NC0004987_Characterization Study Plan (ESCP) 2016_20160531

�' DUKE Environmental Services
ENERGY, Duke Energy
526 South Church Street
Charlotte,NC 28202
Mailing Address:
Mail Code EC13K/P.O.Box 1006
Charlotte,NC 28201-1006
May 25, 2016
Mr. Tom Belnick, Supervisor
NPDES Complex Permitting
NC DEQ/DWR/WQ Permitting Section
1617 Mail Service Center
Raleigh, NC 27699-1617
Subject: Marshall Steam Station
National Pollutant Discharge Elimination System - Permit No. NC0004987
316(b) Entrainment Characterization Study Plan (ECSP)
Dear Mr. Belnick:
Please find enclosed the final Entrainment Characterization Study Plan (ECSP) for Marshall Steam
Station (MSS). The plan incorporates comments from the fisheries biologists peer reviewers, as
well as comments from the North Carolina Department of Environmental Quality (NCDEQ)
received on March 7, 2016. Responses to comments from NCDEQ are also enclosed.
If you have any questions or comments, please contact Nathan Craig at Nathan.Craig@duke-
energy.com / 704-382-9622 or Scott La Sala at [e-mail] / 828-478-
7820.
Sincerely,
„Wid Zijej/..v
Rick Roper
Marshall Steam Station, Station Manager
General Manager III
FHO- Carolinas West RECEIVEDINCDEQIDR
Attachments MAY 31 2016
Br J.Scott LaSala DE Water Quality
Brad Loveland—DE '"Section
Nathan Craig-DE Permitting
Linda Hickok-DE
Tom Thompson-DE
Matt McKinney—DE
Ty Ziegler—HDR,Inc.
UPS- 1Z8955X40391792486
www.duke-energy.com

Response to NCDEQ
Comments:
Entrainment Characterization Study
Plan — Marshall Steam Station
Prepared for:
(a DUKE
'r ENERGY
Prepared by:
HDR Engineering, Inc.
RECEIVED1NCDEQIDWR
1 _..A -.'`: MAY 31 2016
Water Quality
econ
Permitting

Entrainment Characterization Study Plan
Response to NCDED Comments—Marshall Steam Station
Contents
1 Introduction 1
2 Development of Entrainment Characterization Study Plans 1
3 Twice per Month Sampling for Estimating Entrainment 3
3.1 Conclusion 5
4 Clarifications on'Variability"and a Method for Estimating 95 Percent Confidence Intervals 6
4.1 Data Analysis and Confidence Intervals 6
4.2 Conclusion 8
References 9
Duke Energy I I

Entrainment Characterization Study Plan 101
Response to NCDEQ Comments—Marshall Steam Station
1 Introduction
The U.S. Environmental Protection Agency's (EPA) rule implementing §316(b) of the Clean
Water Act(the rule)was published on August 15, 2014 in the Federal Register. The rule applies
to existing facilities with design intake flows (DIF)of more than 2 million gallons per day (MGD)
that withdraw from Waters of the United States, use at least 25 percent of that water exclusively
for cooling purposes, and have or require an NPDES permit. Existing facilities with an actual
intake flow a 125 MGD are required to provide an entrainment characterization study (at 40
CFR 122.21(rX9)) as part of its permit renewal application materials. Duke Energy facilities in
the Carolinas that meet the 125 MGD flow threshold are proposing a two-year entrainment
sampling program to gather the information necessary to fulfill the entrainment characterization
requirements. The goal of the proposed program is to estimate the seasonal and annual total
abundance of fish eggs and larvae that are drawn into the cooling water systems. Entrainment
sampling at each facility is proposed from March through October of 2016 and 2017.
2 Development of Entrainment Characterization
Study Plans
An Entrainment Characterization Study Plan (ECSPs) was developed Marshall Steam Station
(MSS). The ECSP describes the sampling design and site-specific approach being used at MSS
to sample for entrainment; the rationale for the selection of gear type, sampling location, and
other components of the sampling design; and a discussion on how the study fulfills the
requirements of the rule. While not required to undergo a peer-review, a draft of the ECSP was
sent to a subject matter expert in fisheries biology for an independent review. After addressing
comments and incorporating relevant changes, the final ECSP was sent to the Director for
comment. Comments were received from Bryn H.Tracy(Senior Environmental Specialist, North
Carolina Department of Environmental Quality [NCDEQ]) by email on March 7, 2016. We thank
Mr. Tracy for his comments and have provided responses in Table 2-1.
Duke Energy I 1

KespUrlseLUIVI.VC�./laUlnIllcnW IVIOI,11011JLC/5111lulouv„
Table 2-1. Responses to Bryn H. Tracy (Sr. Environmental Specialist, North Carolina Department of Environmental Quality) Comments on the Entrainment Characterization Study Plan for Marshall Steam
Station
Facility Comment Response and Resolution
Page 15,Section 5(same as Allen and Belews Steam Stations comments),first and second paragraphs.a)Please provide reason(s) •
Marshall as to why you decided on bi-weekly sampling(or is it just twice a month?)vs.weekly sampling. I attended an NC AFS workshop in Please see Section 3 for a discussion on twice per month sampling.
2005 lead by Dr.Doug Dixon(EPRI)and I have in my notes that bi-weekly sampling is more biased than more frequent sampling.
Marshall b Please rovide citations 'usti m as to wh twice per month sam lin is sufficient.; Please see Section 3 for a discussion on twice per month sampling.
P O1 fy. 9 Y P g n..:..-..
w,. 1r"
Replication in entrainment abundance sampling is rare, except when paired Bongo nets are used.Generally,sampling of a large volume of water
(e.g., 100 m3)for each sample is understood to integrate temporal variation in entrainment densities. Neither the EPRI(2014)entrainment abundance
c)One sample collected everysix hours leaseprovide whythere is no replication within a six hour period. I read your guidance document nor the Final 316(b)rule or preamble considers the need to collect duplicate or replicate samples.This suggests that neither EPA
P P reason(s) P p nor EPRI believes it to be a critical issue for entrainment sampling. Because entrainment varies temporally and spatially,true replication would require
Marshall response in Appendix B(and I wondered about this issue even before I got to Appendix B),but wonder why you chose not to replicate, simultaneous sampling with a second set of gear,which would roughly double the equipment costs.The increased effort in the field during sampling
even before any samples are collected? Is this common practice not to replicate? Has this been done at other projects? would be marginal,but the laboratory processing effort(and associated costs)would also double.Samples are collected during four periods at each
bimonthly sampling event for eight samples in each month.Confidence intervals can be generated utilizing data from these eight samples as one way
to better understand variability without increasing sampling(See Section 4).
Page 27, Section 7.1,second paragraph:A very good explanation.Also,please keep in mind that the reservoir's productivity ma
Marshallll . ... , .. .. , ,..- ..
have changed during the past 40 years and that the fish community now includes several nonindigenous species which were not o response necessary.
present 40 years ago. AMMINI10.1011111MIK
- _. , ... ..r. . ,. "
E.
Marshall Page 29, Section 7.3.1.a)Please provide reason(s)as to why you decided on monthly sampling vs.weekly or biweekly sampling for Please see Section 3 for a discussion on twice per month sampling.
this study.
Samples will be collected during one event in each month.Two boats will be used to collect the samples—one on each side of the skimmer wall.
b)My understanding from the second paragraph and from Section 7.3.2 is that you will be sampling in front of and behind the skimmer During each sampling event,four shallow samples will be collected on each side of the skimmer wall(for 8 shallow samples). In addition,the study
wall with two boats each collecting two separate samples from a 10 ft.depth for a total of four samples,plus a single bottom tow for a plan indicated one deep sample will be collected on each side of the skimmer wall(for two deep samples).As a result,on each sampling event 10
grand total of 10 samplesper month.So for this study,from the 10 ft.depthyou will have repeat sampling n=4 Isunderstandingsamples are proposed to be collected five from each side of the skimmer wall.
P P P P 9( )• my
correct? Note that during recent equipment testing activities,safety concerns regarding the deep samples were identified which may result in the inability to
collect deep samples.As of the writing of this document,the feasibility of collecting deep samples continues to be evaluated.
RECENEDINCDEOID
MAY 31 2616
Water Q Saton
Permitting Section

