NC0000272_Appendix D_20201204

Appendix D
Analysis of September 7, 2007 Fish Kill
Pigeon River -Blue Ridge Paper Mill
In its February 22, 2010 letter to Ms. Coleen H. Sullins of the North Carolina Department
of Environment and Natural Resources, James D. Giattina, Director of the Water
Protection Division, U.S. EPA, Region 4, wrote that EPA's concern for insufficiency of
the 2006 § 316(a) Demonstration by the Blue Ridge Paper Mill ("Paper Mill" or "Mill")
is "heightened by a North Carolina Wildlife Resources Commission report indicating that
a September 2007 fish kill in the Pigeon River was, in part, due to elevated temperature."
(The cited report is NC WRC 2007) This appendix analyzes the cited fish kill and puts it
into perspective with local and regional environmental conditions and regulatory history.
The Fish Kill
The cited fish kill is reported in the North Carolina Division of Water Quality's 2007
report on fish kills as event AS07004 in Haywood County below Canton where "8,000"
fish were killed (http://h2o.enr.state.nc.us/esb/Fishkill/documents/Events07.pdf). The
"[k]ill event [was] attributed to low flow/DO and high water temperatures brought on by
ongoing drought conditions." The report added "Investigators observed numerous live
fish during the investigation." The full fish kill report on which this entry was based (NC
WRC 2007) noted that at approximately 1600 hrs on September 7, dead fish were
observed below the Paper Mill where temperature was 35.4°C (no temperature was taken
upstream of the Mill at that time, and the exact location of the downstream measurement
was not recorded, according to the fish kill report). At about 1900 hours, the temperature
upstream of the Mill was 24YC and that below the Mill was 33.2°C (a temperature rise
of 8.9°C). "[T]his measurement and others collected by Mill personnel during the kill
did not exceed the limits specified in the current permit." (Letter from Shannon L.
Deaton, NC Wildlife Resources Commission to Tom Belnick, NC Division of Water
Quality, February 25, 2010). The next day (September 8) investigators found
"[a]pproximately 8,434 total fish dead" by sampling three 100-meter sites over a distance
of approximately 6 km (estimated from the Mill to "behind Caring Place Loop"). The
report identified the fish species as brown bullhead, northern hogsucker, silver shiner,
central stoneroller, tangarine darter, Tuckaseegee darter, channel catfish, redbreast
sunfish, smallmouth bass, warpaint shiner, whitetail shiner, greenfin darter, and rock
bass. Although the exact location of the fish kill was not stated in the NC WRC Fish Kill
Report, it probably occurred in the close vicinity of the Mill (e.g., from the Mill to
Fiberville) with stressed and dying fish being transported by currents downstream to the
lowermost sampling locations.
Mill Operations
Operating data for the Blue Ridge Paper Mill and the Mill's data for the river at the Mill
during early September show no exceptionally high discharges, but exceptionally low
river flow. The Mill discharge was steady at 24.7 — 26.7 mgd for September 1-7 (25.75

mgd on Sept 7). Upstream temperatures ranged 20.6 — 22.4°C during the week preceding
September 7 (20.6 on 9/7), while at Fiberville Bridge they were 25.5 - 33. VC (highest on
9n). River flows on September 4-7 were 34.9 — 38.8 cfs, with the lowest flow on 9/7).
The 8.9°C temperature rise late in the day on September 7, according to the fish kill
report, did not cause the permitted temperature rise to be exceeded. As noted above, the
NC Wildlife Resources commission acknowledged that the Mill was operating within its
permit limits.
Physical Conditions at the Site
At the time of the fish kill and in the week preceding it, the Pigeon River was at an
extremely low flow and warm temperature, while the Mill discharge was at normal
operational levels of both flow rate and temperature (Table 1). The Mill discharge flow
was actually below the historical maximum flow rate. If the river flow as recorded at the
Canton Gauge upstream of the Mill was reduced by Mill withdrawals equal to the amount
of water discharged from the Mill, the river flow at the thermal discharge would have
been zero. It is not clear whether the Mill actually withdrew this amount of water from
the river, but without any withdrawal by the Mill the discharge flow would have been 52-
54% of the river flow at the thermal discharge.
Table 1. Physical conditions in the Pigeon River and Mill discharge during the week of
the 2007 fish kill. Data from plant records and state fish kill report.
River flows at Canton Gauge
34.9 to 38.8 cfs
(upstream of mill)
River temperatures upstream of
20.6 to 22.4°C; NC WRD measurement of
Mill
24.3°C in the impoundment
Mill discharge flow
38.2 to 41.3 cfs (24.7 to 26.7 MGD)
Mill discharge temperature
33.8 to 36.9°C
River temperature below Mill
NC WRD spot measurements of 33.2 after the
kill and 35.4°C during the kill (locations not
reported)
River flow at discharge if amount
0
of water discharged equals the
water withdrawal by the Mill
Discharge as percentage of river
51.6 to 54.0 %
flow assuming no water
withdrawal
Historical Mill discharge flows
18.7 to 48.7 cfs (12.1 to 31.5 MGD)

