vyEPA
Region 10 Fact Sheet February 25, 1991
Dioxin Loading to Waters
in the Columbia River Basin
-------
&EPA
i
Region 10 Fact Sheet February 25, 1991
Dioxin Loading to Waters
in the Columbia River Basin
BACKGROUND The water quality of the Columbia River and segments of
the Snake and Willamette Rivers is currently considered
impaired due to concentrations of a form of dioxin. The
pollutant, 2,3,7,8-TCDD, is the most toxic of a group of
compounds known as polychlorinated dibenzo-para-
dioxins. These compounds, although occuring naturally at
very low concentrations, can be found at elevated levels as
a result of human activities such as the manufacture of
chlorinated herbicides, the combustion of domestic and
industrial wastes, and the production of chlorine-bleached
pulp. Concentrations of TCDD measured in fish tissue in
several areas of the Columbia River basin exceed levels
protective of human health. Pulp mills which use chlorine to
bleach paper products have been associated with some of
the highest concentrations of TCDD in surface waters.
Information also exists quantifying levels of TCDD in
effluents from chlorine-bleaching pulp mills in the Columbia
River basin. In order to reduce discharges of TCDD to
acceptable levels in the Columbia River basin, additional
controls are needed on known sources.
WHAT IS A TMDL? Section 303(d) of the Clean Water Act requires each state
(1) to identify waters for which effluent limitations normally
required are not stringent enough to attain water quality
standards and (2) to establish total maximum daily loadings
(TMDLs) on such waters for the pollutant(s) of concern.
The process of developing a TMDL involves the calculation
of the loading capacity (the amount of loading that the river
can receive without violating water quality standards) and
the allocation of allowable loads to point sources, nonpoint
sources, and background. A TMDL, by definition, is the
sum of the individual allocations to point sources, nonpoint
sources and background. It is effectively an implementation
plan for achieving water quality standards which includes an
appropriate margin of safety which takes into account any
lack of knowledge concerning the relationship between
source concentrations and water quality.
-------
Concern about issues of equity in the multi-state basin led
the states of Oregon, Washington, and Idaho to request
that EPA establish the TMDL for dioxin in the Columbia
River basin as a federal action. EPA issued a public notice
of the proposed TMDL and solicited comments on June 14,
1990, pursuant to the requirements of s303(d).
THE FINAL TMDL EPA has considered oral and written testimony received
during the public comment period, and is now establishing
a final TMDL, effective February 25,1991, which provides a
framework to control dioxin discharges to the Columbia
River basin. The TMDL defines the loading capacity of the
entire basin to be about 6 milligrams of 2,3,7,8-TCDD per
day. This value was derived based on an allowable
concentration of .013 parts per quadrillion for 2,3,7,8-TCDD
and the volume of water in the Columbia River.
Because of the vastness of the Columbia River basin and
the need to ensure attainment of water quality standards for
TCDD at all points, the TMDL establishes key checkpoints
within the system. Loading capacities, or "watershed
targets," have been established for the Willamette River
basin (0.54 mg TCDD/day) and the Snake River basin (1.18
mg TCDD/day). Allocations to sources within each of those
watersheds must fit within these watershed targets.
EPA has also established specific wasteload allocations for
the eight chlorine-bleaching pulp mills located in the
Columbia River basin in Oregon, Idaho, and Washington.
These allocations lead to an estimated 95% reduction in
dioxin discharges from these facilities relative to the levels
discharged in 1988. Wastewater discharge permits issued
by EPA and the states of Washington and Oregon must be
consistent with this TMDL EPA also estimated future dioxin
discharges for the Celgar pulp mill in Castlegar, Canada
(expected to be less than 5% of the loading capacity).
EPA believes that other sources of dioxin exist in the
Columbia River basin. In the final TMDL, EPA has identified
other major sources and estimated their loadings of TCDD.
Accordingly, the remainder of the loading capacity
(3.59 mg/day or 60 percent) is reserved to account for
-------
these other sources, for future growth, and as a margin of
safety until adequate data has been collected and evaluated
which either confirms the adequacy of the margin of safety
or supports the establishment of additional or modified
allocations. In order to obtain this additional data, EPA
expects that the states will work in cooperation with EPA to
develop strategies to collect the needed information. If
future data identifies the need to make additional
allocations, or to reduce any existing allocations, a modified
TMDL will be established.
The following figure summarizes the overall structure of the
TMDL:
RESERVED (Unallocated)
WASTE LOAD ALLOCATIONS
PUBLIC COMMENTS The public comment period for the proposed TMDL closed
on July 20,1990. In addition, a public hearing to discuss
the TMDL was held July 17, 1990, in Vancouver,
Washington.
-------
A number of substantive comments were received
concerning the proposed TMDL Major revisions to the
proposed TMDL, which were based upon comments
received, were:
> estimates of the loadings from major sources other than
chlorine bleaching pulp mills were made to be
reasonably sure that the final TMDL will result in the
achievement of water quality standards.
¦ an estimate of the projected loadings from the Celgar
pulp mill in Castlegar, Canada, was incorporated into the
TMDL (this replaced the assumption that the entire
TCDD loading available for the Columbia River at the
U.S.-Canada border be reserved for Canadian sources)
Other issues which were raised during public comment but
which did not result in changes to the TMDL were:
¦ expanding the TMDL to include dioxins and furans other
than 2,3,7,8-TCDD
¦ estimating the attenuation or availability of TCDD in the
sediments
> alternative allocation approaches for the pulp mills
(including allocations of zero for the mills, as well as
allocation of the entire load for the mills)
EPA has developed responses to all significant comments
received on the proposed TMDL Copies of the final TMDL
or the responses to comments can be obtained by writing
EPA Region 10, WD-139, 1200 Sixth Avenue, Seattle, WA,
98101, or calling (206) 553-1086.
-------
1
United States
Environmental Protection Agency
Region 10
1200 Sixth Avenue
Seattle, Washington 98101
TOTAL MAXIMUM DAILY LOADING (TMDL)
TO LIMIT DISCHARGES OF 2,3,7,8-TCDD (DIOXIN)
TO THE COLUMBIA RIVER BASIN
In compliance with the provisions of the Clean Water Act, 33 U.S.C. s1251 jrt
seq..as amended by the Water Quality Act of 1987, P.L 100-4, the Environmental
Protection Agency is hereby establishing a TMDL to limit discharges of dioxin to the
Columbia River basin.
This TMDL shall become effective immediately, and is incorporated into the
water quality management plans for the states of Washington, Oregon, and Idaho
under Clean Water Act s303(e). Subsequent state actions must be consistent with this
TMDL
Signed this Q day of , 1991.
QQlaMXjOL
Dana A. Rasmussen
Regional Administrator, Region 10
U.S. Environmental Protection Agency
-------
TOTAL MAXIMUM DAILY LOAD (TMDL)
FOR 2,3,7,8-TCDD
IN THE COLUMBIA RIVER BASIN
Developed pursuant to the provisions of the Clean Water Act, 33 U.S.C. s 1251, et seq.
Decision Document
February 25,1991
Developed pursuant to the provisions of the Clean Water ,
as amended by the Water Quality Act of 1987, P.L 100-4.
TABLE OF CONTENTS
Page
1. SCOPE 1-1
A. Water Quality-Limited Segments 1-1
B. Pollutant Causing Exceedance of WQ Standards 1-1
C. Source Categories Considered 1-2
2. NEED FOR A TMDL 2-1
A. Overview 2-1
B. The Concern 2-2
D. Water Quality Limited Status 2-3
3. DEVELOPMENT OF THE TMDL 3-1
A. Overview 3-1
B. Process 3-1
C. Loading Capacity 3-2
D. Sources 3-2
E. Allocation of Loads 3-4
F. Judicial Review 3-11
4. SUMMARY 4-1
APPENDICES
A. LOADING CAPACITY A-1
B. ALLOCATION ISSUES B-1
C. WASTE LOAD ALLOCATION METHODS CONSIDERED C-1
-------
Final TMDL for Dlaxin Discharges to the Columbia Basin 1-1
TOTAL MAXIMUM DAILY LOAD
FOR
2,3,7,8-TCDD IN THE COLUMBIA RIVER BASIN
Decision Document
1. SCOPE
This total maximum daily load (TMDL) addresses the following segments,
pollutants, and source categories:
WATER QUALITY-LIMITED SEGMENTS:
RIVER SEGMENT APPLICABLE WATER QUALITY RULES:1
Columbia River
(RM
0
745)
WAC 173-201-047 2
n «
(RM
0
309)
WAC 173-201-080(19) 3
tl R
(RM
309
596)
WAC 173-201-080(20) s
n m
(RM
596
745)
WAC 173-201-080(21) 3
n «
(RM
0
86)
OAR 340-41-202 & 205(2)(p)
4,5
n n
(RM
86
120)
OAR 340-41-442 & 445(2)(p)
4.5
n N
(RM
120
147)
OAR 340-41-482 & 485(2)(p)
4,5
ft if
(RM
147
203)
OAR 340-41-522 & 525(2)(p)
4,5
n n
(RM
203
218)
OAR 340-41-562 & 565(2)(p)
4,5
n n
(RM
218
247)
OAR 340-41-602 & 605(2)(p)
4,5
n w
(RM
247
- 309)
OAR 340-41-642 & 645(2)(p)
4,5
Snake River
(RM
0
- 176)
WAC 173-201-047 2
H II
(RM
0
- 176)
WAC 173-201-080(97) 3
6,7
n ti
IDAPA 16.01.2120 & .2200
Willamette River (RM 0
- 187)
OAR 340-41-442 & 445(2)(p)
4.5
1 In addition to the following, all waste load allocations and permit limits must ensure
compliance with applicable water quality standards of downstream states [40 CFR
§ 122.4(d)].
2 WAC 173-201-047 describes Washington's applicable criteria for toxic substances.
9 WAC 173-201-080 describes Washington's classification for specific waterbodies.
4 OAR 340-41-xx2 describes beneficial uses designated by Oregon.
* OAR 340-41-xx5(2) (p) describes Oregon's applicable criteria for toxic substances.
8 IDAPA 16.01.2120 describes the designated uses for the confluence of the Clearwater
and Snake River in Idaho.
7 IDAPA 16.01.2200 describes Idaho's criteria for hazardous and deleterious materials.
POLLUTANT CAUSING EXCEEDANCE OF WQ STANDARDS:
2,3,7,8 - tetrachlorodibenzo-para-dioxin (2,3,7,8-TCDD)
-------
Final TMDL for Dicodn Discharges to the Columbia Basin
1-2
SOURCE CATEGORIES CONSIDERED:
Source
Category
1
2
Allocation
Type
WLA1
Reserved
Source Description
Pulp & Paper Mills - Chlorine Bleaching
All Other Sources:
¦ Pulp & Paper Mills - Non-Chlorine Bleaching
¦ Woodtreaters Using Pentachlorophenol
¦ Municipal Wastewater Treatment Facilities
¦ Canadian Sources
¦ Other Point Sources
¦ Port Activities
¦ Urban Areas
¦ Other Nonpoint Source
¦ Background
Figure 1-1. Columbia River Basin.
CELCAR
(Casiiefar. ELC)
BOISE CASCADE
(Vftlluia. VA)
LONG VIEW FIBER.
(Loo|vi*v VA)
WEYERHAEUSER.
(Longrtcv. va)
JAMES RIVER.
(Vauu, (A)
BOISE CASCADE.
(St Helena, OK)
POPE * TALBOT"
(H&tery. OR)
rmoT*
POTtATCH CORPORATION
(Uvlaba, ID)
R%*
1 WLA = waste load allocation
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
2-1
2. NEED FOR A TMDL
A. Overview
The Columbia River and segments of the Snake and Willamette Rivers are
currently water quality-limited due to the presence of excessive levels of 2,3,7,8-TCDD.