Entrainment Characterization Study Plan
Response to NCDEQ Comments—Marshall Steam Station
3 Twice per Month Sampling for Estimating
Entrainment
Sampling interval — or the time between sampling events — is typically assigned to a regular
schedule(e.g., one month,two weeks, one week, etc.). In some cases,the interval may change
on a seasonal basis (EPRI 2014). In the rule and its preamble, EPA provides no guidance on
the sampling interval associated with entrainment characterization studies required under
§122.21(r)(9). The ECSPs reviewed by NCDEQ are based on twice per month 24-hour
sampling events where each sampling event contains four samples collected throughout the 24-
hour period, resulting in eight depth integrated entrainment samples per month. This sampling
frequency was selected based on careful consideration of study objectives, gear selection, and
historical 316(b) studies. This section provides additional clarification on sample frequency
selection based on NCDEQ comments.
The Electric Power Research Institute (EPRI), in its 2014 entrainment abundance monitoring
support document, took advantage of a large dataset from an intensive entrainment monitoring
program undertaken at the Indian Point Generation Station on the Hudson River, NY. These
data were collected over 5 years (1983-1987) continuously or near-continuously (24-hours per
day, 7-days per week). Using these data, different sampling frequency and intensity scenarios,
including monthly, twice per month, and weekly sampling frequency, were modeled for their
effect on entrainment abundance estimates.
A sampling frequency of one day in a twice per month interval produced a coefficient of variation
(CV—the ratio of the standard deviation to the mean)of annual entrainment estimates ranging
from 750 percent for taxa entrained at very low densities (0.001 per 100 m3) to roughly 50
percent at the highest densities (100 per 100 m3). The CV declined as the frequency of
sampling increased (Table 3-1). At densities between 0.1 and 1.0 per 100 m3 and greater, the
CV stabilized between 25-50 percent for weekly sampling and 50-75 percent for twice per
month sampling.
The Hudson River estuary is a dynamic and variable environment that is affected daily by tides,
salinity and temperature gradients, and freshwater and nutrient inputs. At locations such as this,
species compositions and their relative abundance can fluctuate rapidly. In addition, the
spawning season is more spread out than in freshwater (i.e., spawning can take place in any
month of the year). Comparatively, a southeastem Piedmont reservoir represents a more stable
ambient sourcewater system and the spawning season is discrete within the year. Based on
these factors, it is assumed that comparable levels of precision can be achieved with less
frequent sampling in southeastem Piedmont reservoirs than would be needed on the Hudson
River. Even Brunswick Steam Electric Plant (BSEP), which withdraws from an estuary, exhibits
similar estimates in entrainment based on monthly and twice per month sampling as described
below.
Organisms collected at very low densities will generally not contribute substantially to the
economic benefits valuation and as a result selection of fish protection technologies will not
Duke Energy 13

Entrainment Characterization Study Plan
Response to NCDEQ Comments—Marshall Steam Station
typically be driven by species rarely entrained'. Therefore, greater variability (CV) in estimates
for these rarely collected species should be acceptable for the Entrainment Characterization
Study and the analyses these data support. In addition, as shown in Table 3-1, there is
relatively little difference in the CV between the weekly and twice per month sampling
frequencies for the more abundant species which have a greater affect on the valuation of
economic benefits and the selection of fish protection technologies.
Table 3-1. Estimated Mean Coefficient of Variation of Annual Entrainment Estimates
Based on 100 Iterations of Different Sampling Scenarios (100 m3 per Sample) Applied to
Indian Point Entrainment Data from 1983-1987 (Modified from EPRI 2014)
Approximate Coefficient of
Sample Mean Variation (%)
Density Weekly Twice per Month
(No./100 m3) Sampling Sampling
0.001 275-525 425-750
0.01 125-50 350-150
0.1 50-75 50-150
1 25-50 50-75
10 25-50 50-75
100 25-50 50-75
We note EPRI (2014) did not provide entrainment estimates associated with the sample
frequencies modeled. Our assumption is the total estimated entrainment generally remains
consistent regardless of the frequency of sampling, but the CV around the estimate decreases
with increased sampling. This is consistent with observations at BSEP. Data from yearly
sampling at BSEP from 1979 to 2004 were used to demonstrate that estimates of entrainment
based on weekly sampling were similar to those based on a monthly sampling frequency (see
green and blue lines in Figure 3-12). Based on this data, and consultation with NCDEQ, Duke
Energy decreased the sampling frequency at BSEP from weekly to monthly.
' A potential exception to this rule is when Federally-listed species may be involved in entrainment,
however, no Federally listed species are associated with potential for entrainment at these Duke Energy
facilities.
2 Baseline and Post-baseline refer to prior to and after the installation of entrainment reduction technology
and operation measures (fine-mesh panels and seasonal flow reduction) at BSEP, respectively. The
post-baseline data includes all sampling frequencies(i.e., weekly and monthly combined).
Duke Energy 14

