NC0055786_Correspondence_19851029

NPDES DOCUMENT !MCANNINO COVER :SHEET
NPDES Permit:
NC0055786
Lexington Regional WWTP
Document Type:
Permit Issuance
Wasteload Allocation
Authorization to Construct (AtC)
Permit Modification
Complete File - Historical
Engineering Alternatives (EAA)
Correspondence
Owner Name Change
Instream Assessment (67b)
Speculative Limits
Environmental Assessment (EA)
Document Date:
October 29, 1985
This document is printed on reuse paper - ignore any
content an the rezrerse side

State of North Carolina
Department of Natural Resources and Community Development
Division of Environmental Management
512 North Salisbury Street • Raleigh, North Carolina 27611
James G. Martin, Governor R. Paul Wilms
S. Thomas Rhodes, Secretary Director
October 29, 1985
Elizabeth Southerland
Monitoring and Data Support Division
USEPA
(WH-553)
401 M Street, SW
Washington, DC 20460
Dear Ms. Southerland:
I appreciate your interest and willingness to help North Carolina
with the development of a mercury (Hg) wasteload allocation for the
City of Lexington. I realize that your interest in the Lexington WWTP/
Abbotts Creek Hg problem extends beyond wasteload allocations. However,
any insight that you offer toward resolution of this problem will be
greatly appreciated.
Enclosed you will find:: a memorandum which describes the current
situation, upstream and downstream ambient monitoring data for Hg, a
map showing ambient station locations relative to the discharge site,
effluent monitoring data, USGS streamflow gaging data, and some mis-
cellaneous information regarding water quality in Abbotts Creek.
Problems that you may encounter in using this information for a "monte
carlo" analysis include: 1) the large number of ambient observations
at or below detectable limits, and 2) the age of the USGS flow data
(collected between 1941 and 1957).. Please let me know if further
information can be provided.
As we discussed in our telephone conversation, I had planned to
attempt a monte carlo analysis to examine the probability of standard
violations given particular wasteloads. However, I ran into a problem
of what to use for a background Hg concentration distribution because
of the number of values at or below detectable limits. I am interested
to know how you would approach this problem, since it is one that we
frequently encounter.
Pollution Prevention Pays
P.O Box 27687, Raleigh, North Carolina 27611-7687 Telephone [phone]
An Equal Opportunity Affirmative Action Employer

Elizabeth Southerland
October 29, 1985
- page two -
In addition, I am interested in how you would approach incorporating
an exposure assessment into the above analysis. Currently, we rely upon
a steady-state assumption of a 7Q10 flow, even for toxics. The present
justification for use of the 7Q10 is that our standard is set to protect
to the chronic no -effect level and that chronic effects could occur over
a seven day period. The criterion is also based upon practical aspects
such as existing information availability and the limitation of resources
for chemical—specific1time-variable models. Although the monte carlo
analysis would help determine the percent of time that a specified level
would be exceeded, it will not provide information regarding the duration
of specific exposure levels.
Your cooperation in this problem is greatly appreciated. I look
forward to hearing from you soon.
Sincerely,
J. Trevor Clements
Technical Support Unit
JTC:mlt
Enclosure

DIVISION OF ENVIRONMENTAL MANAGEMENT
October 28, 1985
MEMORANDUM
TO: File
FROM: Trevor Clements
SUBJECT: Lexington, NC
Mercury Wasteload Allocation
The Lexington WWTP NPDES Permit (NC0024384) is up for renewal, and
will require a limit for mercury (Hg) . Lexington receives Hg from the
local Duracell Battery Manufacturing Plant, and from low level concen-
trations in their domestic waste. Effluent concentrations of Hg ranged
between 0.5 ug/1 and 5.7 ug/1 until plant modifications in 1984 reduced
the average to below 0.5 ug/l. The existing final permit limit for Hg
is 0.5 ug/1 and was derived via a simple, steady-state, mass -balance
dilution model. Model inputs reflected a 7Q10 design flow and assumed
an upstream Hg concentration of 0.0 ug/l.
Since the original issuance of Lexington's NPDES permit, DEM has
established several ambient monitoring stations both upstream and down-
stream of Lexington's discharge. Hg concentrations have been measured
in the water column, sediments, and fish tissue. The fish tissue and
sediment data collected downstream of the discharge indicate that con-
ditions have not substantially improved in the lower Abbotts Creek
segment. High percentages of fish (25 to 50 percent) sampled at the HWY
47 Station continue to possess levels of Hg above the 1.0 mg/kg critical
level recognized by the FDA. Levels in the downstream sediment have not
appeared to decline and may be providing an internal source of loading
back into the water column.
The challenge facing DEM is to determine a wasteload allocation
(WLA) for Hg in light of current knowledge of the Abbotts Creek system.
The WLA should be based upon two criteria: 1) protecting aquatic life

