nationwide that may warrant further analysis.
EPA has issued effluent guidelines for 56 industrial categories encompassing 35,000 to 45,000
facilities that discharge directly to waters of the United States. The effluent guidelines program has also
issued pretreatment standards that apply to approximately 12,000 facilities that discharge to publically
owned treatment works (POTWs). Each year, effluent guidelines and pretreatment standards prevent the
discharge of more than 1.2 billion pounds of toxic pollutants.
EPA does not use TWFs in water quality-based permitting or risk assessment because they do not
provide the information on site-specific water chemistry, environmental alteration of chemicals, exposure
pathways, and organism susceptibility required to adequately set permit limits or assess risk. Moreover,
EPA only uses TWFs to help inform decision making, and never as a sole source of information on which
to base policy.
6.1 Use of TWFs in Cost-Effectiveness Analysis
As explained in the introduction of this report, TWFs are used to calculate Toxic Weighted Pound
Equivalents (TWPEs), which are employed in cost-effectiveness analysis. Cost effectiveness is defined as
the incremental annual cost, expressed in 1981 constant dollars, per incremental toxic weighted pound of
pollutant removed by a treatment technology (U.S. EPA, 1988). As shown in Table 6-1, the use of
constant 1981 dollar values allows effluent guidelines developed and revised at different times to be
compared with one another.
Cost-effectiveness analysis is used to help evaluate a set of treatment technologies to identify the
one that has the lowest cost per pound of toxic pollutants from a waste stream. It should be noted that
there is no absolute scale for determining whether or not a treatment technology option is cost effective.
The cost effectiveness of treatment technologies are considered to be high or low relative to one another.
Although not required by the Clean Water Act, cost effectiveness analysis is a useful tool for
effluent guidelines regulatory analysis. It is, however, just one of many factors considered during
evaluation of treatment technology options. Moreover, cost-effectiveness analysis is not always used in
effluent guidelines regulatory development. It is only used to help assess regulatory options for industries
6-1
-------
that discharge toxic chemicals classified by the Clean Water Act (CWA) as "Priority Pollutants," which
are 126 chemicals specified in the Act, and for substances the CWA classifies as "Non-conventional
Pollutants," such as chlorides, iron, and ammonia. Substances the CWA classifies as "Conventional
Pollutants," such as oil and grease, biochemical oxygen demand, and total suspended solids are not
included in cost-effectiveness analysis because they are addressed in a different analogous analysis that
pertains to all pollutants, not just toxic chemicals.
Table 6-1. Examples of Effluent Guidelines Cost-Effectiveness Analysis Ranked by Dollars Per
TWPE Removed
Indusln
I.N(iin;ik'(l
T\\ IT.
Dischiiriiod
Prior to Rule
(Ihoiisiiiids)
I.N(iiii;iU'(l l \\ IT
Dischiii'uod iil'lcr
Impk-iiK'iiliilioii of
Rule (Ihoiisiiiids)
l.N(iin;iU'd TWI'i:
RcinoM'd b\ Rule
(Ihoiisiiiids)
( »sl-r.lTccli\eiK'ss
ol Rule (I'JXI
Dolliirs per TWIT.
mno\cd)
Inorganic Clieniiculb 1
1,290
31,213
1
Battery Manufacturing
4,126
5
4,121
2
Nonferrous Metals
Manufacturing I
6,653
313
6,340
4
Organic Chemicals
54,225
9,735
44,490
5
Inorganic Chemicals II
605
27
578
6
Nonferrous Metals
Manufacturing II
1,004
12
992
6
Porcelain Enameling
1,086
63
1,023
6
Centralized Waste
Treatment
3,372
1,267
2,105
5-7
Canmaking
12
0
12
10
Metal Finishing
3,305
3,268
37
12
Pesticides
2,461
371
2,090
14
Copper Forming
70
8
62
27
Iron & Steel
1,053
853
200
27
Pulp & Paper
61,713
2,628
59,085
39
Coil Coating
2,289
9
2,280
49
Metal Products &
Machinery
140
70
70
50
Nonferrous Metals Forming
34
2
32
69
Foundries
2,308
39
2,269
84
Pharmaceuticals
89,790
471
89,320
47-96
Aluminum Forming
1,340
90
1,250
121
Electronics I
9
3
6
404
Source: Adapted from Cost-Effectiveness Analysis of Effluent Limitation Guidelines and Standards for the
Centralized Waste Treatment Industry (U.S. EPA, 2000).