Entrainment Characterization Study Plan
Response to NCDEQ Comments—Marshall Steam Station
100 -
ami 80 -
u
°- 60 -
a,
40 J
50th Percentile 80th Percentile
Nu. % Nu.
20 Baseline 9.45 18.46
U Weekly 1.44 84.7 4.13 77.6
Monthly 2.38 74.9 5.06 72.6
Post-baseline 1.79 81.1 4.79 74.0
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Number entrained (millions)
Baseline
—Weekly —Monthly Post-baseline
Figure 3-1 Cumulative Distribution of Total Entrainment at Brunswick Steam Electric
Plant under Varying Sampling Frequency
3.1 Conclusion
In determining sampling frequency, twice per month sampling was selected to balance the level
of uncertainty and costs with rule requirements and considerations of how these data would be
used in the later §122.21(r) reports. Changing the sample frequency from twice per month to
weekly would double or more3 the program costs yet not materially improve the quality of the
larger technology evaluations required under the rule. For these reasons, the twice per month
sample frequency was proposed in the study plans and remains the recommendation after both
peer review and NCDEQ comments have been received.
3 Increasing sample frequency from twice per month to weekly will at least double the cost of the
entrainment characterization program. Given the scale of Duke Energy's fleet-wide entrainment
characterization effort, such an effort may require hiring of additional contractors and field laboratory
staff raising the potential for even further cost escalation.
Duke Energy 15

Entrainment Characterization Study Plan 101
Response to NCDED Comments—Marshall Steam Station
4 Clarifications on "Variability" and a Method for
Estimating 95 Percent Confidence Intervals
The rule requires "sufficient data to characterize annual, seasonal, and diel variations in
entrainment, including but not limited to variations related to climate and weather differences,
spawning, feeding, and water column migration." Through this language, the rule is requiring
sufficient sampling to collect data over the range of conditions at the intake and to prevent bias
through selective sampling. For example, samples should be collected during the day and night
to capture the variation in entrainment associated with diel periods. Many marine and freshwater
fish species exhibit diel vertical migrations (Hellman 1993; Cech et al., 2005; Mehner et al.
2007;Adams et al., 2009). The typical pattern is for organisms to ascend in the water column in
the evening and return to deeper water during the day — although some species have the
reverse pattern (Lampert 1989). In addition to species and life stage, other factors such as light,
temperature, turbidity, and presence of predator and prey species, can influence vertical
migration. Because vertical migration is common, samples should be collected from throughout
the water column during day and night to capture this variation.
The rule requires that the entrainment data capture variations in entrainment rates from biotic
and abiotic factors, but EPA did not intend for facilities to tease apart the causal relationships
between these factors and entrainment. First, covariance of several factors identified by EPA
would make it difficult, if not impossible, to design a study to separate their effects on
entrainment. Using the example from above, separating differences in entrainment associated
with "water column migration"from those due to "diel variation" would not be possible for many
species, because vertical migration behavior is triggered by time of day (i.e., diel period).
Second, for some of the factors identified by EPA, understanding their relationship to
entrainment has no practical application toward determining Best Technology Available. For
example, entrainment rates may change with weather, but use of weather as a decision-making
tool for electrical generation and as a method for compliance with 316(b) is unreasonable.
4.1 Data Analysis and Confidence Intervals
The proposed Entrainment Characterization Study has sufficient "replication" to allow the
development of a 95 percent confidence intervals (95% CI) around the annual estimated
entrainment. Samples are collected during two 24-hour sampling events in each month. With
four samples collected during each sampling event, a total of eight samples are collected each
month at each facility. These individual samples can be used as replicates within each month to
estimate a month specific variance and a 95% CI around the annual entrainment estimate. It is
important to note that this 95% CI would be conservative, i.e., it would err on the side of over
estimating the upper and lower bounds on the 95%CI, because it would include variance due to
assignable causes(e.g., diel variation). Nonetheless, it may be useful in the context of 316(b)as
it provides and upper and lower bound on the estimated annual entrainment. The following
provides one potential method for developing 95% CI based the sampling design contained in
the ECSPs.
Duke Energy 16

Entrainment Characterization Study Plan
Response to NCDEQ Comments—Marshall Steam Station
Organism densities, expressed as number per 100 m3, would be calculated from entrainment
data for each species and life stage. Calculation of sample event densities provides seasonal
abundance trends while diel densities provide abundance trends throughout a day based on all
samples combined. Densities would be calculated as the sum of the total collected divided by
the total sample volume in m3, for the relevant interval, times 100. These densities can then be
used to calculate the total number of early life stage fish entrained during March through
October and the associated 95% CI can be calculated. First, the average concentration of
organisms per unit volume in the h'"stratum(i.e., month sampled), xh , would be calculated as:
1 "'
3C; = �x"
nh 1=1
where:
nh=the number of samples in the h"'stratum
xh,is the f observation in the h"'stratum.
The total number entrained(E)is then,
E=EVkxk
h=1
where:
H=total number of months sampled
VI,=volume of water withdrawn by the station in the if stratum.
The variance of the estimated total entrained is:
S2
Var(E)=EVh(1—fh)—"
h=1 nh
where:
Sh =variance of the if stratum
i(/xhi—ib/\1z
_ 1=1
nh —1
Duke Energy 17

Entrainment Characterization Study Plan 101Response to NCDEQ Comments—Marshall Steam Station
and fa=finite population correction for the h"'stratum. The finite population correction would be
computed as the volume of facility flow sampled in the month (i.e., to stratum) divided by the
total plant flow during the month. This factor becomes important when a substantial percentage
(>10 percent)of the total flow is sampled.
The 95 percent CI can then be computed as:
Eu,per=E+tg,df iVar(E)
E =E—ter fJYar(E)
where:
a=specified probability of Type I error, in this case 0.05
df=degrees of freedom, n-1
4.2 Conclusion
The final rule does not require replication nor is there an obligation to generate confidence
intervals or bounds around the individual entrainment estimates for a given sampling episode or
stratum (month or year). The use of the term 'variation" in §122.21(r)(9) is not synonymous with
"replication"; rather, variation refers to the natural temporal and spatial changes in organism
densities. The Entrainment Characterization Study must be sufficient to characterize diel,
monthly, and annual variation, which our proposed studies address. While not required by the
rule, the proposed Entrainment Characterization Studies generate data sufficient to develop
reasonable estimates of entrainment and 95 percent CI to estimate variability, if needed.
Duke Energy 18