-2-
at the chronic no -effect level, and 2) reducing the potential for bio-
accumulation in fish above the FDA recommended level. Based upon an
earlier EPA recommendation, North Carolina has established a water quality
standard for mercury at 0.2 ug/1. Although EPA has revised their recom-
mendation to 0.012 ug/1, the current standard of 0.2 ug/1 must still apply.
Therefore, the water quality standard will provide the performance measure
for the first objective. A food -chain model is needed to adequately
address the latter concern.
North Carolina's standard for Hg applies under a 7Q10 design flow
criteria. The USGS provided the following hydrologic information for
Abbotts Creek at the Lexington WWTP outfall:
Station No. 0212152100
Drainage Area 183 sq. mi.
Average Flow 165 cfs
7Q10 3.0 cfs
The standard will apply, therefore, at or above flows of 3.0 cfs.
A preliminary WLA can be derived using a simple, steady-state,
mass -balance dilution model. This model should not provide a bad pre-
diction, since there are no other direct dischargers of mercury. The
highest concentration can be expected to occur at the mixpoint. The
following model form will be used:
cw =
QD CD - QU Cu
QW
where: C = Wasteflow Concentration
QD = Downstream Flow (QW + 7Q10)
CD = Downstream Concentration (W.Q. Standard)
QU = Upstream Flow (7Q10)
C = Upstream Concentration
QW = Wasteflow (design capacity)

-3-
Most of the model inputs are straight forward, with the exception
of upstream concentration. Although an ambient monitoring station is
located just upstream of the discharger (0.7 miles), much of the data
collected is listed at the detectable level of 0.2 ug/1. Therefore, it
will be impossible to choose a background concentration that does not
contain bias. If the 0.2 values are included as -is or if they are
eliminated, then the estimate of the mean will be biased on the high side.
Conversely, if these values are assumed zero, then the estimate of the mean
concentration will be underestimated. For lack of a better method, all
values listed at detectable limits will be halved and included in the
estimate. Values from 1982 forward will be used, since the detectable
limit before that time was 0.5 ug/1. This results in the following:
No. Observations 44
Minimum Conc. (ug/1) 0.1
Maximum Conc. (ug/1) 1.9
Mean (ug/1) 0.168
Standard Deviation lug/1) 0.275
It should be noted that 36 of the 44 observations (82 percent) were listed
at detectable levels. In addition, the sample (1.9.ug/1) collected on
5/16/85 is probably a strong influence on the mean since it is a magnitude
of order higher than the other observations. The coefficient of variation
is 164 percent and indicates that the mean is not a very good description
of central tendency. However, the median may not provide a very good
number either since it will be defined by our policy of halving the
detectable limit. The model will be run at both the mean and the median,
and the results will be compared to evaluate the significance of this
assumption.
The design parameters for Lexington are:
QU = 3.0 cfs
Cu = 0.168 ug/1 or 0.1 ug/1
Qw = 5.5 MGD (8.5 cfs)
QD = 3.0 + 8.5 = 11.5 cfs
CD = 0.2 ug/1 (standard)

-4-
Using the mean background concentration, the allowable effluent concen-
tration is:
C (11.5 cfs) (0.2 ug/1) - (3.0 cfs) (0.168 ug/1)
W 8.5 cfs
= 0.211 ug/1
If the median is used, the allowable effluent concentration becomes:
C _ (11.5) (0.2) - (3.0) (0.1)
W 8.5
= 0.235 ug/1
If it is assumed that there is no background mercury, then the result is:
C (11.5) (0.2) - 0
W 8.5
= 0.270 ug/1
In any of the above cases, the wasteload allocation will be fairly
stringent. In order to more accurately assess the probable impacts to
Abbotts Creek given any one of these allocations, it would be interesting
to run a Monte Carlo analysis to develop response curves for each waste -
load. In this manner, a better idea of the probability of a standard
violation can be obtained. This approach will be pursued by the Technical
Support Unit.
In addition, it may be possible to set up a simple food chain model
to approximate the time period needed before Hg concentrations in game
fish drop below the FDA criterion. A model such as that developed by
John Connolly, of Manhattan College, might be applicable. His model
describes changes in fish concentration (by age class) as a function of
uptake from food, uptake from water, excretion, and growth. There may be
enough field data and literature on the subject to develop this model for
the Abbotts Creek arm of High Rock Lake. This idea will be further
investigated.
JTC:mlt

Altstopit
•
•
(*ccJ ,.t7-4 0. 2
1 � 1
re_J i OV$ \ btu 04- 0 . O S
e 440
Ni,a VosAu:k tAutA _soc?-7, GAL701.6
coNA-44:6vIc
? � to pi= 3 g c-�s
) • Ca us)
0•2 I1.3;5)
0.7 8st 4—
(a, /k-3 u&L4
3oOz. = 1-7•4 cks
0-.
�iQ