With little (or possibly no) flow in the river other than the thermal discharge, the thermal
plume likely spread nearly across the river between the Mill and Fiberville Bridge (the in -
stream monitoring location). In this distance, temperatures likely did not decline much
from the discharge temperature. There were no detailed temperature measurements taken
in the thermal plume during the fish kill, although the state's fish kill report noted a spot
temperature measurement of 35.4°C below the Mill.
Regional and Local Environmental Conditions
There is strong evidence that an exceptional regional condition contributed to the fish kill
in the Pigeon River. The evidence shows that (1) this was not an isolated fish kill, (2)
river flow records independent of the Mill indicate near record low flows in the Pigeon
River from low regional precipitation, and (3) air temperatures at the time (which
strongly influence stream temperatures upstream of the Mill) were much above normal.
The Division of Water Quality's 2007 fish kill report lists another kill on August 29
(AS07002) in the Broad River in Rutherford County (also Southwestern North Carolina)
where 100 fish were killed. The "[e]vent occurred following several weeks of drought
conditions, hot weather and low flow." This is similar to the attribution for the Pigeon
River at Canton.
There is clear evidence that exceptionally low-water conditions existed in the Pigeon
River at the time of the fish kill of September 7. River flow data for the Canton USGS
station (Station Number 03456991) show a daily average river flow for September 7 of
49 cubic feet per second (cfs)
(http://wdr.water.usgs.gov/wy207/pdfs/03456991.2007.pdf). This is only 1.2 cfs above
the all-time record minimum September mean flow of 47.8 cfs for water years 1932-
2007. It followed the month of August in which the minimum mean daily flow was 45
cfs and the mean daily flow was only 61.6 cfs, with 12 days at 50 cfs or below. J. Curtis
Weaver, USGS hydrologist, was quoted in the Hendersonville Times -News on Friday,
November 9, 2007 as saying that most of the 210 stream gages in NC were around 10%
of normal stream flow by October, and that the stream flow in the West Fork of the
Pigeon River was the lowest ever recorded. These exceptionally low flows were below
the low -flow statistics generally used as worst -case conditions for establishing NPDES
permits [e.g., the seven-day, one in 10 year low flows or 7Q10, which is the worst -case
condition standard for § 316(a) demonstrations (Wabash and Cayuga Generating
Stations, Public Service of Indiana, NPDES Appeal #78-6, 1979)].
Climate records compiled by the National Climatic Data Center for 2007 confirm that the
time of the fish kill was exceptionally warm and dry. September 2007 was the eighth
warmest on record for air temperatures in the contiguous United States, with worsening
drought in the Southeast (www.ncdc.noaa.gov/oa/climate/research/2007/sep/sep07.html).
That web site noted "Raleigh -Durham International Airport reached a high of 101
degrees F (38 degrees C) on September 10, the latest date in any calendar year with a
maximum daily temperature greater than 100 degrees since records began in 1944." The
week ending September 8, 2007 had temperatures "much above normal" in southwestern