Entrainment Characterization Study Plan
Response to NCDEQ Comments—Marshall Steam Station
References
Adams, C.F., R. J. Foy, J. J. Kelley, and K. 0. Coyle. 2009. Seasonal changes in the diel
migration of walleye Pollock (Theragra chalcogramma) in the northern Gulf of Alaska.
Environmental Biology of Fishes 86: 297-305(as cited in Donner and Eckmann 2011).
Cech, M., M. Kratochvil, V. Drastik, and J. Matena. 2005. Diel migrations of bathypelagic perch
fry. Journal of Fish Biology 66:685-702(as cited in Donner and Eckmann 2011).
Donner, M.T. and R. Eckmann. 2011. Diel vertical migration of larval and early juvenile burbot
optimises survival and growth in a deep, pre-alpine lake. Freshwater Biology 56: 916-925.
Electric Power Research Institute (EPRI). 2014. Entrainment Abundance Monitoring Technical
Support Document: Updated for the New Clean Water Act §316(b) Rule. 3002001425.
EPRI, Palo Alto,CA. 156 pp.
Heitman, G.S. 1993. Fish behavior by day, night, and twilight, In: Behavior of Teleost Fish (T.J.
Pitcher, Ed.) pp. 366-387. The Johns Hopkins University Press, Baltimore. (as cited in
Donner and Eckmann 2011).
Lampert,W. 1989. The adaptive significance of diel vertical migration of zooplankton. Functional
Ecology 3: 21-27.
Mehner, T., P. Kasprzak, F. HOlker. 2007. Exploring the ultimate hypothesis to predict diel
vertical migrations in coregonid fish. Canadian Journal of Fisheries and Aquatic Sciences
64: 874-886. (as cited in Donner and Eckmann 2011).
Duke Energy 19

I
I FY
I
I
I
I
I
I Entrainment Characterization
I Study Plan
& Supplemental Studies
IPrepared for:
I * DUKE
' ENERGY
I Prepared by:
HDR Engineering, Inc.
IApril 15, 2016
I
Marshall Steam
I
Station
I
I
I
I

Entrainment Characterization
Study Plan
& Supplemental Studies
Prepared for:
1
cts NERGY
' Prepared by:
HDR Engineering, Inc.
April 15, 2016
Marshall Steam
Station
1

1 Entrainment Characterization Study Plan )1
Marshall Steam Station r
Contents
1 Introduction 1
I1.1 Regulatory Background 1
1.2 Study Plan Objectives and Document Organization 3
I 2 Generating Station Description 4
2.1 Source Waterbody 4
I 2.2 Station and Cooling Water Intake Description 5
2.2.1 Intake Structure 5
3 Historical Studies-Entrainment 10
' 4 Threatened and Endangered Species 10
5 Basis for Entrainment Sampling Design 11
I 6 Entrainment Characterization Study Plan 16
6.1 Introduction 16
I 6.2 Sample Collection 17
6.2.1 Location 18
6.3 Sample Sorting and Processing 23
I6.4 Data Management 24
6.5 Entrainment Data Analysis 24
I 6.6 Field and Laboratory Audits 25
6.7 Laboratory Quality Control 26
6.8 Reporting 26
I
6.9 Safety Policy 26
7 Evaluation of the Skimmer Wall to Reduce Entrainment 27
I7.1 Introduction 27
7.2 Historical Studies 27
I
7.2.1 Methods and Materials 27
7.2.2 Results and Discussion 28
7.3 Proposed New Skimmer Wall Evaluations 29
I7.3.1 Sampling Location and Frequency 29
7.3.2 Sampling Gear Specifications and Sampling Protocol 32
I7.3.3 Data Management and Analysis 33
8 References 34
1 APPENDIX A—Select Species Spawning and Early Life History Data 35
' Life History References 38
I
Duke Energy I I
I

Entrainment Characterization Study Plan r IL�1
Marshall Steam Station
APPENDIX B—Response to Informal Review Comments 39
APPENDIX C—Comparison of Pumps and Nets for Sampling Ichthyoplankton 45
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Duke Energy I ii

Entrainment Characterization Study Plan
Marshall Steam Station
Tables
Table 1-1. §316(b)Rule for Existing Facilities Submittal Requirements Summary 2
Table 2-1. Design Flow Rate by Unit and Capacity Factor by Unit and Year at Marshall Steam Station6
' Table 5-1. Advantages and Disadvantages of Hoop Nets and Pumped Samplers for Estimating
Ichthyoplankton Density in Cooling Water Intake Structures (some information adapted from EPRI 2014)
13
' Table 5-2. Potential Disadvantages of Pumped Ichthyoplankton Sampling at Marshall Steam Station14
Table 5-3. Summary of Approach for Development of §122.21(r)(9) Required Entrainment
Characterizations 16
Table 6-1. Entrainment Sampling Details 17
' Table 7-1. Mean Monthly Densities (larvae/1,000 m3)of Larval Fish in the upper 5 meters (16 feet) of the
water column upstream of the Marshall Steam Station skimmer wall,April through September 1975 28
Table 7-2. Study Design Matrix 30
Table A-1. Life Histories of Selected Species Present Near Marshall Steam Station 36
Table B-1. Directed Charge Questions 39
Table B-2. Peer Reviewer Responses to Directed Charge Questions 41
Table C-1. Advantages and Disadvantages of Hoop Nets and Pumped Samplers for Estimating
Ichthyoplankton Density in Cooling Water Intake Structures (some information adapted from EPRI 2014)
' 47
Table C-2. Total Number (N) and Mean Densities (MD) (mean number of shad/ 1,000 m3) of All Shad
' Collected with Comparison Gear and Shad <28 mm Total Length Collected with a Tucker Trawl on Lake
Norman, North Carolina, 6-10 June 1982, with Average Volume of Water Filtered per Sample (m3)
(Leonard and Vaughn 1985) 56
Table C-3. Summary of Major Studies Designed to Comparatively Evaluate the Sampling Efficiency of
Various Large-Volume Pumps and Tow Nets (Taggart and Leggett 1984) 58
1
1
Duke Energy I iii
t

Entrainment Characterization Study Plan
Marshall Steam Station r)1
' Figures
Figure 2-1. Marshall Steam Station Vicinity Map 5
Figure 2.2. Site Configuration of Marshall Steam Station 7
' Figure 2.3. Plan View of Marshall Steam Station Cooling Water Intake Structure(Source:Alden 2012) 8
Figure 2.4. Section View of Marshall Steam Station Cooling Water Intake Structure (Source: Alden 2012)
9
1 Figure 4-1.Geographical Boundary of the IPAC Search 11
Figure 6-1. Plan View of the Marshall Steam Station's Cooling Water Intake Structure with Approximate
Sampling Locations—Sampler not to Scale(Image Modified from:Alden 2012) 19
Figure 6.2. Section View of the Marshall Steam Station's Cooling Water Intake Structure with
Approximate Location of Sample Inlets at Three Depths — Sample not Shown to Scale (Image Modified
from:Alden 2012) 20
Figure 6.3. Deck Level View Showing Approximate Locations of the Sampling Gear(Image Modified from:
Bing Maps) 21
' Figure 6.4. Example Entrainment Pump Sampling System Configuration 22
Figure 7-1. Approximate Sampling Locations at the Skimmer Wall of Marshall Steam Station (Source:
' Image Modified from: Google Earth) 31
Figure 7-2. Example Tucker Trawl Net(Source:www.algalita.org) 32
t
t RECEIVEDINCDEQIDWR
MAY 31 2016
' Water Quality
Permitting Section
Duke Energy i iv
1