This pollutant is the most toxic of a group of compounds known as polychlorinated
dibenzo-para-dioxins (dioxin). The concern over dioxin levels in the Columbia River is
based on data describing concentrations of 2,3,7,8-TCDD in effluents and treatment
piant sludges at chlorine-bleaching pulp mills as well as in fish tissue below these mills.
Section 303(d)(1)(C) of the Clean Water Act (CWA) and EPA's implementing
regulations (40 CFR Part 130) require each state to identify waters for which existing
required pollution controls are not stringent enough to attain applicable water quality
standards. For these water quality-limited segments, each state is then to establish
total maximum daily loads (TMDLs) for appropriate pollutants of concern. By definition
(40 CFR, s 130.2), a TMDL is the sum of the individual waste load allocations (WLAs)
for point sources and load allocations (LAs) for nonpoint sources and natural
background. The CWA states that the TMDL:
"shall be established at a level necessary to implement the applicable
water quality standards with seasonal variations and a margin of safety
which takes into account any lack of knowledge concerning the
relationship between effluent limitations and water quality."
Thus, the TMDL is effectively an implementation plan for achieving water quality
standards using an appropriate margin of safety. A margin of safety may be provided
(1) by using conservative assumptions in the calculation of the loading capacity of the
waterbody and (2) by establishing allocations that in total are lower than the defined
loading capacity. The water quality standard being protected by this TMDL is
0.013 parts per quadrillion (ppq) 2,3,7,8-TCDD in the water (see Appendix A).
The national focus on toxics discharges as evidenced in the 1987 amendment
to Section 304 of the CWA, 33 U.S.C. s 1314(1), gives additional urgency to the
establishment this TMDL Congress intended s304(1) to focus state water quality
protection programs on immediately addressing water quality problems due to point
source discharges of toxic pollutants. States are required to develop lists of impaired
waters, identify point sources and amounts of toxic pollutants they discharge,and to
develop individual control strategies (ICSs) for each such point source. An ICS may
be a draft or a final National Pollutant Discharge Elimination System (NPDES) permit.
The s304(1) lists developed for Washington, Oregon, and Idaho have identified dioxin
levels in the Columbia, Snake, and Willamette Rivers as exceeding applicable water
quality standards. Limits included in ICS's, developed under s304(l), must be
consistent with waste load allocations (WLAs) where a TMDL has been established.
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
2-2
B. The Concern
Dioxins are produced as a result of human activities, such as the manufacture
of chlorinated herbicides, the combustion of domestic and industrial wastes, and the
production of chlorine-bleached wood pulp. Both water column concentrations of
dioxin in the Columbia River and the water quality standard for 2,3,7,8-TCDD are below
levels which can be measured with current analytical technology. However, because
some organisms, such as fish, accumulate dioxin in their bodies, 2,3,7,8-TCDD has
been found at detectable levels in the tissue of fish taken from the Columbia River
basin. As discussed below, these tissue levels are of concern and indicate that these
waters exceed state water quality standards.
The state water quality standard applicable to 2,3,7,8-TCDD in the Columbia
River basin has been determined to be 0.013 ppq (see Appendix A). The EPA criterion
on which this standard is based was derived from human health concerns resulting
primarily from consumption of contaminated fish. In establishing EPA's 1984 2,3,7,8-
TCDD criterion values, the following factors were developed and used: a biocon-
centration factor (this relates the concentration in fish tissue to the concentration in the
water in which the fish lives), fish consumption rates, and a cancer potency factor.
These factors relate water column concentrations to fish tissue concentration and
cancer risk. A fish tissue concentration of 0.07 ppt and a water concentration of 0.013
ppq (the applicable water quality standard) are both estimated to result in a life-time
cancer risk of 10* (one excess cancer per one million people).
In 1987, EPA initiated a National Bioaccumulation Study (NBS) designed to
gather screening information on the prevalence and concentrations of selected toxic
compounds in fish tissue and other aquatic organisms. This study was conducted on
a broad scale across the United States and included testing for 2,3,7,8-TCDD.
Sampling sites included relatively undisturbed background areas, streams below
industrial, agricultural, and urban activities, and segments below mills using chlorine to
bleach pulp. The NBS identified concerns related to chlorine-bleaching kraft pulp mills.
Fish samples collected at several locations below chlorine-bleaching pulp mills on the
Columbia River within EPA Region 10 (from the Canadian border to the mouth) have
shown detectable concentrations of 2,3,7,8-TCDD. Another EPA study, the "104 Mill
Study" (1988), subsequently confirmed, through testing of effluents and sludges, that
chlorine-bleaching pulp mills are a significant source of 2,3,7,8-TCDD.
Figure 2-1 displays estimates of risk of excess cancer resulting from
consumption of fish at various locations along the length of the river. The risk
estimates were obtained by applying the fish consumption and cancer potency factors
used in developing the EPA criterion for 2,3,7,8-TCDD to fish tissue concentrations
actually measured. Fish tissue data used came from EPA's National Bioaccumulation
Study (1987), the Northwest Pulp & Paper Associations^ Columbia River Fish Study
(Beak Consultants, 1989), the Washington Department of Ecology's work on Lake
Roosevelt (1989-1990), and from efforts in Canada. The resulting risk estimates
(Figure 2-1) are consistently higher than the 10T8 level, confirming that the water quality
standard and, therefore, the loading capacity of the system, are being exceeded. This
is consistent with, and supported by, predicted water column concentrations of
2,3,7,8-TCDD (based on in-stream dilution of pulp mill discharges as measured in the
104 Mill Study) which also exceed the water quality standard.
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
2-3
Figure 2-1. Columbia River Fish Tissue Data
.y
in
Bosi s for Load i ng Capac i ty
200
BOO
River Mile
Fish tissue TCDD
C. Water Quality-limited Status
Oregon has identified the Columbia River (river miles 0 - 309) and the Willamette
River (RM 0-187) as being water quality-limited for 2,3,7,8-TCDD. Washington has
similarly identified the Columbia and Snake Rivers within that state as being water
quality-limited for 2,3,7,8-TCDD. The state of Idaho has also identified the confluence
of the Clearwater and Snake Rivers as being water quality-limited for 2,3,7,8-TCDD.
On June 14,1990, EPA approved these listings pursuant to CWA Section 303(d).
On March 21,1990 the states of Oregon, Washington, and Idaho stated that
they would not adopt a TMDL for dioxin in the Columbia River as state actions but
rather requested that EPA establish this TMDL as a federal action. The states
acknowledged that while the development of a TMDL has been a cooperative effort,
the interstate nature of the Columbia River Basin and the desirability of consistency
and equity in regulating dischargers in this basin necessitated that the TMDL be a
federal action. Therefore, on June 14,1990, pursuant to Section 303(d), EPA formally
disapproved the expressed intent of Washington, Oregon, and Idaho to not submit
TMDLs and, subsequently, developed this final TMDL for dioxin discharges to the
Columbia River basin as a federal action.
This TMDL provides a framework to control 2,3,7,8-TCDD discharges to the
Columbia River Basin and achieve compliance with water quality standards. The
following sections of the decision document describe the established TMDL and the
process used to develop it.
-------
Final TMDL for Dicodn Discharges to the Columbia Basin
3-1
3. DEVELOPMENT OF THE TMDL
A. Overview
Development of a TMDL provides a process for weighing the needs of
competing activities which affect water quality in a watershed and creating an
integrated pollution control strategy for point and nonpoint sources. This process
allows regulatory agencies to take a holistic view of water quality problems from the
perspective of in-stream conditions.
The total load of a pollutant to a waterbody is attributable to point sources,
nonpoint sources, and natural background. The TMDL process distributes portions of
the stream's loading capacity to the various sources, including background conditions,
in a way that will achieve water quality standards. The level of refinement reflected in
actual allocations depends on the amount of available data. The Water Quality
Management Regulations [40 CFR, s 130.2] state, for example, that:
"Load allocations are best estimates of the loading, which may range
from reasonably accurate estimates to gross allotments, depending on
the availability of data and appropriate techniques for predicting the
loading."
As previously pointed out, Section 303(d) states that a margin of safety should
be used which takes into account any lack of knowledge concerning the relationship
between effluent limitations and water quality. Thus, the law indicates that the TMDL
process should move forward using available information. As new information
becomes available in the future, the TMDL can be refined.
B. Process
The TMDL identifies the amount of a pollutant that may be discharged to a
water quality-limited stream. TMDLs can be expressed in terms of either chemical
mass per time, toxicity, or other appropriate measure. The TMDL for a particular
waterbody is dependent on such factors as the location of sources, stream flow, water
quality standards, background conditions, and in-stream pollutant reactions. The
process of developing and implementing a TMDL for 2,3,7,8-TCDD in the Columbia
River basin consists of several steps:
¦ define the loading capacity of the river at key points
¦ Identify sources which potentially contribute loads of 2,3,7,8-TCDD
¦ allocate loads to point sources, nonpoint sources (NPS), and background
¦ implement the TMDL through Water Quality Management Plans and NPDES
permits
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
3-2
C. Loading Capacity
WLAs and LAs represent the allocated portions of a receiving water's loading
capacity. The loading capacity is the greatest amount of pollutant loading that the
river can receive without violating water quality standards. A TMDL must not exceed
the loading capacity of a waterbody.
Two fundamental issues must be determined at the outset when establishing a
TMDL These are (1) the definition of upstream and downstream boundaries of the
waterbody for which the TMDL is being determined and (2) the flow conditions
(design flow) appropriate for calculating the loading capacity or amount of pollutant
which can be assimilated. Having defined the extent of the waterbody and the
appropriate flow conditions, the loading capacity is calculated to achieve the applicable
water quality standard (see Appendix A for discussion of applicable standards for
dioxin and river flow rates occurring in the Columbia River Basin).
A loading capacity of approximately 6 mg of 2,3,7,8-TCDD per day has been
calculated for the Columbia River at its mouth.
D. Sources
The Columbia River is over 1200 miles long and drains an area of about
259,000 square miles. Land use and terrain in the basin are diverse. General activities
affecting water quality in the basin include areas of urban development, industry,
agriculture, and forestry. In terms of 2,3,7,8-TCDD, chlorine bleaching pulp mills have
been identified as a major source based on their effluent and sludge data.
Within EPA Region 10, eight chlorine-bleaching pulp mills currently discharge to
the Columbia River system. These mills, one in Idaho, four in Washington, and three
in Oregon, are shown in Figure 3-1. The eight mills currently produce over 7,000 tons
per day of bleached pulp. Another chlorine-bleaching pulp mill which discharges to
the Columbia River is located near Castlegar, British Columbia, about 30 miles above
the U.S. - Canadian border. Known sources of 2,3,7,8-TCDD are thus affecting the
Columbia River within EPA Region 10, from the mouth near Astoria, Oregon to the
Canadian border (river mile 745) and the Snake and Willamette Rivers, major drainages
within the Columbia River system. Consequently, the entire Columbia River basin,
including the Snake and Willamette Rivers, are included in the TMDL Tributaries
outside of EPA Region 10, such as the Clark Fork in Montana, have also been
considered in developing the TMDL
Besides chlorine bleaching pulp mills, other potential source categories include
woodtreaters using pentachlorophenol, major municipal wastewater treatment plants,
agricultural areas, industrial sites, urban areas, and release from bottom sediments.
Data on dioxin discharges from these sources, however, are minimal or nonexistent for
the following reasons:
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
3-3
m Concern over the extent of dioxin pollution is relatively recent.
¦ Many of the point sources have been considered minor dischargers in the
past and had minimal monitoring requirements.
¦ It is difficult to characterize loadings from stormwater or nonpoint sources.
These inputs are not continuous and are generally driven by weather related
events such as rain storms or snow melt.
¦ There are analytical obstacles associated with measuring 2,3,7,8-TCDD. The
water quality standard of 0.013 parts per quadrillion (ppq) is several orders
of magnitude below a typical detection limit of 10 ppq for water column
measurements.
The available data are not adequate to develop WLAs or LAs for these sources.
However, current loadings for some of these other dioxin sources of concern in the
Columbia basin are estimated in Appendix B and summarized later in the following
section.