North Carolina, being 5.0 to 8YC (9-15°F) above normal and "extremely dry"
(www.ncdc.noaa. gov/oa/climate/research/us-
weekly.php?year=2007&month=09&day=8&submitted=Submit). Stream temperatures
in small watersheds like the Pigeon River are known to closely follow air temperature
trends.
Fish Species
The NC WRD fish kill report identified 13 species of dead fish being collected. These
were brown bullhead, northern hogsuckeer, silver shiner, central stoneroller, tangerine
darter, Tuckaseegee darter, channel catfish, redbreast sunfish, smallmouth bass, warpaint
shiner, whitetail shiner, greenfin darter, and rock bass.
Although the NC WRD's fish kill report indicated dead fish were found up to about 6 km
from the Mill, it is likely that the dead fish collected at the most downstream locations
actually died farther upstream, indicating that the fish kill occurred in the reach close to
the Mill. Also, the high temperatures that may have caused the fish kill observed on
September 7 likely occurred over a several -day period prior to the observed kill. This is
because it takes time for lethally high temperatures to be manifested in loss of
equilibrium (which causes dying fish to be washed downstream) and death.
Temperatures in the mill Mischarge and river immediately downstream of the Mill (when
dilution was likely nil) were within the range that would be lethal to many riverine fishes
(Table 2). A search of the literature located upper lethal temperature tolerance data for 6
of the 13 species collected, and data for related species that seem to be reasonable
surrogates for four others. As footnoted in the table, the usual test to determine lethal
temperatures is reported as a temperature for 50% mortality. It is conventional to
subtract 2°C from this temperature to estimate the temperature where there is 100%
survival (Table 2, column 2). Also, the test generally lasts for several days (usually 96 hr
to one week). A 24-hour survival temperature is somewhat higher, generally estimated to
be 1-2°C above the UILT minus 2°C for longer exposures (Table 2, column 3).
The fish species that were recovered by the fish kill investigation team in approximately
6 km of river below the thermal discharge demonstrated existence of a diverse
assemblage. Many of the species are categorized as pollution intolerant. Others are
valued sports fish. While it is regrettable that some individuals of these species were
killed, the kill served as a sampling of species' composition that indicated that a highly
diverse fish assemblage had occupied the most thermally affected zone of the river
immediately downstream of the Mill in summer. The fish kill did not eradicate fish from
this zone, because the investigators noted live fish swimming in the area of the kill as the
dead fish were being observed and collected.
Regulatory Perspective
It is important to view the fish kill of September 7 in the context of the 2006 § 316(a)
Demonstration provided by the company in its application for a renewed NPDES permit.

The EPA letter implied the fish kill negated the basis of a § 316(a) Demonstration: that a
balanced indigenous population or community (BIP/BIC) be shown to exist. There is
relevant information in the statute, regulations, guidance and administrative precedent to
inform a different view.
Table 2. Upper temperature tolerances of some fish species collected by the North
Carolina Wildlife Resources Commission team investigating the September 7, 2007 fish
kill at the Blue Ridge Paper Products Mill in Canton, NC.
Species in fish kill (a)
UILT — 2°C (b)
Safe daily
maximum
exposure
Brown bullhead
32.8
34
Northern hogsucker
32
33
Common shiner
31.5
33
(surrogate for all other
shiners)
Central stoneroller
29 (c)
32 (c)
Channel catfish
35
36
Bluegill sunfish
31.8
33
(surrogate for redbreast
sunfish)
Smallmouth bass
33
34
Rock bass
33
34
(a) Species of fish found in the 2007 fish kill for which data are available in literature
summaries (or closely related species that seem appropriate surrogates).
(b) Upper Incipient Lethal Temperature (UILT) is that temperature at which 50% of
the sample is dead after exposures generally 96 hr to one week. Subtracting 2°C
approximates the temperature for no mortalities. Acclimation temperatures are all
25°C or above except central stoneroller, which was tested with field samples
captured at 12-30°C. Sources: NAS/NAE 1973, Brown 1974, Wismer and
Christie 1985.
(c) Due to some fish having been tested at field acclimation temperatures as low as
12°C, this value is likely low for summer acclimation.