IEntrainment Characterization Study Plan 101
Marshall Steam Station r
IAcronyms and Abbreviations
I
°C degrees Centigrade
' °F degrees Fahrenheit
AIF actual intake flow
Alden Alden Research Laboratory, Inc.
I
AOQL Average Outgoing Quality Limit
BTA Best Technology Available
cm centimeter
ICSP continuous sampling plan
CWIS cooling water intake structure
I
DIF design intake flow
Director National Pollutant Discharge Elimination System Director
DTRM double-trip release mechanism
I
Duke Energy
El. Duke Energy Carolinas, LLC
elevation
ECSP Entrainment Characterization Study Plan
I
EPRI Electric Power Research Institute
fps feet per second
gpm gallons per minute
' HDR HDR Engineering, Inc.
IPAC Information for Planning and Conservation (website)
m3 cubic meter
IMarshall Marshall Steam Station
MW megawatt
I
pm
mm micrometer or micron
millimeter
MGD million gallons per day
I MIL-STD military-standard
NPDES National Pollutant Discharge Elimination System
NCDENR North Carolina Department of Environment and Natural Resources
I NCWRC North Carolina Wildlife Resources Commission
Normandeau Normandeau Associates, Inc.
PVC Polyvinyl chloride
IQA Quality Assurance
QC Quality Control
I
SOP
USFWS Standard Operating Procedures
U.S. Fish and Wildlife Service
I
I
Duke Energy l iv

Entrainment Characterization Study Plan L1�
Marshall Steam Station rJ
1 Introduction
1 1 .1 Regulatory Background
t The Clean Water Act was enacted in 1972 and introduced the National Pollutant Discharge
Elimination System (NPDES) permit program. Facilities with NPDES permits are subject to
§316(b) of the Act, which requires that the location, design, construction and capacity of cooling
' water intake structures (CWIS) reflect best technology available (BTA) for minimizing adverse
environmental impacts. Cooling water intakes can cause adverse environmental impacts by
drawing early life-stage fish and shellfish into and through cooling water systems (entrainment)
' or trapping juvenile or adult fish against the screens at the opening of an intake structure
(impingement).
On August 15, 2014, the final §316(b) rule for existing facilities was published in the Federal
Register. The rule applies to existing facilities with design intake flows (DIF) of more than 2
million gallons per day (MGD) that withdraw from Waters of the United States, use at least 25
' percent of that water exclusively for cooling purposes, and have or require an NPDES permit.
The rule supersedes the Phase II rule, which regulated large electrical generating facilities until
' it was remanded in 2007, and the remanded existing-facility portion of the previously
promulgated Phase III rule. The final rule became effective on October 14, 2014.
' Facilities subject to the new rule are required to develop and submit technical material, identified
at §122.21(rX2)-(14), that will be used by the NPDES Director (Director) to make a BTA
determination for the facility (Table 1-1). The specific material required to be submitted and
' compliance schedule are dependent on actual intake flow rates (AIF) at the facility and NPDES
permit renewal date. Existing facilities with an AIF 2 125 MGD are required to address both
impingement and entrainment and provide explicit entrainment studies which may involve
' extensive field and economic studies (§122.21(r)(9)-(13)). Existing facilities with AIF < 125 MGD
have fewer application submittals. For such facilities, the Director must still determine BTA for
entrainment on a site-specific basis and the applicant may supply information relevant to the
' Director's decision. Facilities are to submit their §316(b) application material to their Director
along with their next permit renewal, unless that permit renewal takes place prior to July 14,
2018, in which case an alternate schedule may be negotiated.
' Duke Energy Carolinas, LLC's (Duke Energy) Marshall Steam Station (Marshall) is subject to
the existing facility rule and, based on its current configuration and operation, is anticipated to
be required to develop and submit each of the §122.21(r)(2)-(13) submittal requirements with its
next permit renewal in accordance with the rule's technical and schedule requirements. Within
' the §122.21(r)(2)-(13) requirements, (r)(4), (7), (9), (10) and (11) have specific requirements
related to entrainment evaluations (refer to Table 1-1 for additional detail). Duke Energy will also
undertake an evaluation of the existing skimmer wall to quantify the reduction in ichthyoplankton
' available for entrainment at Marshall. This document provides an Entrainment Characterization
Study Plan (ECSP) and the methods for evaluating the skimmer wall to support §316(b)
compliance at the facility with consideration of these specific requirements. As a part of
development of this Study Plan, Duke Energy submitted an earlier draft of this document to
Duke Energy I 1