Figure 3-1. Location of Chlorine-Bleaching Pulp Mills
in the Columbia River Basin
CELGAR
(CwUegar, B£)
BOISE CASCADE
(VaUuU)
POTLATCH CORPORATION
(LbvMos)
JAMES Rival
LONGVIEW FIBER
/ (Lom*to»)
Jakes rives
/ (Cunu)
WftUM
BOISE CASCADE
(SL Bctena)
Portland
POPE ft TALBOT
(Hftbey)
-------
1
Final TMDL for Dlodn Discharges to the Columbia Basin 3-4
E. Allocation of Loads
Having identified major sources of 2,3,7,8-TCDD to the Columbia River basin,
the TMDL must establish allocations sufficient to control discharges within the loading
capacity. These allocations are made considering technical, socioeconomic, and
institutional constraints. Historically, individual states have used various allocation
schemes on a case-by-case basis or specified that a particular method be used.
Technical guidance has been prepared which describes 19 potential approaches for
allocation of loads (Technical Guidance Manual for Performing Waste Load
Allocations", U.S. Environmental Protection Agency, 1986). When evaluating various
methods, conditions that favor one approach over another must be considered.
With respect to this TMDL there are some potential problems in using the more
common methods described in the technical guidance:
¦ The geographic scale associated with the Columbia Basin and the number of
potential sources is considerably larger than the scale typically encountered
in most TMDL situations.
¦ Common methods focus on waste load allocations for point sources.
Background sources (e.g. release from bottom sediments) and nonpoint
source loads, however, may be significant considerations for 2,3,7,8-TCDD
in the Columbia River basin.
¦ There are few data on 2,3,7,8-TCDD discharges from source categories
other than chlorine bleaching pulp mills in the basin.
¦ There are complexities in addressing persistent and highly bioaccumulative
pollutants such as 2,3,7,8-TCDD.
The last three of these points mean that data and methods of analysis (e.g. predictive
models) are not available to adequately characterize all pollution sources at this time.
However, the lack of information about some pollution sources or processes is not a
reason to delay implementation of water quality-based controls for known sources
contributing to violations of water quality standards. The key is to work within a logical
framework that will lead to the attainment of water quality standards. After
consideration of the above problems and the issues discussed in Appendix B, the
following approach was developed for this TMDL:
¦ Identify watershed targets to be used as a framework to guide allocation
decisions;
¦ Establish WLAs for the major source category for which there are currently
sufficient data to do so;
¦ Estimate current loadings for other source categories;
-------
Final TMDL for Dloxin Discharges to the Columbia BasJn
3-5
m Reserve some of the unallocated loading capacity (beyond that necessary to
cover the WLAs established and estimated current loadings for other
sources) to provide an additional component of the margin of safety, some
of which could be used for future growth.
This approach provides for further pollution reduction from known sources while
additional data are collected to: (1) confirm that the reductions required by this TMDL
are leading to water quality standards attainment; and (2) provide additional
information necessary to refine estimates of assimilative capacities and TMDL
allocations. This TMDL establishes WLAs that will form the basis of more stringent
limits for dioxin discharges from confirmed point sources. It also estimates loadings
from other sources and incorporates a margin of safety to account for existing
uncertainties. Where new data show that modification of the TMDL is appropriate, the
TMDL will be revised accordingly. By allowing future modification of the TMDL,
regulatory agencies can avoid delays in controlling known sources while they continue
to investigate other possible sources. Decisions on the use of the unallocated load will
be made through a joint effort by the States and EPA.
Watershed Targets:
The Oregon Department of Environmental Quality (DEQ) has utilized the
concept of watershed targets for developing TMDLs in Oregon. Watershed targets are
particularly useful for TMDLs designed to achieve water quality standards in large
waterbodies adversely affected by a pollutant coming from a variety of sources.
Allocations for major sources are established after watershed targets are identified.
The watershed targets serve as internal check points to determine that water quality
standards will be met at key locations within the drainage. This same technique is also
being used for the Columbia River in this TMDL
Watershed targets can be set within the basin by simply identifying the loading
capacity at key points in the drainage system. To determine these targets, the only
data requirements are a water quality criterion and a design flow (in this case, the
mean harmonic stream flow). The watershed targets focus on high priority tributaries.
In the case of the Columbia, there are three logical points in addition to the lower
Columbia near Bradwood (below Longview) for which loading capacities should be
calculated. These locations are shown in Figure 3-2 and relevant data are summarized
in Table 3-1.
The Willamette Basin is the most industrialized and populated area in the
Columbia River system. There are high numbers of both industrial and municipal
dischargers in the drainage compared to other sub-basins in the Columbia River
system. The most logical approach is to establish the watershed target as equal to the
loading capacity for the Willamette River at Portland (0.54 mg/day). The sum of all
allocations to sources in the Willamette Basin must not exceed this watershed target.
By the same token, loading capacity attributed to flow produced by the Willamette is
not currently available for use in the mainstem Columbia.
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
3-6
Because the Willamette Basin is entirely within Oregon, the Oregon Department
of Environmental Quality (ODEQ) has the option, within the context of a TMDL, to
adjust allocations for specific sources which would still meet this watershed target. In
fact, Oregon has already initiated dioxin controls in the Willamette through issuance of
an NPDES permit to Pope & Talbot at Halsey with effluent limits for 2,3,7,8-TCDD
(0.19 mg/day). Furthermore, DEQ has committed to developing a TMDL for dioxin in
the Willamette which will meet the watershed target1 A Willamette Basin TMDL could
include different limits for Pope & Talbot, based on needs determined by ODEQ.
Figure 3-2. Location of Watershed Targets (¦) Relative to Pulp Mills
Watershed targets were also evaluated at two other locations in the Columbia
system: 1) at the mouth of the Snake River and 2) at the U.S. - Canadian border.
Far fewer sources exist upstream of these locations than is the case with the
Willamette River basin. However, significant levels of 2,3,7,8-TCDD have been
measured in tissue of fish taken from sites associated with each of these watersheds.
The fish tissue concentrations indicate that the water quality standard and, therefore,
the loading capacity for 2,3,7,8-TCDD is currently exceeded.
1 This TMDL will be reviewed by EPA in accordance with 5303(d) of the Clean Water Act.
(VaOola)
POTLATCH CORPORATION
(U*M«d)
JAMES RIVER,
(liuu)
Portland
-------
Final TMDL for Dlasdn Discharges to the Columbia Basin
3-7
Based on currently available data, reductions in 2,3,7,8-TCDD loads are needed
to meet all three of these watershed targets. These watershed targets must be
achieved in order to ensure attainment of water quality standards where those
watersheds enter the Columbia River. To the extent that the TMDL results in loading
reductions beyond that necessary to meet the watershed targets, the difference is
available for other downstream uses, future growth, or margin of safety.
Table 3-1. Loading Targets for 2,3,7,8-TCDD to Selected
Watersheds in the Columbia River System
Harmonic
Loading
Watershed
Mean Flow
Capacity
(cfs)
(mg/day)
TOTAL COLUMBIA RIVER BASIN
188,000 1
5.97
SELECTED SUB-BASINS
Watershed N. of WA/Canada Border
72;700 2
2.31
Snake River Watershed
37,0003
1.18
Willamette River Watershed
17,100 4
0.54
TOTAL FOR SUB-BASINS
4.03
1 Row at Columbia River near Bradwood
2 Row at Columbia River at WA/Canada border
3 Row at Snake River below Ice Harbor Dam
4 Row of Willamette River at Portland
Establish WLAs
This TMDL focuses on developing waste load allocations for the chlorine
bleaching pulp mills in the basin. These mills constitute the only source category in
the Columbia River basin where site specific quantitative information exists describing
effluent quality and waste loads for 2,3,7,8-TCDD. Nationally, the median 2,3,7,8-
TCDD concentration in tissue of fish collected below pulp mills using chlorine bleaching
was higher than for fish collected below any other source category studied in the
National Bioaccumulation Study (1987). In additioh, the s304(1) listings under the
Clean Water Act specifically identified these mills in the Columbia River Basin as point
sources requiring individual control strategies (ICS's). The basis of this listing was not
only data describing concentrations of 2,3,7,8-TCDD in fish tissue below the mills but
also measured concentrations of 2,3,7,8-TCDD in effluents and treatment plant sludges
at these mills. The analysis undertaken in developing this TMDL indicates that this
source category would lead to exceedance of water quality standards even if no other
sources existed.
-------
Final TMDL for Dloadn Discharges to the Columbia Basin
3-8
The proposed TMDL (public notice issued on June 15,1990) discussed several
alternative methods to establish waste load allocations for chlorine bleaching pulp mills.
The waste load allocation methods evaluated are summarized in Appendix C. The
proposed TMDL allocated approximately 2 mg/day (not including the Canadian Celgar
mill or the planned expansion at Pope & Talbot) to the chlorine bleaching pulp mills. A
major criterion for evaluating alternative methods for establishing WLAs for chlorine
bleaching pulp mills was the need to verify compliance with resulting NPDES permits.
Allocations for each mill were derived based on the lowest verifiable concentration
(long term average of 4.7 ppq 2,3,7,8-TCDD in the bleached wastestream) in an
assumed average wastewater flow per quantity bleached pulp produced (14,470
gallons/ton). Such an approach yields WLAs which are equal in terms of mass
discharge per unit production of bleached pulp product (0.257 »g 2,3,7,8-TCDD/ton).
Table 3-2 displays WLAs based on updated production figures including
planned production increases for Celgar [based on comments from R.W. Sweeney,
Celgar Pulp Co.] and Pope & Talbot [based On comments from CH2M-Hill for James
River and Pope & Talbot; July 20,1990]. WLAs resulting from allowing 4 different
quantities of 2,3,7,8-TCDD per ton of bleached pulp produced are given in the table.
Three of the options reflect some of the comments received during the public
comment period for the proposed TMDL
Option 1. This option reflects the belief by the pulp and paper industry that
they should be given the entire loading capacity of the river system.
An allowed discharge rate of 0.68 ng 2,3,7,8-TCDD per ton of
bleached product results in 100% of the calculated loading capacity
being allocated to the existing pulp and paper mills in the basin.
Option 2. This option is generally equivalent to the WLAs proposed in the
draft TMDL submitted for public comment. Two differences are
noted: (1) the WLA for Pope & Talbot at Halsey is increased based
on planned production increases and the NPDES permit recently
issued by DEQ; and (2) a WLA has been calculated for the Celgar
mill based on planned production increases and the discharge rate
(0.257 pQ 2,3,7,8-TCDD per ton of bleached product) allowed for
the other mills. The calculated WLA for Celgar has no regulatory
authority, but is used for comparison purposes and as an estimated
loading which should be achievable by Celgar.
Option 3. This option reflects the concern by the local pulp mills that the
proposed TMDL did not provide equity with the Celgar mill at
Castlegar, British Columbia. Based on information submitted by
both the Celgar mill and the British Columbia Ministry of
Environment (see Appendix B), the proposed modernization project
at Celgar will result in 2,3,7,8-TCDD discharges which are less than
0.05 mg/day (or 0.042 /xg/day per ton bleached pulp). The
technology planned for use at Celgar is being or has been installed
at several bleached kraft mills in other parts of the world. Option 3
applies this discharge rate to all the affected mills and results in 7%
of the calculated loading capacity being allocated to the existing
pulp and paper mills in the basin.
-------
Final TMDL for Dloxin Discharges to the Columbia BasJn
3-9
Option 4. This is the zero discharge option requested by many commenters.
The environmental community believes that zero discharge is the
only viable option, because of dioxin's persistence and cumulative
build-up in the sediments and biota.