Data used in conducting cost-effectiveness analyses may come from a variety of sources. Most
typically, however, data are compiled through industry surveys and effluent sampling and analysis
specifically designed for individual effluent guidelines rulemakings. The number of chemicals assessed in
6-2
-------
a cost effectiveness-analysis may range from only a few to more than 100 depending on the nature of an
industry's waste streams. For the centralized waste treatment industry, for example, cost effectiveness
was assessed for technologies to treat a set of 146 chemicals.
6.2 Use of TWFs in Effluent Guidelines Program Planning
CWA Section 304(b) requires EPA to annually review industrial categories for which effluent
guidelines have been promulgated to determine if any revisions are warranted, and to identify possible
new industrial categories for regulation. Section 304(m) requires EPA to publish a biennial plan
announcing the schedule for the annual reviews, as well as the schedule for completing any effluent
guidelines revisions or new regulations identified during the annual reviews. EPA has published a draft
program plan every odd-numbered year since 1989 and a final program plan every even-numbered year
since 1990.
EPA began using TWPEs in 2003 to rank industrial categories by their total annual discharge
amounts to support development of the 2004 effluent guidelines program plan (U.S. EPA, 2004). The use
of TWPEs as an indicator of the relative potential for industrial waste streams to cause harm prior to
discharge assists EPA in setting priorities for effluent guidelines rulemakings so as to achieve the greatest
environmental and health benefits from toxic discharge reduction. It is important to note that TWPE-
based industry ranking is just one of four factors considered in effluent guidelines program planning, in
addition to consideration of public comments. EPA uses the following three additional factors in its
annual review of industrial categories:
1. The performance and cost of applicable and demonstrated wastewater treatment
technologies, process changes, or pollution prevention alternatives that could effectively
reduce the pollutants in the industrial category's wastewater and, consequently, reduce
the potential harm to human health or the environment associated with pollutant
discharges;
2. The affordability or economic achievability of the identified wastewater treatment
technologies, process changes, or pollution prevention measures; and
3. Opportunities to eliminate inefficiencies or impediments to pollution prevention or
technological innovation or opportunities to promote innovative approaches. This factor
might also prompt EPA during an annual review to decide against identifying an existing
set of effluent guidelines or pretreatment standards for revision where the pollutant
source is already efficiently and effectively controlled by other regulatory or
nonregulatory programs.
EPA calculates two TWPE estimates for each industrial category based on data from individual
facility discharge monitoring reports (DMRs) and EPA's Toxics Release Inventory (TRI). Facilities are
required to submit DMRs to EPA or state permitting agencies as a condition of their National Pollutant
6-3
-------
Discharge Elimination System (NPDES) water discharge permits. DMRs are stored in the Permit
Compliance System (PCS) and the Integrated Compliance Information System for the National Pollutant
Discharge Elimination System (ICIS-NPDES). EPA began replacing PCS with ICIS-NPDES in 2006.
Until the transition is complete, EPA retrieves certain data from both databases.
TRI is the common name for Section 313 of the Emergency Planning and Community Right-to-
Know Act (EPCRA) (U.S. EPA, 2009a). Facilities above certain reporting thresholds, such as size and
amount of toxic chemicals discharged, must annually report estimates of discharges of certain toxic
chemicals listed under EPCRA. Unlike DMRs, TRI reports are primarily based on estimates rather than
on sampling and analysis. TRI reports, however, include estimates of toxic chemical waste streams sent to
POTWs, which DMRs do not. TRI also includes estimates of some pollutant discharges that are not
currently regulated (i.e., pollutants not in DMRs).
When facilities discharge indirectly, POTWs remove varying amounts of toxic chemicals from
industrial waste streams, depending to the type of treatment process they use, before discharging the
treated effluent to surface waters. This type of discharge is referred to as an indirect discharge. For the
purpose of effluent guidelines program planning, EPA primarily uses TRI data to help estimate the
amount of indirect discharges from industrial sources. EPA only uses TRI data to support effluent
guidelines program planning, not for cost-effectiveness analysis.