Entrainment Characterization Study Plan
Marshall Steam Station EN
treview by a subject matter expert in the field of fisheries (see Appendix B) and identified to the
State as a peer reviewer.
While the equipment and methods contained in this Study Plan were developed with the intent
to be implemented as written, changes to the Study Plan may be required based on facility
requirements and/or situations encountered during execution.
1 Table 1-1. §316(b) Rule for Existing Facilities Submittal Requirements Summary
Submittal Requirements Submittal Descriptions
Iat1122.21(r)
(2) Source Water Characterization of the source water body including intake area of influence
Physical Data
1 (3) Cooling Water Intake Characterization of cooling water system;includes drawings and narrative;description of operation;
Structure Data water balance
Source Water Characterization of biological community in the vicinity of the intake;life history summaries;
(4) Baseline Biological susceptibility to impingement and entrainment;must include existing data;identification of missing
I
Characterization data data;threatened and endangered species and designated critical habitat summary for action area;
identifies fragile fish and shellfish species list(<30 percent impingement survival)
;...�,.rj' arrative description of cooling water system and intake structure;proportion of design flow used;aer
System Data ter reuse summary;proportion of source water body withdrawn(monthly);seasonal operation
ummary;existing impingement mortality and entrainment reduction measures;flow/MW efficiency
114 Chosen Method of Provides facility's proposed approach to meet the impingement mortality requirement(chosen from
6 Compliance with seven available ocompliance,if required b
( ) Impingement Mortality options);provides detailed study by Standard selected compliance option;addresses entrapmentlan for monitoring where required
I Entrainment Provides summary of relevant entrainment studies(latent mortality,technology efficacy);can be
POO Ce studies from the facility or elsewhere with justification;studies should not be more than 10 years old without
justification;new studies are not requir
Provides operational status for each unci;age ori Capa lza o s for est v@ y t3'
I 8) Operational Status upgrades within last 15 years;uprates and Nuclear Regulatory Commission relicensing status for
nuclear facilities;decommissioning and replacement plans;current and future operation as it relates
to actual and design intake flow
IRequires at least two years of data to sufficiently characterize annual,seasonal,and diel variations
in entrainment,including variations related to climate,weather,spawning,feeding,and water
$ column migration;facilities may use historical data that are representative of current operation of the
Entrainment facility and conditions at the site with documentation regarding the continued relevance of the data
to document total entrainment and entrainment mortality;includes identifications to the lowest taxon
II
Characterization
Study possible;data must be representative of each intake;must document how the location of the intake
in the water body and water column are accounted for;must document intake flows associated with
the data collection;documentation in the study must include the method in which latent mortality
would be identified(including QAQC);sampling and data must be appropriate fora uantitative
I
mpsurvey
r
Coi�ensrve".�'
(10) Technical Feasibility Provides an evaluation of technical feasibility and incremental costs of entrainment technologies;
&Cost Evaluation Net Present Value of facility compliance costs and social costs to be provided;requires peer review
I
Study
Provides a discussion of monetized and non-monetized water quality benefits of candidate
entrainment technologies from(r)(10)using data in(r)(9);benefits to be quantified physical or
Benefits Valuation biological units and monetized using appropriate economic valuation methods;includes changes in
1 (71)si,Study fish stock and harvest levels and description of monetization;must evaluate thermal discharges,
facility capacity,operations,and reliability;discussion of previous mitigation efforts and affects;
benefits to environment and community;social benefits analysis based on principle of willingness-to-
pay;requires peer review
I
Duke Energy 1 2
I

Entrainment Characterization Study Plan
Marshall Steam Station F
' Submittal Requirements Submittal Descriptions
at§122.21(r)
12 Non-Water Quality Provides a discussion of non-water quality factors(air emissions and their health and environmentalp noise,safety,grid reliability,Environmental and
Other Impacts impacts,energypenalty,thermal discharge, consumptive water use,
( ) p
Assessment etc.)attributable to the entrainment technologies; requires peer review
Documentation of external peer review,by qualified experts,of submittals(r)(10),(11),and(12).
(13) Peer Review Peer Reviews must be approved by the NPDES Director and present their credentials.The
applicant must explain why it disregarded any significant peer reviewer recommendations.
g
(14) New Units Identify the chosen compliance method for the new unit
' 1 .2 Study Plan Objectives and Document Organization
The ECSP provided in this report was developed to support Marshall's §316(b) compliance
' through development of a site-specific entrainment study plan with the following key objectives
in mind:
I1. Collect data to support development of§122.21(r)(9) which requires at least two years of
entrainment studies be conducted at the facility;
2. Collect data to support development of §122.21(r)(7) which allows for summaries of
relevant technology efficacy studies conducted at the facility;
3. Collect data to support Duke Energy's objective of having data sufficient to evaluate
' biological efficacy of potential alternative intake technologies that may require site
specific evaluations and support social cost-benefit analyses as a part of the
§122.21(r)(10)-(12) compliance evaluations; and
4. Collect data to evaluate the entrainment reduction effectiveness of the deep skimmer
wall located at the entrance to the facility's intake cove'.
' While not a primary objective, the entrainment data gathered will help support development of
§122.21(r)(4) which requires a listing of species and life stages most susceptible to entrainment
at the facility.
' To meet these objectives, this document provides summaries of the station's configuration and
operations (Section 2), historical biological sampling efforts conducted at the facility that are
' relevant to cooling water intake evaluations (Section 3), a summary of Threatened and
Endangered Species identified near the facility (Section 4), a sampling program design
justification based on this information (Section 5), the recommended study methods including
' key parameters of gear, schedule, frequency, and quality control procedures (Section 6), the
rationale and methods for an evaluation of the skimmer wall, including historical information
(Section 7), references cited (Section 8), life history information on species likely to be entrained
' Under the final rule,the best technology available for entrainment reduction is determined on a site-specific basis by
' Directors based on several factors including the feasibility and social costs of entrainment reducing technologies
and their potential social benefits. As part of the determination process, facilities may take credit for features at a
facility that reduce entrainment, such as existing fish protection technologies.
Duke Energy 13
1

Entrainment Characterization Study Plan
Marshall Steam Station
' that supports reducing sampling from 12-months per year to 8-months per year (Appendix A),
documentation of the subject matter expert review of an earlier draft of this plan (Appendix B)
and a white paper on the use of pumped samplers in entrainment monitoring (Appendix C).
2 Generating Station Description
This section presents background information on the source waterbody (Lake Norman) from
which Marshall withdraws cooling water and the design and operation of the cooling water
' intake structure.
2.1 Source Waterbody
Marshall is located on Lake Norman, an impoundment of the Catawba River formed by the
Cowans Ford Dam in 1963 (Figure 2-1). The Catawba River Basin lies in the south-central
portion of western North Carolina and originates in the eastern slopes of the Blue Ridge
Mountains in Old Fort, McDowell County, North Carolina. From its source, the Catawba River
flows eastward, then southward toward the City of Charlotte. The Catawba River Basin and the
Broad River Basin join to form the headwaters of the Santee-Cooper River system.
Lake Norman and the Cowans Ford Dam are part of the Catawba Chain of Lakes system
' formed by 11 Duke Energy hydroelectric dams. Construction on the hydroelectric developments
began in the early 1900s with the final hydroelectric development, Cowans Ford on Lake
' Norman, completed in 1963. Lake Norman is the largest reservoir in the Chain of Lakes and lies
between Lookout Shoals Lake upstream to the north and Mountain Island Lake downstream to
the south.
' The surface area of Lake Norman is approximately 32,510 acres, its mean depth is 33.5 feet,
and it drains a 1,790-square mile watershed (NCDENR 2003). Lake Norman is 34 miles long
' from the tailrace of Lookout Shoals Lake downstream to Cowans Ford Dam (NCDENR and
NCWRC 2001). Marshall is approximately in the middle of this distance on what would be the
west bank of the river. Other electric generating facilities on Lake Norman are the Cowans Ford
' hydroelectric development at the 350-MW Cowans Ford Dam and the 2,360-MW McGuire
Nuclear Station in Huntersville, North Carolina adjacent to the Cowans Ford Dam.
1
Duke Energy 14
1