Table 3-2. Waste Load Allocation Options for Chorine-Bleaching Pulp Mills
Pulp Mill *• Location
Production of
Waste Load Allocations
(mg 2,3,7,8-TCDD/day, long term average)
Option 1
(0.68)
Option 2
(0.257)
Option 3
(0.042)
Option 4
(0.00)
(tons/day)
(X)
Potlatch -- Lewiston, ID
1,509
15.1
1.03
0.39
0.06
0.00
Boise Cascade — Wallula, IM
James River -- Camas, UA
Longview Fibre — Longview, UA
Weyerhaeuser -- Longview, UA
957
1,650
310
1,026
9.6
16.5
3.1
10.3
0.65
1.12
0.21
0.70
0.25
0.42
0.08
0.26
0.04
0.07
0.01
0.04
0.00
0.00
0.00
0.00
Pope & Talbot -- Halsey, OR
Boise Cascade -- St. Helens, OR
James River -- Uauna, OR
1,500
1,035
800
15.0
10.4
8.0
0.19
0.70
0.54
0.19
0.27
0.21
0.06
0.04
0.03
0.00
0.00
0.00
Celgar -- Castlegar, B.C.
1,200
12.0
0.82
0.31
0.05
0.00
TOTAL Source Category Allotment
9,987
100.0
5.96
2.38
0.40
0.00
X of Basin Loading Capacity
100X
40X
7X
ox
Mote: a) The value shown parenthetically infer each option represents the equivalent quantity of
2,3,7,8-TCDD discharged in jig per ton of bleached pulp produced.
b) The ULA listed for Pope & Talbot infer Options 1 and 2 has been adjusted to the long term
average of 0.19 mg/day identified in the NPDES permit issued by the Oregon Department of
Environmental Quality (Novetrtoer 7, 1990). See discussion in "Watershed Targets" section.
c) The WLAs listed for Celgar are included for comparison purposes only. EPA has no authority to
establish enforceable WLAs for a Canadian source.
All available information has been carefully considered. Based on that
information the "zero discharge" option is not necessary to achieve water quality
standards and would not be enforceable due to the fact that the analytical detection
limit is significantly higher than zero. Option 3 has similar difficulties, especially with
respect to measuring compliance. This leaves Options 1 and 2 as still reasonable.
The existence of other sources (see below), the lack of information on processes
affecting the distribution of 2,3,7,8-TCDD, and the ,concern over the potential release
from 2,3,7,8-TCDD stored in sediments and aquatic biota make Option 1 inappropriate.
Consequently, Option 2 is the most reasonable approach at this time and the WLAs
listed under that option are being established as final in this TMDL. EPA has
concluded that these WLAs are the lowest levels consonant with analytical practicalities
at this time and, as discussed below, can be accommodated within the available
loading capacity taking into account other existing sources. NPDES permits issued
subsequent to this TMDL must be consistent with these waste load allocations.
-------
Final TMDL for Dioxfn Discharges to the Columbia Basin
3-10
EPA recognizes that, as NPDES permits are developed, some adjustment of the
above WLAs to reflect differences in particular mill capabilities may be appropriate.
Such adjustments, if needed, will be determined on a case-by-case basis in
consultation with the affected states.
Estimated Loadings From Other Sources
There is insufficient information, at this time, to establish WLAs for other point
sources or LAs for nonpoint sources. However, in order to be reasonably certain that
total loadings under this TMDL will not exceed the loading capacity of the system,
loadings from some of the most significant other source categories are evaluated in
Appendix B and summarized below.
Canada:
The Celgar pulp mill is the only Canadian source of dioxin to the Columbia River
for which 2,3,7,8-TCDD has been measured in the effluent. As pointed out in the
previous section, however, EPA has no authority to establish an enforceable WLA for
the Celgar pulp mill in Canada. In this TMDL, EPA estimates that 2,3,7,8-TCDD
loadings from sources upstream of the U.S.-Canada border will be no more than the
0.31 mg/day which we would allocate to Celgar if it were a Region 10 mill (Table 3-2,
Option 2). Since Celgar is expected to reduce its 2,3,7,8-TCDD loadings to
0.05 mg/day by 1994, the higher 0.31 mg/day estimate provides some room to cover
other unidentified sources upstream of the U.S.-Canada border and/or a margin of
safety for the possibility that Celgar may not fully achieve anticipated reductions in its
2,3,7,8-TCDD loading to the Columbia River.
Other U.S. Point Sources:
As detailed in Appendix B, woodtreating facilities and municipal wastewater
treatment plants are estimated, in total, to contribute current loadings of less than
2.3 mg/day 2,3,7,8-TCDD. Establishing WLAs for these facilities is not feasible at this
time due to the shortage of data. Recent Resource Conservation and Recovery Act
(RCRA) regulations for woodtreaters and NPDES regulations and guidance for
stormwater discharges will lead to better information and control of discharges from
these sources in the future. WLAs will be established, if appropriate, for those point
source discharges with existing NPDES permits when information becomes available.
Other Sources and Background:
The remaining 22% of the loading capacity (1.29 mg/day) will be held in reserve
as part of the needed margin of safety. This will cover contributions from (1) nonpoint
sources such as agricultural or atmospheric inputs, (2) other industrial sources such
as non-chlorine bleaching pulp mills, (3) background levels of 2,3,7,8-TCDD stored in
the sediments and aquatic biota, and (4) possible future growth.
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
3-11
Data Collection
The establishment of this TMDL is not the conclusion of EPA's efforts with
respect to controlling dioxin in the Columbia River basin. A more comprehensive data
collection program is planned to confirm assumptions made in the development of this
TMDL Monitoring efforts will be designed to obtain better baseline information and to
fill recognized data gaps, particularly with respect to other potential sources of
2,3,7,8-TCDD and the role of sediments. If necessary, the TMDL will be revised based
on new information.
EPA will work cooperatively with the states to take the following actions:
¦ Develop a strategy to address water quality concerns related to
2,3,7,8-TCDD inputs from woodtreating facilities. The proposed strategy
should identify individual sources in each state to be considered for
allocations, a sampling plan for determining reductions needed, and a
schedule for implementation of the strategy. This should be done in
conjunction with activities required by NPDES regulations as implemented
under recent guidance for controlling stormwater discharges.
¦ Address other point source concerns, such as other major industrial NPDES
dischargers and major municipal NPDES facilities with formal pretreatment
programs, by States forwarding to EPA existing state data on
concentrations of dioxin in sludge.
¦ Develop a strategy that addresses the other source categories such as
urban runoff and agriculture.
F. Judicial Review
Parties seeking to challenge this TMDL are advised that exclusive review of this
TMDL might be in the United States Court of Appeals because arguments could be
made that this TMDL includes "effluent limitations" or is part of a determination as to a
State permit program, or is inextricably bound to the issuance or denial of NPDES
permits. If that is the case, any petition for such review would have to be filed within
120 days of EPA's action in establishing the TMDL, as described in 40 CFR
Section 23.2.
1 This information collection is exempt from the Paperwork Reduction Act because it is being
sought from fewer than 10 sources.
-------
Final TMDL for Dlaxin Discharges to the Columbia Basin
4-1
4. SUMMARY
Although certain types of data are currently lacking, available information
highlights several concerns. Concentrations of 2,3,7,8-TCDD in fish tissue in several
areas of the Columbia River basin exceed levels protective of human health at the 10T6
risk level and indicate that the state water quality standards are currently being
exceeded. Regional and national data strongly suggest that pulp mills which use
chlorine to bleach are the most significant sources of 2,3,7,8-TCDD to surface waters.
Direct measurements of effluent samples taken from chlorine-bleaching pulp mills in
the Columbia River basin confirms 2,3,7,8-TCDD levels requiring control.
There is a remaining need to refine information on contributions from other
potential sources such as woodtreaters, as well as to describe the effect of attenuation
and the role of sediments. This TMDL reserves a portion of the calculated loading
capacity as unallocated because of this need for information. The TMDL established
herein for 2,3,7,8-TCDD discharges to the Columbia River Basin completes the
following actions:
¦ Establishes waste load allocations to individual pulp mills which use chlorine
bleaching, at this time. Use equal mass discharge per unit production
(Table 3-2, Option 2) to allocate waste loads to individual pulp mills in that
source category. NPDES permit limits for these pulp mills must be consistent
with this TMDL
¦ Estimates loading from Columbia River sources upstream from the U.S.-
Canada border. The total loading reserved for this source category is
0.31 mg/day. By 1994 the Celgar pulp mill, is expected to reduce its
contribution to approximately 0.05 mg/day. The remainder of the 0.31 mg/day
is reserved as a margin of safety to cover other unidentified sources upstream
of the U.S.-Canada border and/or a shortfall by Celgar in achieving anticipated
reductions.
¦ Estimates loading from some Region 10 point sources other than the pulp mills
for which WLAs were established. Appendix B describes the evidence
suggesting a total 2,3,7,8-TCDD loading from these sources of less than
2.3 mg/day.
¦ Reserves the remaining loading capacity (1.29 mg/day, after subtracting the
WLAs and estimated loadings for the sources identified above) for (1) other
undesignated sources, (2) an additional margin of safety to account for
uncertainties in the assumptions used in developing this TMDL, and (3) future
growth. This reserved portion is equal to approximately 22% of the total loading
capacity. As uncertainties are reduced, more of the reserved capacity could be
allocated to new or existing sources.
Table 4-1 and Figure 4-1 summarize the overall structure of the Final TMDL with the
allocations based on currently available information.
-------
Final TMDL for Diaxln Discharges to the Columbia Basin
4-2
Table 4-1. Waste Load Allocations for Chlorine-Bleaching Pulp
Mills in Context of Watershed Targets
ULA
Loading
Capacity
LOADING CAPACITY FOR ENTIRE COLUMBIA RIVER BASIN
5.97
ColLMbia River Basin above Washington/Canada border
Watershed target
EstiMated Canadian Loading including Celgar mill
[0.31] 1
2.31
Snake River Basin above Ice Harbor Dam
Watershed target
Pulp Mill WLAs: Potlatch (Lewiston, ID)
0.39
1.18
Willamette River Basin above confluence with Columbia R.
Watershed target
Pulp Mill WLAs: Pope & Talbot (Halsey, OR)
0.19 2
0.54
Remainder of Coluifcia R. Basin
Pulp Mill WLAs: Boise Cascade (Wallula, WA)
James River (Camas, WA)
Longview Fibre (Longview, WA)
Weyerhaeuser (Longview, WA)
Boise Cascade (St. Helens, OR)
James River (Wauna, OR)
TOTAL
0.25
0.42
0.08
0.26
0.27
0J1
1.49
SUM OF WLAs FOR REGION X PULP MILLS IN BASIN
2.07
' This is not a WLA, but is included for purposes ot comparison with the WLAs for U.S. mills.
2 This Is the same WLA Identified in ODEQ's NPDES permit (issued 11/7/90) for this facility.
i
Figure 4-1. Overall Division of Columbia River Basin Loading Capacity
RESERVED (Unallocated)
60%
WASTE LOAD ALLOCATIONS
-------
Final TMDL for Dioodn Discharges to the Columbia Basin
A-1
APPENDIX A. LOADING CAPACITY
Waste load allocations (WLAs) for point sources and load allocations (LAs) for
nonpoint sources represent the allocated portions of a receiving water's loading
capacity. The loading capacity is the greatest amount of loading that the river can
receive without violating water quality standards. A TMDL must not exceed the loading
capacity of a waterbody. To determine the appropriate loading capacity available for
allocation requires:
¦ the water quality standard applicable to 2,3,7,8-TCDD and the Columbia River
basin.
¦ the river flows used to calculate the loading capacity of the Columbia River
basin at key locations.
1. Applicable Water Quality Standards
The pollutant of concern for this TMDL, 2,3,7,8-TCDD, is the most toxic of a
group of compounds known as polychlorinated dibenzo-para-dioxins. These
compounds are produced as a result of human activities such as the manufacture of
chlorinated herbicides, the combustion of domestic and industrial wastes, and the
production of chlorine-bleached pulp.