Table 6-2. TWPE Ranking of Industrial Categories Based on 2008 DMR Data3
Kiink
Poiul Source Csilcgon
Number of
Dischiiriicrs
l oliil Aniuiiil Pounds
1 Olill l \\ PI
1
Nonferrous Metals
Manufacturing
68
1,760,000,000
955,000
2
Fertilizer Manufacturing
54
160,000,000
818,000
3
Petroleum Refining
727
7,740,000,000
618,000
4
Organic Chemicals, Plastics
And Synthetic Fibers b
545
4,740,000,000
512,000
5
Pulp, Paper And Paperboard
267
2,350,000,000
510,000
6
Ore Mining And Dressing
117
558,000,000
339,000
7
Textile Mills
106
39,500,000
247,000
8
Waste Combustors
306
3,970,000,000
245,000
9
Inorganic Chemicals
Manufacturing b
206
1,790,000,000
228,000
10
Miscellaneous Foods And
Beverages
152
110,000,000
193,000
Source: Technical Support Document for the Preliminary 2010 Effluent Guidelines Program Plan (U.S. EPA,
2009b).
a Excludes categories currently or recently under review for a potential rulemaking or rulemaking revision.
b Excludes facilities manufacturing chlorine or chlorinated hydrocarbons, as these discharges are currently
undergoing review for possible rulemaking revision.
6-4
-------
Table 6-2 shows the ten industrial categories with the highest current TWPE discharges according
the 2008 DMR data analyzed for the 2010 effluent guidelines program plan. Although DMR and TRI data
show that EPA regulatory programs are reducing discharges of toxic chemicals over time, discharge
amounts of individual toxic chemicals vary from year to year, depending largely on economic conditions.
The heaviest industries, such as metal manufacturing, petroleum refining, and ore mining, produce the
highest toxic discharges, and thus tend to be among the highest ranked industries from year to year.
Table 6-3 shows the ten industrial categories with the highest TWPE discharges according the
2008 TRI data analyzed for the 2010 effluent guidelines program plan. The discrepancies between Table
6-2 and Table 6-3 are the result of differing DMR and TRI reporting requirements. TRI contains
discharge data for fewer facilities due to reporting thresholds that only require larger facilities to submit
data. TRI also reflects a lower estimate of the pounds of pollutant discharged, for the same reason.
Table 6-3. TWPE Ranking of Industrial Categories Based on 2008 TRI Data"
Number of
1 (M;il Anniiiil
1 (Hill
Kiink
Poiul Soiiiyc Csilcgon
Dischii r^crs
Pounds
i w pi:
1
Pulp, Paper And Paperboard
250
19,300,000
523,000
2
Petroleum Refining
298
25,700,000
410,000
3
Organic Chemicals, Plastics And Synthetic Fibers b
679
27,600,000
137,000
4
Ore Mining And Dressing
34
491,000
109,000
5
Plastics Molding And Forming
123
2,200,000
74,700
6
Inorganic Chemicals Manufacturing b
162
8,510,000
71,300
7
Nonferrous Metals Manufacturing
122
4,190,000
38,700
8
Pesticide Chemicals
73
2,090,000
35,500
9
Carbon Black Manufacturing
7
279
27,600
10
Timber Products Processing
122
48,400
27,300
Source: Technical Support Document for the Preliminary 2010 Effluent Guidelines Program Plan (U.S. EPA,
2009b).
a Excludes categories currently or recently under review for a potential rulemaking or rulemaking revision.
b Excludes facilities manufacturing chlorine or chlorinated hydrocarbons, as these discharges are currently
undergoing review for possible rulemaking revision.
For the purpose of effluent guidelines program planning analysis, EPA adds the TWPEs
calculated from TRI and DMR data to produce a total TWPE value for each industrial category. EPA
takes this approach because the combination of DMR and TRI data helps to overcome certain limitations
in each data source and provides a more comprehensive portrait of total toxic discharges. A thorough
discussion of the effluent guidelines program, the effluent guidelines planning process, use of TWFs and
TWPEs in industry ranking, and the complete list of industry rankings is contained in the Technical
Support Document for the Preliminary 2010 Effluent Guidelines Program Plan (U.S. EPA, 2009b).
6-5
-------
6.3 EPA's Use of TWFs to Support Effluent Guidelines Rulemakings and Program Planning
References
1. U.S. EPA. 1988. U.S. Environmental Protection Agency. Office of Water. Ehrensberger, K., and
R. Rico. Cost-effectiveness Analysis for Effluent Guidelines. Office of Water. Washington, DC.
(May).