1 Entrainment Characterization Study Plan ���
Marshall Steam Station
I1.
Lake r-
Thekory
VS: '1111111V11 OSvilN
ail
W
Hickory
•
4
=�. onove Ili
INOwton 17
I •Sherrllls Ford
MARSHALL STEAM STATION
MOOMIal -3 Maiden I yOrsh �0 Oesk°O •Mooresrille
t lake Audi Pr
Landis lee ,
`.`,' f Davidson
n
I .1.1 . olrtton
•
an•Cornelius
olls
- 14 i
Ia . *HiroletsvIlle IlikConcord
herryville +. -
• ni
Figure 2-1. Marshall Steam Station Vicinity Map
1 2.2 Station and Cooling Water Intake Description
Marshall has four coal-fired units with a combined electric generating output of 2,110-megawatts
I (MW). Units 1 & 2, each rated at 385 MW began operation in 1965 and 1966, respectively. Unit
3 and Unit 4, each rated at 670 MW began operation in 1969 and 1970, respectively. Marshall
uses a once-through cooling system and was operated between 45 and 65 percent between
I 2011 and 2013 (Table 2-1). The design pumping capacity is 1,463 MGD based on pumping
rates of 190,000 gallons per minute (gpm) each for Unit 1 and Unit 2 and 318,000 gpm each for
Unit 3 and Unit 4.
I2.2.1 Intake Structure
I The single cooling water intake is located at the end of a 1.3-mile long, 200-acre cove (Figure
2.2). At the mouth of the cove is a skimmer wall that is 60-feet deep at full pond and has a 10-
foot by 270-foot opening for withdrawing cool, hypolimnetic water from near the bottom of the
I lake. Water entering the intake cove passes under the skimmer wall, which selectively
withdraws from 60 to 80 feet below the surface. The intake cove is dendritic with a surface area
of approximately 81 ha, and a theoretical retention time of 1.15 days at full station flow rate.
IOn the face of the intake are trash racks with steel bars spaced 3-inch on-center. Unit 1 and
Unit 2 each have three intake bays and two circulating pumps, and Unit 3 and Unit 4 each have
I five intake bays and three circulating pumps, for a total of 16 intake bays and 10 circulating
pumps (Figures 2-3 and 2-4). Typically, one pump per unit is operated during winter months or
during reduced plant loading and more pumps are brought online to meet discharge
IDuke Energy 15

Entrainment Characterization Study Plan
Marshall Steam Station
temperatures during periods of peak demand (i.e., typically July through September). Heated
water is discharged into a 1-mile long cove that is downlake from the intake cove and has a
surface area of 75 acres.
Table 2-1. Design Flow Rate by Unit and Capacity Factor by Unit and Year at Marshall
Steam Station
Design Flow Rate Capacity Factor by Year(Percent)
Unit (MGD) 2011 2012 2013
1 274 43.6 v WP439.Mi
2 274 56.2 41.4 45.0
1 3 458 69.1 56.6 32.4
4 458 70.7 67.4 64.1
Facility 1,464 59.9 49.5 45.3
I
1
I
I
I
I
I
I
Duke Energy 16

M N all MI r r N N r — 1 — all i I — S E 11111
Entrainment Characterization Study Plan LYR
Marshall Steam Station I
\ .._
.air .
V '•
INTAKE r , .. 4,,,.0-.... �.
I ! COVE ����� ,j ' ''t` 4
* ax
fr. r r" SKIMMER " r
I, WALL ,, 4if s" k
111100 -
INTAKE `� R ..rs, �,,'iTrt �i j
r •;� w
1 CANAL f r. , i 1r'
a ...01111111 EXISTING ' ;:*` fi �. ..�^ �F/' a.41.114 9
t
CWIS -----// '4tl` ';,d �;
y :.�§ R,, a { e '6'. al •
, if..
MARSHALL �: 4.
f7 4 i vi alg
STEAM h --- �'`. !',1•41..,
STATION .'?_ ,...� �, -r w ,
DISCHARGE i „;
* 44,:...
,.. , .....
CANAL yr •
_
• ?_
USGS ��' �t esri....
Su U NA1A MGA UtiGS
Source.Esrl,DlgllulGlohe,GeuEye.I tubed.USDA,USES.AIX.Gelrn.,ynung.Aerogrld.'1t,. IGI'. ...t Ione.and the GIS U.er Curnnrun.1y
Figure 2.2. Site Configuration of Marshall Steam Station
Duke Energy 17

Ell Ell Al In E M M 1 M r E O M MB a — r M N
Entrainment Characterization Study Plan EN
Marshall Steam Station
-- 153.3' ►r 93.5'
0 0 .2= -'
_._. -,_-t.. -tet. —' ��— _.. _
II t i E
4 , 1 ; I �
i I - l- J t_. 1 1 J '1.2'
TRASH • — FIXED-PANEL BULKHEAD ' CIRCULATING (TYP.)
RACK(TYP.) SCREEN(TYP,) GATE(TYP.) WATER PUMP(TYP.)
�-- UNIT 4 ►;-r UNIT 3 - i-...— — UNIT 2 --a-+ UNIT 1
0 15' 30'
Figure 2.3. Plan View of Marshall Steam Station Cooling Water Intake Structure (Source: Alden 2012)
Duke Energy 18

— NM In NB MN e /I — EN r MN — NS NM S NM all N r
Entrainment Characterization Study Plan EN
Marshall Steam Station
F
5x
BULKHEAD
i / GATE
TOP DECK EL.7702 GUIDES
_ >--1 \-_.
. \ \
\ •
\ \
\ \ I
\ \ T FULL POND El.760.0'
- - \ \ I
\ \ \
\\ \ , -TRASH RACK
. ' : \ \
\ \ y EXPECTED LOW WATER EL.750.0' _
• \ \ \
\ \ \
\ \ \ Z MIN.DRAWDOWN EL.7450
. \ \ \
\ \ \
\\ \
\\ \
• i i \\ \
FIXED \\ \
• SCREEN \\ \
GUIDES \\ \
L_______‘ \\ \
i
I1 _ , INV.EL.725.8' 1 1 ,\
. , • • • • . r 0 5' 10'
is
Figure 2.4. Section View of Marshall Steam Station Cooling Water Intake Structure (Source: Alden 2012)
Duke Energy 19