Oregon, Washington, and Idaho have adopted water quality standards for toxic
substances which apply to parts of the Columbia River basin including the Snake and
Willamette Rivers. Because the purpose of this TMDL is to provide a framework for
attaining all applicable water quality standards for dioxin, this multi-state TMDL must be
protective of the waters with the most stringent of those standards. A brief description
of individual state standards follows.
Oregon has adopted a numeric criterion for 2,3,7,8-TCDD. Oregon
Administrative Rules (OAR) Chapter 340, Division 41 summarizes water quality criteria
for toxic substances applicable to all basins. This includes the Columbia River from its
mouth to river mile 309 and the Willamette River from its mouth to river mile 187. OAR
340-41-205(p)(B), for example, states:
*Levels of toxic substances shall not exceed the most recent criteria
values for organic and inorganic pollutants established by EPA and
published in Quality Criteria for Water (1986). A list of the criteria is
presented in Table 20/
The ambient water concentration listed in Table 20 for protection of human health from
carcinogenic effects caused by 2,3,7,8-TCDD is 0.000013 ng/L, or 0.013 parts per
quadrillion (ppq). This value represents the 10"® risk level, the concentration at which a
lifetime exposure results in a probability of one excess cancer case per one million
people. It considers the consumption of contaminated water as well as fish or other
aquatic organisms.
-------
Final TMDL for Dladn Discharges to the Columbia Basin
A-2
Washington has identified the Columbia River from the mouth to river mile (RM)
596.6 as a Class A waterbody and from RM 596.6 to the Canadian border (RM 745) as
a Class AA waterbody. Washington has also identified the Snake River from the mouth
to RM 176.1 as a Class A waterbody. Washington's rules which apply to toxic
substances are found in WAC 173-201-047. The narrative part of the rule indicates
that:
"Toxic substances shall not be Introduced above natural background
levels In waters of the state which may adversely affect characteristic
water uses, cause acute or chronic conditions to the aquatic biota, or
adversely affect public health"
WAC 173-201-047 also states that appropriate concentrations for toxic substances in
Washington are to be determined in consideration with EPA's Quality Criteria for
Water (1986). in the process of developing its lists of degraded waters as required by
s304(1) of the Clean Water Act, Washington interpreted its standard for 2,3,7,8-TCDD in
a manner consistent with Oregon's numeric Standard, i.e. 0.013 ppq of 2,3,7,8-TCDD
as an ambient water concentration needed to protect human health.
Idaho has narrative standards which are intended to protect the beneficial uses
of its waters including the Snake River. The standard, found in IDAPA 16.01.2200,
states:
"As a result of man-caused point or nonpoint source discharge, waters of
the State must not contain: 01. Hazardous materials... in concentrations
found to be of public health significance or to adversely affect designated
or protected beneficial uses. 02. Deleterious materials... in
concentrations that impair designated or protected beneficial uses
without being hazardous."
In the process of developing Idaho's s304(1) short list, EPA interpreted this standard
also in a manner consistent with Oregon's numeric standard.
As stated above, this TMDL has been developed to achieve attainment of the
water quality standards of all affected states. Although the wording of the applicable
state standards for Idaho, Oregon, and Washington differs, EPA has interpreted these
standards as being equally stringent. Even if this is not the case, however,
2,3,7,8-TCDD loading to upstream segments still must be restricted to levels ensuring
the attainment of water quality standards applying to downstream segments.1 Where
this document refers to "the standard" or "the criterion" for 2,3,7,8-TCDD, this means
the 0.013 ppq criterion at the 10^ risk level and, by implication, the assumptions which
form the basis of that criterion as established by EPA. That criterion, adopted by the
State of Oregon, is the controlling water quality standard which this TMDL protects.
1 The Superior Court of Washington for Thurston County recently found that the manner in
which the State applied their water quality standards to the listing under §304(1) of three pulp
and paper mills was invalid. EPA believes that this decision does not affect the use of
0.013 ppq as the water quality standard for dioxin in developing this TMDL because all waste
load allocations and permit limits must ensure compliance with applicable water quality
standards of downstream states [40 CFR 5122.4(d)]. Oregon's water quality standard is
clearly stated as being 0.013 ppq for 2,3,7,8,-TCDD.
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
A-3
2. River Flow:
The loading capacity of a stream is determined using the water quality criteria
value and a design flow for the receiving water. Typically, loads are expressed as
chemical mass per time such as pounds per day. In the case of 2,3,7,8-TCDD, loads
have been expressed as milligrams (mg) per day and are calculated as follows:
Load (mg/day) = 0.00245 * Concentration (ppq) * Row (cfs)
The 0.00245 is the factor needed to convert the units of parts per quadrillion (ppq) and
cubic feet per second (cfs) to milligrams per day (mg/day)
The design flow significantly affects the determination of the loading capacity.
The choice of design flow used to calculate the loading capacity for the Columbia River
basin was based on the characteristics of the 2,3,7,8-TCDD water quality criterion.
That criterion, 0.013 ppq 2,3,7,8-TCDD, is based on human health concerns over a
lifetime. In order to address human health concerns, the harmonic mean flow is
recommended as the appropriate stream design flow (Draft Technical Support
Document for Water Quality-based Toxics Control, U.S. Environmental Protection
Agency, 1990).
The harmonic mean flow was used to develop this TMDL because it provides a
more reasonable estimate than the arithmetic mean to represent long-term average
river flow. Rood periods in naturally flowing rivers bias the arithmetic mean above
flows typically measured. This overstates available dilution. The calculation of the
harmonic mean, however, dampens the effect of peak flows. As a result, the bias is
reduced. The harmonic mean is also an appropriate conservative estimate of long-
term average flow in highly regulated river basins, such as the Columbia. In a
regulated river basin, the harmonic mean and the arithmetic average are often much
closer numerically.
Table A-1 summarizes the loading capacity for 2,3,7,8-TCDD in the Columbia
River system at several key locations. A long-term flow record must be used in order
to minimize the effect of either droughts or wet years. It is also important to recognize
the effect that reservoirs have had on flows in the Columbia basin. Many of the major
dams were constructed before 1950. Thus, flow records used to determine the
loading capacity in the Columbia River were those reported by the U.S. Geological
Survey from 1950 to present
Table A-1. Loading Capacity for 2,3,7,8-TCDD in the Columbia River
Drainage
Harmonic
Loading
Gage
Location
Area
Mean Flow
Capacity
(sq.mi.)
(cfs)
(mg/day)
12399500
Coluifcia River at International Boundary
59,700
72,700
2.31
12472800
Coluifcia River below Priest Rapids
96.000
95,100
3.03
14019200
Coluifcia River at McNary Dam
214,000
143,000
4.54
14105700
Coluifcia River at The Dalles
237.000
152,000
4.83
14144700
Coluifcia River at Vancouver
241,000
159,000
5.04
14222880
Coluifcia River at Coluifcia City
254,000
180,000
5.73
14246900
Coluifcia River below Longview
256.900
188,000
5.97
-------
Final TMDL tor Dloxtn Discharges to the Columbia Basin
A-4
Rows at three locations on the Columbia River were estimated because of
inadequate long-term records. These locations are at Vancouver (gage #14144700),
at Columbia City (gage #14222880), and below Longview (gage #14246900). The
estimates were based on gaged flows from tributary rivers for the corresponding
segments. Average flow yield from the tributaries for a particular segment was used to
estimate flow from the ungaged portion of that segment. These gaged tributaries are
listed in Table A-2.
Table A-2. Loading Capacity for 2,3,7,8-TCDD in the Columbia River Tributaries
Drainage
Harmonic
Loading
Cage
Location
Area
Mean Flow
Capacity
(sq.mi.)
Ccfs)
(mg/day)
13343500
Snake River near Clarkston
103,200
35,700
1.14
13353000
Snake River below Ice Harbor Dan
108,500
37,000
1.18
14113000
Klickitat River near Pitt
1,297
1,207
0.04
14120000
Hood River near Hood River
279
612
0.02
14123500
White Salmon River near Underwood
386
951
0.03
14125500
Little White Salmon River near Cook
134
317
0.01
14128500
Wind River near Carson
225
514
0.02
14142500
Sandy River below Bull Run River
436
1,009
0.03
14143500
Washougal River near Uashougal
108
234
0.01
14166000
Willamette River at Harrtsburg
3,420
7,600
0.24
14211720
Willamette River at Portland
11,100
17,100
0.54
14220500
Lewis River near Ariel
731
2,396
0.08
14222500
East Fork Lewis River near Heisson
125
196
0.01
14223500
Kalama River near Kalama
198
618
0.02
14243000
Cowlitz River at Castle Rock
2,238
5,721
0.18
-------
Final TMDL for Dlosdn Discharges to the Columbia Basin
B-1
APPENDIX B. ALLOCATION ISSUES
In determining appropriate allocation methods for the Columbia, several
concerns have been identified that affect decisions on the TMDL. Issues identified
which were considered in developing allocations for 2,3,7,8-TCDD to the Columbia
River include:
¦ Loading from the British Columbia pulp mill
¦ Loading from other potential sources such as woodtreaters
¦ Fate, transport, and attenuation
¦ Role of bottom sediments (cumulative effects and resuspension)
¦ Framework for addressing future allocations (both growth within the
pulp industry and allocations to other source categories)
1. British Columbia Pulp Mill
Celgar Pulp Company operates a bleached kraft pulp mill located in Castlegar,
British Columbia. Wastewater from this mill is discharged to the Columbia River
approximately 30 miles upstream from the United States - Canada border (Figure B-1).
Studies conducted by Canadians have shown elevated concentrations of 2,3,7,8-TCDD
in lake whitefish collected below the Celgar mill (Mah et a/., 1989; EVS, 1990). In
addition, follow-up analyses by the Washington Department of Ecology of fish from
Lake Roosevelt found elevated levels of TCDD and TCDF (Johnson, 1990). Lake
Roosevelt is the impoundment formed by Grand Coulee Dam on the Columbia River
downstream from the Celgar mill. Because of concern over the amounts of TCDD and
TCDF detected in fish tissue, the Washington Department of Health took action in
August 1990. A health advisory was issued that children under age four and under 40
pounds should not eat whitefish from Lake Roosevelt. Subsequent sampling by the
Washington Department of Ecology suggests that concentrations of 2,3,7,8-TCDD may
also be elevated in sturgeon as well.
Figure B-1. Location of Celgar Pulp Co. (Castlegar, B.C.)
BOISE CASCADE
(Vallala)
POTLATCH CORPORATION
(LevMoa)
CELCAR
(Costle(ar. H£)
-------
Final TMDL for Dloxin Discharges to the Columbia Basin
B-2
The discovery of elevated levels of dioxins and furans below Celgar and other
British Columbia pulp mills resulted in action by the Canadian government New
regulations under the Canadian Environmental Protection Act (CEPA) have been
proposed to regulate the discharge of chlorinated organics. The Canadian federal
government is proposing limits of non-detectable amounts of dioxins and furans by
January 1994. In addition, the Province of British Columbia (B.C. Environment) has
adopted regulations to control adsorbable organic halides (AOX) discharged from
bleached kraft pulp mills. The control of AOX requires reductions in the use of chlorine
which, in turn, decreases the formation of dioxins and furans. The new regulations
require that, by 1993, AOX be limited to 2.5 kg per metric tonne of pulp produced.
Over the past decade, the B.C. Ministry of Environment has been trying to get
various owners of the Celgar pulp mill to resolve water pollution problems caused by
their failure to meet waste permit requirements. The identification of chlorinated
organics as a health issue has resulted in increased urgency on the part of the
Canadians to install pulping technology and effluent treatment works to resolve
problems. To meet these government requirements, Celgar has proposed a mill
modernization effort.
The most recent measurements of effluent quality discharged by the Celgar mill
were obtained during the Canadian Pulp and Paper Association survey (CPPA, 1990).
Information on present and projected levels of 2,3,7,8-TCDD and -TCDF have been
provided by Celgar. These are summarized in Table B-1. The load measured in early
1990 from the Celgar pulp mill is less than 1.37 mg/day. Since this survey, the mill
has made several improvements that were designed to further reduce dioxin and furan
levels in the effluent. Results of the follow-up sampling will be available later this year.