Statute
Section 316(a) of the federal Clean Water Act refers to a balanced indigenous population
in the "body of water." That has been understood to mean that the Congress's interest
was the whole body of water, and not necessarily every part of it. In evaluating this
potential uncertainty, the Environmental Appeals Board has said that § 316(a) applies to
the "receiving waters:" "In other words, to the extent thermal discharge limitations that
are less stringent than the otherwise applicable effluent limitations will nonetheless
preserve a balanced community of indigenous aquatic life in the receiving waters, EPA
may incorporate such less stringent limits into an NPDES permit." In re Aurora Energy,
L. L. C., 2004 EPA App. LEXIS 30 *6 (E. A. B. Sept. 14, 2004). The body of water in
this case is the Pigeon River downstream of the Mill, in which the zone of the fish kill
was but a small part.
Regulations
The implementing regulations for § 316(a) seem to state quite clearly that the BIPBIC is
intended to apply outside a zone in which the effluent is initially mixed with the rest of
the water body. Section 125.62(c)(2) expressly says that a BIP must exist "immediately
beyond the zone of initial dilution of the applicants modified discharge" and "in all areas
beyond the zone of initial dilution." Section 125.73(a), like the statute, refers to the
"body of water." The most consistent interpretation of the regulations would mean that
the permitting authority should view the big picture of the water body, even if there is a
violation in some small area. The September 7, 2007 fish kill in the Pigeon River
appeared to occur in the zone immediately downstream of the thermal discharge where
the effluent is incompletely mixed.
Guidance
Guidance by EPA, both general and thermal, supports the view that the BIPBIC standard
and a lack of any appreciable harm does not need to be demonstrated in the immediate
vicinity of the discharge, where the fish kill apparently occurred. Mixing zone provisions
of the Clean Water Act have been applied generally across a range of pollutants,
including thermal. (see,
hqp://www.epa.jzov/waterscience/standards/mixin zo�pics.html and
http://www.epa.gov/waterscience/standards/mixingzone/docs.html, for example). Mixing
zones are areas where EPA intends that water quality criteria and standards do not need
to be met. In concert with application to other pollutants, it has generally been assumed
that the § 316(a) BIPBIC biological "standard" also should be applied outside the mixing
zone. The "mixing zone" appears to be the direct analog of the "zone of initial dilution"
of Section 125.62(c)(2).
An exception from the BIPBIC standard in the zone near the discharge was specifically
discussed in the § 316(a) Guidance Manual (EPA and NRC 1977). The Guidance
Manual specifies a "Master Rationale" or concluding argument of the demonstration.
Guidance for the Master Rationale specifically considers a legal mixing zone where

damage may occur: 'Receiving water temperatures outside any (State established) mixing
zone will not be in excess of the upper temperature limits for survival, growth, and
reproduction, as applicable, of any RIS occurring in the receiving water." Section 3.8.4,
page 71.
Other guidance documents prepared by EPA are relevant to the general matter of mixing
zones as zones near a pollutant discharge where some impairment is allowed. EPA's
1991 Technical Support Document for Water Quality -Based Toxics Control (EPA 1991)
advises that impacts in mixing zones not impair the integrity of the water body "as a
whole." (p. 70). EPA's 1998 Guidelines for Ecological Risk Assessment (EPA 1998)
includes evaluation processes similar to the Guidance Manual for thermal discharges. An
important element of the EPA ecological risk assessment guidance is spatial scale: does
the area of impact constitute a large percentage of the "landscape?" The guidance
indicates that factors to be considered "include the absolute area affected, the extent of
critical habitats affected compared with a larger area of interest, and the role or use of the
affected area within the landscape." (EPA 1998, p. 117). "Landscape" is used in its
broad sense, because the guidelines are written for terrestrial as well as aquatic
assessments. "Landscape" would equate to water body or water body segment in aquatic
analyses.
Congress specifically recognized the availability of the mixing zone concept as a
mechanism for dealing with thermal discharges pursuant to § 316(a) of the Act. During
the House debate on the Conference Report, Representative Clausen, a member of the
conference managers group appointed by the House, stated: "Section 316(a) in effect
recognizes the temporary localized effects a thermal component may have as well as the
potential beneficial effects. It encourages the consideration of alternative methods of
control, including mixing zones, so long as the controls assure the protection and
propagation of a balanced indigenous population of shellfish, fish, and wildlife."
Administrative Precedent
Litigation and administrative decisions that characterized the early years of
implementation of § 316(a) established the precedent that a reduction in the population of
particular species in the immediate area of the discharge did not preclude a successful §
316(a) demonstration. For example, a reduction in the population of a particular species
in a localized area was found to be acceptable by the Administrator after appeal of the
§ 316(a) decision for the Wabash and Cayuga generating stations. In re Public Service
Co. of Indiana, Inc. (Wabash/Cayuga Generating Stations), NPDES Appeal No. 78-6,
1979 EPA App. LEXIS 4, 22, 1 E.A.D. 590 (Nov. 29, 1979). Although the overall fish
populations in the Wabash River were unaffected by operations of the generating
stations, some species were virtually eliminated from the power plant sites. The appeal
decision stated: "... [In] attempting to judge whether the effects of a particular thermal
discharge are causing the system to become imbalanced, it is necessary to focus on the
magnitude of the changes in the community as a whole and in individual species" and
then determine these "changes are appreciable." Id. The zones near the discharges did
not meet the BIC criterion, but the broader ecosystem did. Similarly, the area to be