2. U.S. EPA. 2000. U.S. Environmental Protection Agency. Office ofWater. Cost-Effectiveness
Analysis of Effluent Limitation Guidelines and Standards for the Centralized Waste Treatment
Industry. Washington, DC. (December). Available online at:
http://water.epa.gov/scitech/wastetech/guide/treatment/upload/2006 12 28 guide cwt final effe
ctive.pdf.
3. U.S. EPA. 2004. Technical Support Document for the 2004 Effluent Guidelines Program Plan.
EPA 821R-04-014. Washington, DC. (August). EPA-HQ-OW-2003-0074-1346 through 1352.
4. U.S. EPA. 2009a. U.S. Environmental Protection Agency. 2009 Toxics Release Inventory
National Analysis Overview. Washington, DC. (Unknown). Available online at:
http://www.epa.gOv/tri/tridata/tri09/nationalanalvsis/overview/2009TRINAOverviewfmal.pdf.
5. U.S. EPA. 2009b. Technical Support Document for the Preliminary 2010 Effluent Guidelines
Program Plan. EPA 821R-09-006. Washington, DC. (October). EPA-HQ-OW-2008-0517-0515.
6-6
-------
7. Limitations and Appropriate Uses of Toxic Weighting Factors
Appropriate uses of TWFs are constrained by the limitations of the data used to derive them, and
by limitations in the discharge data to which TWFs are applied. This section discusses the appropriate
uses of TWFs given the data limitations.
7.1 Limitations of Source Data Used to Derive TWFs
Due to the lack of a single comprehensive source of toxicity information for the range of
chemicals assessed by the effluent guidelines program, EPA must combine disparate sources of effects
information in the calculation of TWFs. The data used to derive TWFs were compiled for different
purposes at different times. Effects information used in the aquatic life component of TWFs is compiled
primarily from three different sources, each based on different underlying analytical methodologies.
Similarly, RfDs and cancer potency factors used to derive the human health component of TWFs are
compiled from a hierarchy of multiple sources based on data quality and availability. Inconsistent source
data is the primary reason why TWFs are not suitable for more rigorous applications, such as analyses to
support water-quality based permitting or risk assessment.
In addition, complete data are lacking for the aquatic life and human health components of the
majority of chemicals for which TWFs have been calculated. Only 37 percent of TWFs (349 of 1,064) are
based on both aquatic life and human health effects information. The majority of TWFs, 62 percent (661),
are based solely on aquatic life values. The remaining five percent of TWFs (54) are derived solely from
human health values.
7.2 Limitations of Discharge Data to Which TWFs Are Applied
There are usually no data limitations issues when conducting cost-effectiveness analyses because
the data are typically compiled through industry surveys and effluent sampling specifically designed for
individual effluent guidelines rulemakings. For the screening-level analysis, however, TWF-based
analysis of PCS/ICIS-NPDES data, must be caveated to note the limitations of these two databases, which
are the only sources of national level facility discharge data available to EPA for effluent guidelines
program planning analysis.
One of the biggest limitations of PCS/ICIS-NPDES is that for many industries, annual discharge
amounts can only be calculated for facilities classified as major dischargers, which are typically the
largest facilities with the most toxic discharges. Facilities with NPDES permits are not required to report
the actual quantities of chemicals they discharge. Instead, they must only report the discharge flow rate,
such as gallons per day, and the concentration of chemicals in the waste stream. EPA combines the flow
and concentration information to produce annual discharge load estimates expressed as pounds discharged
7-1
-------
per year. Facilities classified as minor dischargers, however, are not required to report their discharge
flow. Some states report both flow and concentration information for minor dischargers, others do not. In
addition, there is considerable inconsistency in how state permitting authorities classify facilities as either
major or minor. Moreover, the majority of facilities are classified as minor dischargers in some industries
with significant toxic discharges, such as most coal and ore mining facilities. State permitting authorities
are required to consider the following six factors when classifying facilities as major or minor (U.S. EPA,
2010):
1. Toxic pollutant potential;
2. Discharge flow to stream flow ratio;
3. Conventional pollutant loading;
4. Public health impact;
5. Water quality factors; and
6. Proximity to coastal waters.
In addition to incompleteness of data reported to PCS/ICIS-NPDES, there are several other
important limitations to the database (U.S. EPA, 2009).
PCS/ICIS-NPDES contain data only for the pollutants facilities are required to monitor as
stipulated in their NPDES permits. Facilities are not required to monitor or report all of
the pollutants they actually discharge.