The amount of 2,3,7,8-TCDD measured from the Celgar mill in the 1990 survey is
significantly less than the loading capacity of 2.3 mg/day for the Columbia River at the
International Boundary. This does not consider other potential sources upstream of
the border. However, no other sources have been identified where 2,3,7,8-TCDD has
been detected.
Table B-1. Concentrations of TCDD and TCDF from Celgar Pulp
2,3,7,8-TCDD
Concentration Load
(ppq) (mg/day)
2,3,7,8-TCDF
Concentration Load
(ppq) (mg/day)
CPPA 1990 Survey
Projected after modernization
(from bleach plant)
ND (14) < 1.37
<0.0485
310 30.4
< 0.0485
-------
Final TMDL for Dloxln Discharges to the Columbia Basin
B-3
Celgar is also seeking government approval to increase the mill's production
from 560 to 1200 air dried metric tonnes of pulp per day. B.C. Environment recently
completed public hearings regarding the proposed Celgar pulp mill expansion project.
Modifications to the mill's production process are being proposed which include
oxygen delignification, 70% substitution of chlorine dioxide for chlorine, and hydrogen
peroxide bleaching followed by primary and secondary effluent treatment The
improvements to the Celgar mill are expected to be in place by 1994. Concentrations
of TCDD and TCOF in the bleach plant effluent are expected to be below detection
limits of 10 ppq. Maximum daily discharges after modernization are expected to be
<0.05 mg/day for TCDD and <0.05 mg/day for TCDF (Celgar, 1990). Recognizing
problems in the past, B.C. Ministry of the Environment has stated that: "Either Celgar
will have to significantly upgrade pollution control technology in their existing mill to
achieve compliance or they will face heavy penalties for breaking the law."
Several of the U.S. mills criticized the proposed TMDL (June 15,1990) for a
perceived lack of equity with Canada. The final TMDL estimates a loading of
0.31 mg/day from Celgar. This is equal to the loading which would be allocated to
Celgar if it were a mill in Region 10. This accounts for Celgar's planned production
after modernization (see Table 3-2) and applies a factor of 0.257 A»g/day of
2,3,7,8-TCDD discharged per ton of bleached pulp. This is the same factor used to
calculate the WLAs for the Region 10 mills. This is not a WLA but rather an estimated
loading. This estimate provides a margin of safety to cover other unidentified sources
in Canada and/or a possible shortfall in Celgar's attainment of the projected
0.05 mg/day loading. As additional information is assembled, this preliminary estimate
may be refined.
2. Other Potential Sources
The development of the TMDL needs to consider all potential sources of
2,3,7,8-TCDD in the Columbia drainage. Besides chlorine bleaching pulp mills, other
potential source categories include woodtreaters, major municipal wastewater
treatment plants, agricultural areas, industrial sites, and urban areas. Table B-2
summarizes potential sources of TCDD in the Columbia, the type of available
information on loading rates, and median fish tissue concentrations from the National
Bioaccumulation Study (NBS) associated with the source category. The NBS was
conducted as a screening investigation to determine the prevalence of selected
bioaccumulative pollutants in fish. One of the study objectives was also to identify
general correlations between fish tissue concentrations and sources of these
pollutants.
The NBS results, listed in Table B-2, clearly indicate that the highest levels of
TCDD contamination in fish were found in areas below chlorine bleaching pulp mills.
However, two other site categories from the NBS in the Columbia basin which were
not immediately below pulp mills had elevated levels of TCDD in fish. Both sites are
located in the north Portland area. One of the sites, Columbia Slough, is affected by
nonpoint sources, predominantly urban runoff and a landfill. The other site is located
below a major woodtreating operation (McCormick & Baxter) which uses
pentachlorophenol (PCP). TCDD contamination has been associated with PCP.
-------
Final TMDL for Dioxfn Discharges to the Columbia Basin
B-4
Table B-2. Potential Sources of 2,3,7,8-TCDD in the Columbia Basin
Source Category
Availability of
Data
for
Region 10
National Bioaccuoulation
Study Cooperative Results
(from draft report)
Median Cone, (ppt)
Chlorine Bleaching Pulp & Paper
104 mill study
4.73
Non-Chlorine Bleaching Pulp & Paper
N/A
1.30
Si^erfmd Sites
Renedial Investigations
1.47
Woodtreaters, Incinerators, etc.
TRI , DMR
1.39
Other Industrial Sites
N/A
1.27
Urban Areas
N/A
1.27
Municipal Wastewater Treatment Plants
Sewage Sludge Survey
0.64
Agricultural Areas
N/A
0.56
Other Sites
N/A
0.63
Note: N/A - Not Available
TRI - Toxics Release Inventory (PCP)
DMR - NPDES Discharge Monitoring Reports (PCP)
Woodtreaters:
A number of current and former wood treatment facilities exist in the Columbia
River basin where pentachlorophenol (PCP) has been used as a preservative. A
potential source of 2,3,7,8-TCDD from woodtreating facilities is contaminated PCP.
Thirteen sites near former or existing woodpreserving facilities were sampled during
the National Bioaccumulation Study. The median 2,3,7,8-TCDD concentration in fish
tissue at these sites was 1.39 ppt (compared to 4.73 for the chlorine bleaching pulp
mills). Of the thirteen sites sampled nationally near woodtreaters, only one was in the
Columbia River basin: the Willamette River at Portland (below McCormick & Baxter).
Three species of aquatic organisms were sampled at that site with the following
results:
Species 2.3.7.8-TCDD
Largemouth Bass 0.74 ppt
Sucker 2.22 ppt
Crayfish 2.61 ppt
The values for this site are higher than the median for the NBS. However, organisms
collected from this location are also influenced by other potential sources of 2,3,7,8-
TCDD, such as urban runoff.
These measured values reflect the need to evaluate information on the potential
discharge of 2,3,7,8-TCDD from woodtreating facilities. EPA has recently developed a
data system which contains information from the Toxics Release Inventory (TRI). A
retrieval of reported releases of PCP for 1987 identifies seven facilities (woodtreaters) in
the Columbia Basin (Table B-3). Five of these facilities are located in the Willamette
drainage. Although the TRI information does not contain data on TCDD, the indicated
releases of PCP lead to concern over woodtreaters, particularly in the Willamette basin.
DMR data and inspection reports describing PCP discharges are also available for
-------
Final TMDL for Dloxin Discharges to the Columbia Basin
B-5
several woodpreserving facilities with NPDES permits in the Columbia basin.
Table B-3. PCP Discharges from Columbia Basin Woodtreating Facilities
Cataloging
Unit
Facility Name
Location
NPDES
DMR
Data
TRI Data (lbs.
1987
(Water) (Total)
PCP released)
1988
(Water) (Total)
17010214
17010214
17010216
17010305
B.J. Carney
L.D. McFarland
Poles. Inc.
B.J. Carney Industries, Inc.
Sandpolnt. 10
Sandpolnt, ID
Oldtown, ID
Spokane. WA
C 1,850
C 500
17020003
17020003
Chewelah Log and Post
Colvllie Post and Pole
Chewelah. WA
Colvllle. WA
17040201
17040219
6arland Pole Co.
Penta Post
Idaho Falls. ID
Gooding, ID
17050114
17050114
Pressure Treated Timber
Roundy Pole Fence Co.
Boise, ID
Eagle, ID
C 7
17070105
J.H. Baxter & Co.
The Dalles. OR
17080001
17080001
17080001
17080003
Allweather Wood Treaters
Exterior Wood. Inc.
Pacific Wood Treating
International Paper Co.
Washougal. WA
Washougal, WA
Rldgefleld, WA
Longvlew, WA
I/R
250 2,300
B 1,500
17090001
17090003
17090003
17090008
17090010
17090010
17090012
Jasper Wood Treating
J.H. Baxter & Co.
L.D. McFarland
Taylor Lumber & Treating
Dant & Russell
Permapost
HcCormlck & Baxter
Jasper, OR
Eugene, OR
Eugene, OR
Sheridan, OR
North Plains, OR
Hlllsboro, OR
Portland, OR
X
X
0
X
250 1,250
250 1,500
250 13,488
0 250
31 6.999
200 202
B 750
B 2,150
150 154
Notes TRI data for releases of PCP to: Water (discharge)
Total (Includes water, air and land disposal)
B : 1 - 499 lbs.
C No discharge to water Identified X Loads calculated for PCP
I/R Inspection Report o Only PCP concentration reported
The preamble to a proposed RCRA rule relating to the wood preserving industry
(53 FR 53292, December 30, 1988) describes ranges of chlorinated dibenzodioxin and
chlorinated dibenzofuran as well as PCP concentrations in wastewaters from
woodtreating facilities. Thus, an estimate of potential 2,3,7,8-TCDD releases from
woodtreating facilities can be made based on data on PCP discharges. The TRI data
were considered in estimating TCDD wastewater releases from woodtreaters.
However, there are some apparent problems. Several facilities, for instance, reported
zero discharge to water while others reported the same value of 250 pounds. DMR
data, on the other hand, appear to provide better information on PCP discharges.
Applying assumed ratios of 2,3,7,8-TCDD per unit PCP (derived from Table 7,
53 FR 53292) to the DMR data, EPA estimates that 1 - 2 mg/day 2,3,7,8-TCDD could
be originating from woodtreating operations in the Columbia basin. This estimate
includes the potential release from facilities where no DMR or TRI data exists.
-------
Final TMDL for Dlaxln Discharges to the Columbia Basin
B-6
Levels of 2,3,7,8-TCDD observed in fish and sediments below one major
woodtreating operation plus estimates of potential loads point to the need for
additional data. Any allocation scheme used to develop the TMDL must leave room
for these facilities. Using available information, a range of 1 - 2 mg/day appears to be
a reasonable estimate. However, this estimate is preliminary and data are still being
generated. As additional information is assembled, this estimate may be refined. Most
of the released 2,3,7,8-TCDD is associated with site run-off during rainfall. Thus, the
loading from woodtreaters could be reduced by implementing stormwater controls.
Municipal Wastewater Treatment Facilities:
National data demonstrate that the sludges removed from some municipal
wastewater treatment plants contain dioxins and furans. Generally, octa-chlorinated
forms predominate the dioxins found in these sludges, although 2,3,7,8-TCDD has also
been detected. Where sludges are contaminated, the wastewater discharges could
also contain 2,3,7,8-TCDD. Testing performed for 2,3,7,8-TCDD in sludge nationally
included five municipal wastewater treatment plants in the Columbia basin ("National
Sewage-Sludge Survey Facility Analytical Results", U.S. Environmental Protection
Agency, 1989). Results for these five facilities are listed in Table B-4.
Table B-4. Columbia Basin Sludge Testing for 2,3,7,8-TCDD
Cataloging
Unit
Facility Name
Location
2,3,7,8-TCDD Detection
(ng/kg) Limit
HunlciDal WWTP's
17050114
West Boise STP
Boise, 10
ND (4 7)
H n M
m n
ND ( 6.1)
170B0001
Columbia Blvd. STP
Portland. OR
ND (16.0)
• t ft M
•t »•
ND ( 8.9)
17090005
Stayton STP
Stayton, OR
ND (23.0)
17090006
Lebanon STP
Lebanon, OR
3.3
ft M
N N
2.2
17090012
Tryon Creek STP
Lake Oswego, OR
ND (57.0)
M MM
mm n
ND (43.0)
Chlorine B1. Mills
17060306
Potlatch Corp.
Lewiston, 10
78.0
17070101
Boise Cascade
Wallula, WA '
70 0
17080001
James River
Camas, WA
12.0
17080003
Boise Cascade
St. Helens, OR
4.2
17080003
Longvlew Fibre
Longview. WA
69.0
17080003
Weyerhaeuser
Longvlew, WA
25.0
M M
M M
35.0
17080003
James River
Wauna, OR
19.0 (pri.)