considered for application of the BIC was judged to be the broad area of the water body
segment in the Region 1 Administrator's decision regarding the Pilgrim Power Plant (in
In re Boston Edison Co. (Pilgrim Power Plant) Determination Regarding Issuance of
Proposed NPDES Permit No. MA0025135 (EPA Region 1, March 11, 1977)). The
Administrator found the impact of Pilgrim to be "minimal in comparison to the species
population in the area of impact." Similarly, in In re Central Hudson Gas & Elec. Corp.,
EPA GCO 63, 1977 WL 28250 (July 29, 1977), EPA noted that § 316(a) "permits an
adverse environmental impact so long as the impact does not interfere with the protection
and propagation of the balanced indigenous population in the aquatic ecosystem."
This precedent was upheld recently in regard to the Brayton Point plant, which is called
out for other purposes in the attachment to the EPA letter ("Section 315(a) Report and the
Study Plan for the Subsequent Permit"). A decision on the Brayton Point plant affirmed
that some of Mount Hope Bay, Massachusetts and Rhode Island, could be warmed above
24°C (the upper avoidance temperature for winter flounder) so long as it did not occur in
more than 10% of the bay for more than five days a year. In re Dominion Energy
Brayton Point, L.L.C., NPDES Permit No. MA 0003654, NPDES Appeal No. 07-01,
2007 EPA App. LEXIS 38 (EAB September 27, 2007).
Conclusion
The evidence supports a conclusion that the fish kill of September 7, 2007 was an
extraordinary event, limited in time, brought about by extreme low river flows and high
late summer water temperatures in the region and the Pigeon River as a whole in August
and early September. Further, the fish kill occurred in a limited zone immediately
downstream of the thermal discharge, in which the effluent is incompletely mixed. Such
a zone is expressly excluded from the BIP/BIC criterion by statute, regulations, EPA
guidance and administrative precedents. Blue Ridge Paper Products' Mill operated under
its normally permitted conditions during the time leading up to the fish kill. This resulted
in its thermal discharges not being sufficiently cooled by low ambient river flows less
than the 7Q10 flow, with the result that river temperatures just below the Mill
temporarily exceeded the thermal tolerances of several fish species. The species
composition of the kill indicated that a diverse assemblage of fishes had been occupying
the approximately 6-km reach below the Mill prior to the fish kill, in accord with
BIP/BIC criteria. As an extraordinary and brief event due to abnormal regional climatic
conditions, and in a zone normally excluded from meeting BIPBIC criteria, it is
inappropriate to use this fish kill as a measure for determining the adequacy of a § 316(a)
Demonstration, as was the implication in the EPA Region 4 letter.
References
Brown, H. W. 1974. Handbook of the effects of temperature on some North American
fishes. American Electric Power Service Corporation, Canton, Ohio.
NAS/NAE (U.S. National Academy of Sciences and National Academy of Engineering).
1973. Water Quality Criteria 1972. A report of the Committee on Water Quality Criteria,

Environmental Studies Board. EPA-R-73-033, Environmental Protection Agency,
Washington, DC.
NCWRC (North Carolina Wildlife Resources Commission). 2007. NCWRC Fish Kill
Report. September 7-8, 2007. Raleigh, North Carolina.
Wismer, D. A., and A. E. Christie. 1985. Temperature relationships of Great Lakes
fishes: a data compilation. Ontario Hydro, Toronto.