PCS/ICIS-NPDES include very limited data characterizing indirect discharges from
facilities to POTWs.
Many of the pollutant parameters reported to PCS/ICIS-NPDES are not individual
chemical compounds. For example, "oil and grease" is a common industrial parameter
reported to PCS/ICIS-NPDES that can contain a wide variety of industrial chemicals.
Moreover, many elemental chemicals have multiple ionic species with differing toxic
properties. Levels of these multiple ionic species are frequently reported as a single total
value. For example, the various ionic forms of mercury may be reported together as "total
mercury" levels in a waste stream.
Most reported flow rates do not indicate the type of wastewater. Thus total flow rates of
industrial process wastewater may be diluted by stormwater or noncontact cooling water.
Facilities are identified by SIC code, not by effluent guidelines point source category. For
some SIC codes, it may be difficult or impossible to identify the actual industry that is the
source of the reported wastewater discharge.
A primary advantage of TRI is that, unlike PCS/ICIS-NPDES, facilities are required to report
annual discharge amounts of chemicals rather than separate flow and concentration values. However, a
disadvantage of TRI is that facilities are only required to report estimated discharge amounts rather than
7-2
-------
actual sampled discharge amounts. There are several additional notable limitations of the data reported to
TRI (U.S. EPA, 2009).
Information in TRI does not represent total national discharges because not all facilities
are required to report to TRI. Small business with less than ten employees and facilities
that do not meet discharge amount reporting thresholds are not required to report to TRI.
Thus the completeness of discharge information varies from industry to industry.
The list of chemicals that facilities are required to report to TRI is not all-inclusive and
has varied over time.
TRI requirements encourage facilities to report some compounds as present at one-half
the detection level if a facility suspects that the compound has the potential to be present,
even if analysis of discharge samples shows the compound is below its detection level.
Many companies adopt this conservative reporting approach. For facilities with high
discharge flows, this can result in large estimates of pounds discharged with no sampling
data to determine that the compound was actually present.
As with PCS/ICIS-NPDES, certain chemicals are reported as a class, such as polycyclic
aromatic compounds, dioxin compounds, and metal compounds. Because the individual
compounds in the class may have widely varying toxicities, actual toxicities of waste
streams can be difficult to estimate accurately.
Facilities are identified by NAICS code, not by effluent guidelines point source category.
For some NAICS codes, it may be difficult or impossible to identify the point source
category that is the actual source of the toxic wastewater discharge.
Only facilities in certain NAICS codes are required to report, so some sources of toxic
chemical discharges are not included in TRI.
7.3 Appropriate Uses of TWFs
In spite of the limitations of the TWF source data and the discharge data to which they are
applied, EPA considers TWFs appropriate for use in effluent guidelines planning analysis to identify
industries or facilities that may warrant additional research. Moreover, TWPE-based ranking of industries
is just one of four factors that EPA considers when setting priorities for effluent guidelines program.
EPA also considers TWFs appropriate for use in the calculation of cost-effectiveness values
because such values only serve as indicators of the relative cost effectiveness of treatment technology
options and not as absolute metrics. Moreover, as with TWPE-based industry ranking in effluent
guidelines program planning, cost-effectiveness analysis is not used as the sole basis for decision making
about technology options, but rather as one of many factors that EPA considers.
EPA's effluent guidelines program cautions other organizations that may consider using TWFs
for applications other than the specific ones for which they were designed. Other potential users of TWFs
should carefully evaluate TWF source data limitations, the limitations of EPA national discharge
7-3
-------
databases, and the uncertainties created by the TWF approach of combining aquatic life and human health
effects values before deciding to use TWFs.
7.4 Limitations and Appropriate Uses of Toxic Weighting Factors References
1. U.S. EPA. 2009. U.S. Environmental Protection Agency. Office of Water. Technical Support
Document for the Preliminary 2010 Effluent Guidelines Program Plan. EPA 821-R-09-006.
Washington, DC. (October). EPA-HQ-OW-2008-0517-0515.
2. U.S. EPA. 2010. U.S. Environmental Protection Agency. Office of Water. NPDES Permit Writers
Manual. Washington, DC. EPA-833-K-10-001. (September). Available online at:
http: //www. epa. gov/npde s/pubs/pwm 2010.pdf.
7-4
------- |