N M
M rt
89 0 (sec.) —
17090003
Pope & Talbot
Halsey. OR
31.0
-------
Final TMDL for Dl&dn Discharges to the Columbia Basin
B-7
Of the five municipal facilities whose sludges were examined in the Columbia
basin, only one had detectable levels of 2,3,7,8-TCDD. This indicates that the TMDL
should leave some room for potential allocations to municipal sewage treatment plants.
Analytical results for this treatment plant, however, show that the detected
concentration was at levels much lower than sludge tested at chlorine bleaching pulp
mills (Table B-4). Thus, it can be expected that load estimates for municipal facilities
will be much lower than the loads allocated to the pulp mills based on the sludge data.
Initial estimates of 2,3,7,8-TCDD discharged from municipal wastewater
treatment facilities can be made using available data. Permitted total suspended solids
for each facility and an assumed average 2,3,7,8-TCDD concentration in municipal
sludge form the basis of these calculations. The analysis also assumes that
chlorinated dioxins / furans found in municipal sludge are associated with effluent
solids at the same concentrations. The average 2,3,7,8-TCDD concentration detected
was 2.8 ng/kg. The permitted total suspended solids load from Region 10 municipal
wastewater treatment plants in the Columbia Basin is over 170,000 pounds per day.
Based on this information, these municipal wastewater treatment facilities could, as a
group, contribute an average of 0.2 mg/day 2,3,7,8-TCDD. As additional information is
assembled, this preliminary estimate may be refined.
Other Industrial Sources:
Non-chlorine bleaching pulp mills (Table B-5) and other potential industrial
sources also need to be considered in the allocation process. No data has been
presented on 2,3,7,8-TCDD concentrations in either wastewater or sludges for
Columbia basin non-chlorine bleaching pulp mills. Another potential industrial source
of 2,3,7,8-TCDD is Rhone-Poulenc, located in north Portland. This plant has produced
chlorophenolic herbicides since 1956. The facility discharges boiler blowdown, cooling
water, site runoff, and treated groundwater to the Willamette River (across from
McCormick & Baxter). The effluent is known to contain chlorinated phenols, although
2,3,7,8-TCDD was not detected during a National Dioxin Study.
Table B-5. Non Chlorine Bleaching Pulp Mills in the Columbia Basin
Cataloging
Unit
Facility
Location
17010305
Inland Enpire Paper Co.
Spokane, WA
17080001
Boise Cascade Corp.
Vancouver, WA
17090003
Willamette Industries
Albany, OR
17090004
Weyerhaeuser
Springfield, OR
17090007
Smurfit Newsprint
Newberg, OR
17090012
Janes River II
West Linn, OR
17090012
Smurfit Newsprint
Oregon City, OR
An estimate of loadings from these sources cannot be determined at this time.
With respect to non-chlorine bleaching pulp mills, an analysis cannot be conducted
because no data has been identified which describes 2,3,7,8-TCDD in either effluents
-------
Final TMDL for Dioodn Discharges to the Columbia Basin
B-8
or sludges. As to Rhone-Poulenc, available data from the National Dioxin Study
showed non-detect for 2,3,7,8-TCDD. However, the detection limits were higher than
present day limits. As additional information is gathered, it will be possible to estimate
loadings from these sources.
3. Fate. Transport, and Attenuation
Losses of 2,3,7,8-TCDD in the water column can occur through sedimentation
(see discussion in next section), photolysis, and volatilization, as well as through
uptake by aquatic organisms. 2,3,7,8-TCDD's structural properties, laboratory
bioconcentration experiments, and field observations also indicate a strong potential
for bioaccumulation. Thus, the role of these processes needs to be expressed in
terms of potential bioavailability. Limited information exists which can be used to
provide initial estimates on the effects of fate, transport and attenuation in the
Columbia River system. Readily available, quality data have been considered. This
includes information from the Northwest Pulp & Paper Association's Columbia River
Fish Study (1989), from EPA's National Bioaccumulation Study (1987), from the
Washington Department of Ecology's work on Lake Roosevelt (1989-90), and from
efforts in Canada.
Several approaches exist to evaluate the effects of fate, transport, and
attenuation. Water quality models, using a variety of assumptions, can be used to
assess ambient data and to evaluate the need for additional controls. Available
analytical tools range from simple estimates to complex data-intensive dynamic
models. Analyses can include a loss rate which considers potential adsorption of
TCDD on particulate matter within the water column. The potential release of TCDD
from the sediment to the overlying water or the potential effect of sediment bound
TCDD on the benthic and aquatic life food chain must also be considered. However,
quantitative predications of bioaccumulation for specific cases and regulatory actions
are complicated by many uncertainties. These uncertainties include the degree of
partitioning between dissolved and bound phases, definition of the food chain structure
plus bioenergetic parameters, and the relative importance of other fate and transport
phenomena.
The Clean Water Act specifically states that TMDL's shall be established with a
margin of safety which takes into account any lack of knowledge. Based on the lack
of knowledge concerning attenuation of TCDD in the Columbia River basin,
assumptions must be made with respect to attenuation in determining the loading
capacity of the system and allocations of that capacity. A review of comments
received on the proposed TMDL did not provide conclusive evidence that net
attenuation occurs. Although TCDD may be lost to the sediments, that loss may only
be temporary because of resuspension, desorption, or biological uptake directly from
the sediments.
Figure B-2 superimposes predicted fish tissue concentration data on a graph of
the actual (measured) fish tissue data plotted in Figure 2-1 in Section 2 of this
document. Water column concentrations of 2,3,7,8-TCDD were modeled based on
(1) the results of TCDD sampling in source effluents (the "104-Mill Study), (2) receiving
-------
Final TMDL for D/ox/n Discharges to the Columbia Basin
B-9
water dilution calculated from the harmonic mean flows at the discharge points, and
(3) an assumption of no net attenuation. Predicted fish tissue concentrations were
then calculated using a bioconcentration factor of 5,000 (the factor used in developing
the water quality criterion). As in Figure 2-1, all fish tissue concentrations (both
measured and predicted) are displayed in terms of estimated cancer risk based on the
factors used to calculate EPA's water quality criterion for 2,3,7,8-TCDD. Both the 1CT®
and 1GT4 risk levels are identified. The 1CT* risk level corresponds to the 0.013 ppq
ambient 2,3,7,8-TCDD concentration which is the basis of the TMDL, while 1CT4
represents a level of possible concern due to non-cancer effects. Note that the line
plotted between data predicted based on an assumption of no net attenuation closely
follows the data points based on directly measured fish tissue concentrations.
Figure B-2. Columbia River Fish Tissue: TCDD
-
A
A ^
Concern due to Non-cancer Effects A a
-
—tsJ ' * \
6 fts & I ^ A
B " £f a a &
A Basis for Loading capacity A
-
.. __ • -r -j - 1 i , ,
0 200 400 600 800
River Mlle
~ 10t-miI I study load a Actual fish tissue
Based on the comparison in Figure B-2 of predicted tissue concentrations with
observed values, an assumption of no net attenuation appears to be reasonable.
Thus, for purposes of developing this TMDL, all 2,3,7,8-TCDD discharged is assumed
to remain in the water column and remain biologically available. Because this is a
conservative assumption, this TMDL should lead to the attainment of water quality
standards regardless of the actual level of attenuation. If future studies quantify a net
attenuation rate, allocations can be modified to reflect this. This capacity could be
used to provide an increased margin of safety to account for unknown sources,
increase allocations for existing sources, or accommodate future growth needs. By
the same token, if studies indicate that TCDD releases from historical accumulations in
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
B-10
the sediments constitute a problem, tighter controls may be needed (see discussion in
following section).
4. Role of Bottom Sediments
Sediment concentrations are the result of a complex series of interactions
between TCDD, the overlying water column, solids, aquatic organisms, and the
external loading of TCDD. Because of the hydrophobic nature of dioxin, there is a
tendency for TCDD to move from the water column to the sediments and aquatic
biota. Although attenuation may result in a net loss of TCDD from the water column,
the potential also exists for the sediments to act as a source of dioxin through the
release of TCDD which has accumulated (Figure B-3).
Floure B-3. Exchange of TCDD Between Water Column, Sediments, and Biota
Water
Column
Partitioning
Bioconcentration
Desorption
Resuspension
Decay
Excretion
Deposition
Sedimentation
Release
Uptake
Bioaccumulation
Some fraction of the TCDD which enters a river is quickly associated with solids.
The adsorption of TCDD to particulate matter may ultimately determine levels in fish
tissue. There are a number of different theories about the role of equilibrium
partitioning and bioaccumulation from contaminated sediments. The fate of TCDD in
the aquatic environment is increasingly being discussed in terms of food chain
mechanisms. Dioxins are believed to be adsorbed to bacteria, fungi, and organic
sediment particles. These particles are eaten by filter-feeding benthic invertebrates
which in turn are consumed by fish.
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
B-11
In addition, solids tend to settle to the bottom of the receiving water. In areas
where the river is not filling in, these particles (and the TCDD associated with them) will
continue to be carried downstream as either bedload or resuspended sediments. In
areas of sediment accretion, typically where river velocities are diminished, TCDD will
tend to accumulate in the bottom sediments where it may be available to aquatic
organisms. Resuspension of sediments either through high streamflows, boat traffic,
or dredging activities must also be considered.
Current knowledge of the Columbia system is not adequate to determine the
availability of TCDD associated with particulate matter to benthic organisms or fish on
a basin-wide basis. Existing sediment concentrations probably reflect a combination of
both current and historical discharges of TCDD. Because the Region's pulp mills have
implemented some process changes recently, such as the use of different defoamers,
it is unlikely that existing sediment contamination levels are in equilibrium with current
loadings to the basin. Also, if desorption of dioxin occurs slowly, it may take several
years to observe the effect of reduced discharges in sediments and in biota.
Limited sediment sampling for dioxin has been done in the Columbia system.
Data collected in the mainstem Columbia River below Bonneville Dam have not
detected 2,3,7,8-TCDD. However, current detection limits may be above the level of
concern considering the low organic content of the sediments analyzed. TCDD has
been detected in Willamette River sediments below a woodtreating operation. These
spatial differences reflect both physical characteristics and the influence of specific
sources. Thus, future studies on the effect of sediments should address site-specific
concerns.
Given these conditions it would not be appropriate to assume a permanent loss
of 2,3,7,8-TCDD through sedimentation. Indeed, a portion of the loading capacity
should remain unallocated to account for potential release from the sediments and
from TCDD currently stored in the food chain. As indicated in the discussion on
attenuation, tighter controls will be needed if data show that the cumulative effects of
historical discharges significantly delay attainment of TCDD standards under the
reduced loadings required by this TMDL
5. Future Allocations
TMDLs may provide a framework for dealing with future allocations. Examples
include the assignment of any unallocated portion of the loading capacity to specific
point or nonpoint sources. Future growth of the pulp industry in the Columbia River
basin, either expansion of existing mills or new mills, is a possibility which should be
considered in this TMDL
Developing an equitable framework for future allocations is not an easy task.
This TMDL reserves a portion of the loading capacity as unallocated for 2,3,7,8-TCDD
to account for uncertainties and to provide for future growth. As uncertainties are
reduced, the amount held back can be made available to other sources or for
additional future growth. Decisions on the use of the unallocated load will be made on
-------
Final TMDL for Dlcadn Discharges to the Columbia Basin
B-12
a case-by-case basis by EPA in consultation with the affected States. If proposed
projects are not consistent with this TMDL, a revised TMDL would need to be
established before the proposed increased loadings could be allowed.
-------
Final TMDL for Dioxin Discharges to the Columbia Basin
0-1
APPENDIX C. WASTE LOAD ALLOCATION METHODS CONSIDERED
In developing the proposed TMDL, several alternative waste load allocation
methods were considered for allocating portions of the loading capacity to chlorine
bleaching pulp mills. These alternatives were presented in the Decision Document for
the proposed TMDL to illustrate the effect of assumptions made on resulting WLAs and
to stimulate public consideration of the pros and cons of alternative allocation
scenarios. Included in the presentation of options was one preferred alternative.
There was no information received during the public comment period which has
caused EPA to change its decision about the preferred allocation method (Option 4,
Table C-2). Two additional options were suggested, however. These were:
(1) allocate the entire loading capacity to the bleaching pulp mills, and (2) require zero
discharge of dioxin from the pulp mills. The first suggestion is clearly inappropriate
since other sources, which are presently difficult to control, would cause the loading
capacity of the system to be exceeded. Appendix B includes additional discussion and
estimates of sources other than chlorine bleaching pulp mills which supports the
likelihood of this exceedence. The zero discharge option is also further discussed in
this document and in the response to comments. Zero discharge is not necessary in
order to meet water quality standards for dioxin in the Columbia River basin.
For the convenience of the public, the discussion of options contained in the
Decision Document for the proposed TMDL is repeated here. The alternative
approaches considered fall into several different categories which include:
Effluent Concentrations
Mass Discharge per Unit Production
Percent Reduction
Equal Effluent Concentrations:
One allocation option is to set an equal effluent concentration for each pulp mill
which uses chlorine bleaching. The resultant cumulative load is the portion of the
loading capacity allocated to chlorine bleaching pulp mills located in EPA Region 10.
Some margin of safety is then provided by the difference between the loading capacity
and the WLAs to the chlorine bleaching pulp mills in the Columbia basin of Region 10.
The unallocated amount depends directly on the effluent concentration selected.
A starting point is to look at a long term average effluent limit of 10 ppq (the
current general method detection limit) at each mill. This limit is initially applied at the
Equal
Equal
Equal
-------
Final TMDL for Dloadn Discharges to the Columbia Basin
C-2
point of discharge. Total plant effluent flows are used as a basis to calculate loads.
Discharge monitoring report (DMR) data have been summarized and includes average
effluent discharge rates.
Using a long term average effluent limit of 10 ppq applied at the point of
discharge and current estimates of monthly average flow at each mill, the cumulative
load from all the mills equals 11.7 mg/day (Table B-1). This is greater than the
loading capacity of 5.97 mg/day. Consequently, this option must be rejected because
water quality standards would not be met under conservative assumptions, such as no
attenuation. In addition, this would not account for any 2,3,7,8-TCDD from other
sources. Thus, more restrictive controls are needed.
A permit condition set at a level below the general analytical detection limit
creates a situation where it is difficult, if not impossible, to determine compliance.
Because dioxins and other chlorinated organic compounds are produced in the bleach
plant, concentrations of 2,3,7,8-TCDD are higher in the combined bleach plant flow
than in the total plant effluent This means that waste load allocations which result in
total plant effluent concentration limits that are below the general analytical detection
limit could be monitored for compliance by measuring concentrations in the combined
bleach plant waste stream. Using estimates of bleach plant flows and a long term
average limit of 10 ppq in the combined bleach plant flow, the cumulative load is
3.7 mg/day or approximately 62 percent of the total loading capacity (Table B-1).
Although this option yields a cumulative load from chlorine bleaching pulp mills which
is less than the loading capacity, several concerns exist:
¦ there is very little room for allocations to other potential sources, such as
woodtreaters or the mill in British Columbia (estimates described in
Appendix B indicate current loadings from other sources would exceed
the unallocated portion of the loading capacity)
¦ there would be no margin of safety
¦ future growth in the pulp & paper industry is not addressed
For these reasons, the possibility of yet lower effluent limits was evaluated. This
was accomplished by setting a "maximum" concentration of 10 ppq, rather than using
a long term average of 10 ppq. To understand how this results in a lower allocation,
the relationship between the waste load allocation (WLA) and the actual permit limits
must be examined. In certain cases, permit limits will be different than WLA values.
Because the criteria for 2,3,7,8-TCDD is set to protect human health, the loading
capacity (and WLAs) reflect a long term average. slt is important to consider how the
WLAs address variability in effluent quality. Permit limits are set at the upper bounds of
acceptable performance and are values not to be exceeded. Requirements are usually
expressed using two types of permit limits, either daily maximum or monthly average.
Procedures have been developed for computing monthly average permit limits from
long term average WLAs in EPA's TSD (Technical Support Document for Water
Quality-based Toxics Control", U.S. Environmental Protection Agency, 1985).
-------
Final TMDL for Dloadn Discharges to the Columbia Basin
0-3
Assuming a coefficient of variation (C.V.) of 0.6 describes the effluent variability
for 2,3,7,8-TCDD from pulp mills1 and one sample required to be taken per month, a
monthly average permit limit of 10 ppq converts to a long term average WLA value of
4.7 ppq. Using estimates of bleach plant flows and 4.7 ppq as the long term average
concentration limit for the combined bleach plant flow, the cumulative load is
1.8 mg/day or just over 30 percent of the total loading capacity. This leaves nearly
70 percent of the loading capacity available to cover loadings from other potential
sources. This approach also results in more than a 95 percent reduction in
2,3,7,8-TCDD discharged from these pulp mills when compared to estimates of current
loading based on results of the 104 mill study.
Table C-1. Waste Load Allocations for Chlorine-Bleaching Pulp Mills
Production
(tons/day)
Percent
Option 1
TCDO ULA
(mg/day)
(totion 2
TCDD ULA
(mg/day)
Option 3
TCDO ULA
(mg/day)
Mill
1,509
17.2
1.42
0.71
0.33
Potlatch
Lewiston, ID
957
10.9
0.76
0.14
0.06
Boi se Cascade
-- Uallula, UA
1,650
18.8
2.20
0.87
0.41
James River -
- Camas, UA
310
3.5
2.37
0.23
0.11
Longview Fibre
-- Longview, UA
1026
11.7
2.01
0.57
0.27
Weyerhaeuser
-- Longview, UA
1500
17.1
0.19
0.19
0.19
Pope & Talbot
-- Halsey, OR 2
1,035
11.8
1.29
0.64
0.30
Boise Cascade
-- St. Helens, OR
800
9.1
1.44
0.36
0.17
James River -
- Uauna, OR
7,837
100.0
11.67
3.72
1.84
TOTAL Source Cateaorv Allotment
Option 1: Set Equal Long Term Average Effluent Concentration of 10 ppq
at Point of Discharge
Option 2: Set Equal Long Term Average Effluent Concentration of 10 ppq
at Bleach Plant
Option 3: Set Equal Long Term Average Effluent Concentration of 4.7 ppq
at Bleach Plant
1 A C.V. of 0.6 Is recommended In EPA's TSD (Technical Support Document for Water Quality-
based Toxics Control*, U.S. Environmental Protection Agency, 1985) for situations where there
is insufficient data to estimate a C.V. for a specific pollutant from a specific industrial process. In
the fact sheet accompanying the public notice for the draft TMDL, EPA solicited Information of
use in developing a more appropriate C.V., If available, from the public. No such Information
was provided.
2 The WLAs listed for Pope & Talbot under all options have been adjusted to the long term
average of 0.19 mg/day identified In the NPDES permit issued by the Oregon Department of
Environmental Quality (November 7, 1990).
-------
Final TMDL for Dloxin Discharges to the Columbia BasJn
C-4
Equal Mass Discharge per Unit Production:
A disadvantage of equal effluent concentrations based on current flow rates is
that it may not be equitable for all mills. A common approach for industrial permits is
to consider production levels in establishing effluent limits. To provide for more equity,
each mill could be allocated an equal amount of 2,3,7,8-TCDD for discharge per
quantity of bleached pulp produced. One way to accomplish this is to associate
bleach plant flow rates with production quantity of bleach pulp. In estimating bleach
plant flows, the Washington Department of Ecology used 14,470 gallons of wastewater
generated per ton of bleached pulp produced. Applying this figure to calculate bleach
plant flows and 4.7 ppq as the long term average concentration limit for the combined
bleach plant flow, the cumulative load is 2.07 mg/day (Table B-2) or approximately
35% of the total loading capacity.
Table C-2. Waste Load Allocations for Chorine-Bleaching Pulp Mills
Pption 4: Set Equal Long Term Average Effluent Concentration
of 4.7 ppq at Bleach Plant and Set Rows at 14,470 gallons / ton
bleached pulp)
Production
Percent
TCDD WLA
(tons/day)
(mg/day)
Nil I
1.509
17.2
0.39
Potlatch -- Lewiston, ID
957
10.9
0.25
Boise Cascade -- Wallula, WA
1,650
18.8
0.42
Janes River -- Camas. WA
310
3.5
0.08
Longview Fibre -- Longview, WA
1026
11.7
0.26
Weyerhaeuser -- Longview, WA
1500
17.1
0.19
Pope ft Talbot -- Halsey, OR 1
1.035
11.8
0.27
Boise Cascade -- St. Helens, OR
800
9.1
0.21
James River -- Wauna, OR
7.837
100.0
2.07
TOTAL Source Category Allotment
Although this is an increase of 0.13 mg/day over that shown in Table 5-5, the
approach does address one major problem with using current bleach plant flows. Mills
have been encouraged to recycle internal waste streams to the maximum extent
possible. One example, Boise Cascade at Wallula, practices extensive recycling.
Under the equal effluent concentration method, a mill that does a high level of
recycling receives a lower allocation. However, a mill that does not make efficient use
of water in the bleach plant benefits from a high allocation. This is a major reason for
relating bleach plant flows to pulp production when determining allowable loads. This
1 The WLA listed for Pope & Talbot has been adjusted to the long term average of 0.19 mg/day
Identified In the NPDES permit Issued by the Oregon Department of Environmental Quality
(November 7, 1990).
-------
Final TMDL for Dloxin Discharges to the Columbia Basin
C-5
approach still results in more than a 95 percent reduction in 2,3,7,8-TCDD discharged
from these mills when compared to results of the 104 mill study. Based on the
evaluation in Appendix B, this reduction, although less than obtained by Option 3, is
still sufficient to achieve total 2,3,7,8-TCDD loadings to the basin which are less than
the loading capacity.
Equal Percent Reduction:
Another option considered is equal percent reduction for all source
categories. Because there is an absence of specific data for loadings of TCDD to
the Columbia, this approach can be viewed in several different ways. The first could
use information on the relative magnitude of 2,3,7,8-TCDD in fish collected below
potential sources of dioxin. Using median tissue concentrations summarized in Table
A-1 as a general indicator of these relative contributions, thirty-six percent (36%) of the
loading capacity could be attributed to chlorine bleaching pulp production. The
remaining sixty-four percent (64%) could be attributed to other sources, such as
municipal wastewater treatment plants or agricultural areas. This analysis excludes
refineries because this industry is not known to be a significant source in the Columbia
drainage. Although this approach does offer some advantages by accounting for
other source categories, there are some major drawbacks. These include:
¦ NBS was intended as a screening study and not to describe source
category loadings
¦ fish sampled nationally were collected from streams of varying sizes and
did not account for dilution
¦ results of NBS associated with certain source categories may also
include other sources (i.e. a site directly below a municipal wastewater
treatment plant may also be 30 miles below a bleached kraft pulp mill)
Another option suggested is to use values of 2,3,7,8-TCDD measured in
Columbia River fish and the bioconcentration factor used to develop the water quality
criterion (0.013 ppq) to "back calculate" current TCDD loads. Although it may be
possible to estimate the relative magnitude of present plus historic TCDD loading by
looking at tissue concentrations, other factors besides a weighted average
bioconcentration factor of 5000 must be considered. For instance, bioconcentration
factors specific to the species should be evaluated. The age of the fish and lipid
content of the samples must also be taken into account. The 5000 bioconcentration
factor used to develop the criterion is intended to represent the weighted average
factor for the species mix and lipid content in the "average" American fish / shellfish
diet. The lack of species-specific bioconcentration data, as well as the difficulty in
distinguishing the effects of historic versus current loading, makes using this approach
inappropriate for this TMDL at the present time.
------- |