Entrainment Characterization Study Plan
Marshall Steam Station
There are 16 sets (front and back) of fixed screens, with each screen consisting of upper and
lower removable panels and with one screen set per intake bay. The screens have 3/8-inch
woven wire mesh and are cleaned individually by removing the screens by crane, placing them
horizontally over a steel debris collection pit that can be lifted by crane and emptied, and
washing them with a high pressure spray wash. Normally the screens are washed on an as-
needed basis, but the screen washing process can be continuous during periods of heavy
debris/fish loading.
3 Historical Studies - Entrainment
Historically, Duke Energy sampled entrainment at Marshall March 1 to August 27, 1976.
Samples were collected from a 2-inch diameter gate valve located on the intake structure
behind pumps of Unit 1 and Unit 4. Water from the gate valves was passed through plankton
nets with 794-pm suspended in 55-gallon drums. Valve flow rates were estimated by measuring
the time required to fill the drum. During sampling, times were measured to the nearest minute.
Sample volumes were derived from the flow rates and sample times. Twenty-four hour samples
were collected three times each week from both Unit 1 and Unit 4. Sample volumes average
461 m3 at Unit 1 and 280 m3 at Unit 4. During sampling, one egg and 19 larval fish were
collected. Larval Yellow Perch (Perca flavescens) accounted for 47 percent of the total
entrainment. Other fish entrained included shad (Dorosoma sp.), White Catfish (Ameiurus
catus), Channel Catfish (Ictalurus punctatus), crappie (Pomoxis spp.). The one egg collected
was not identified to species and collected in April. Larval fish were present in samples from 4-
March to 2-August.
4 Threatened and Endangered Species
There are no State or Federal rare, threatened or endangered species or critical habitat
designations near Marshall. The U.S. Fish and Wildlife's map-based search tool (IPAC) was
consulted to generate a resource report and determine the potential presence of Federally-listed
species within Lake Norman and the surrounding land (Figure 41; USFWS 2015). The only
aquatic species identified was the endangered Carolina Healsplitter (Lasmigona decorate).
However, the Carolina Healsplitter is usually found in mud, muddysand, or muddy gravel
substrates along stable, well-shaded stream banks (USFWS 1996). The habitat near Marshall's
intake is not suitable to Carolina Healsplitter and it is not anticipated to reside anywhere near
the Marshall CWIS.
1
' Duke Energy 110

IEntrainment Characterization Study Plan L��
Marshall Steam Station
I
! ,f 1-1 I i
ftatisitut
it
-.,rh .w
J 1 I l /Nil /
Jt� ill n _�Irin ,0p°� ._.�n N.nn.tt
lMi1 t
MI•/I
, { 4r. I I
:ft -I I
I .Illii '.
...
CI
i� H.rn l..e i
1 'II. I1
643s+:mrs1 1. n I , MIFi Oa }i
. ii •
PIry.dJ
l .•vl
i „ \:
Figure 4-1. Geographical Boundary of the IPAC Search
1
5 Basis for Entrainment Sampling Design
IHDR Engineering, Inc. (HDR) and Normandeau Associates, Inc. (Normandeau) participated in a
site visit to Marshall on April 22, 2015 to evaluate potential entrainment sampling options at the
I CWIS and to determine if pumped samples from within the CWIS would be a practicable
sampling method based on best professional judgment and previous experience with
entrainment sampling. Sampling at the intake with a pumped sampler minimizes damage to or
I loss of organisms that can occur if samples are collected at the discharge side of the condenser
cooling water system. In addition, properly designed and operated pumped systems have
shown collection efficiency of 95 percent or greater for fish eggs and larvae with little or no
Iorganism damage (EPRI 2014). In addition, sampling the discharge at Marshall would be
impractical because it enters a wide canal below the surface.
IITwo primary methods that have been historically used to estimate ichthyoplankton entrainment
at power plant intakes are utilizing streamed/towed nets and collecting pumped samples.
Traditional ichthyoplankton nets can be used to filter water as it enters the intake and collect
Iorganisms. Alternatively, pumps can be used to convey water from the intake structure to a fine-
mesh net onshore. Onshore nets are suspended in a buffering tank to minimize damage and
Iextrusion of eggs and larvae.
Each method has advantages and disadvantages and a comparison of the two methods are
Isummarized in Table 5-1. Pumped sampling was selected as the preferred sampling method for
IDuke Energy 111

Entrainment Charaderization Study Plan L1�
Marshall Steam Station rJ
Marshall. The primary advantages of utilizing pumps at this location include metering precise
sample flows, longer sample collection times, reduce the potential to miss samples due to
inclement weather or other events, and increase the ability for technicians to safely observe net
filtering and other aspects of the data collection. Pumped samplers are among the preferred
gear types accepted by the EPA and have been used extensively to successfully monitor
entrainment at power plant intakes for decades. Their versatility includes being utilized in fresh,
estuarine, and marine water environments. Properly designed and operated systems can be
accurate and effective. While no sampling method is perfect, we believe pumped samplers offer
the best, most cost-effective, and consistent sampling method available for Marshall.
1
Duke Energy I 12

Entrainment Characterization Study Plan
Marshall Steam Station
Table 5-1. Advantages and Disadvantages of Hoop Nets and Pumped Samplers for Estimating lchthyoplankton Density in Cooling Water Intake Structures (some information adapted from EPRI 2014)
Gear Type Advantages Disadvantages
Hoop Nets Deployed in the Intake -Large volumes are sampled quickly(less manpower required for the same number of pumped samples). -Can be difficult to deploy and retrieve in the confined space of intake structures—precludes the use of some
1
- If net frames are not used,then there is limited to no modifications to the intake required for deployment. net types(e.g.,standard bongo, neuston nets,or Tucker trawls).
-No potential for mechanical damage associated with pump passage. -Depending on deployment method, may require modifications to intake structures(e.g., frame-mounted nets in
frame guides).
-Less precise flow metering than pumped samplers.
-Large volumes are sampled quickly—capturing less temporal variability as compared to pumped samples.
- Relatively small nets needed to fit in the intake structure offer a small spatial sample. Multiple nets can be used
to increase sampled area at the cost of additional samples to be processed in the laboratory.
-Tow speeds in the range of 1-2 meters per second(commonly used during ichthyoplankton sampling)is above
the intake velocities at the majority of intakes.
-Some active avoidance possible by larger motile life stage(e.g., late larvae and early juvenile). Larger hoops
can be used to decrease potential for avoidance,but would require a larger deployment area, since length is
proportional to opening diameter in properly sized nets2.
-Greater potential for extrusion than pumped samples(no buffering tank).
-Boat deployed nets are subject to weather delays and associated safety concerns.
Pumped Samplers in the Intake -100 m3 samples are collected over roughly 2 hours increasing the potential to capture temporal variability in -Some active avoidance possible by larger motile life stage(e.g., late larvae and early juvenile).
ichthyoplankton densities not observed in net samples. Im ro eri desi ned sam lens can lead to Barna a to organisms Burin sam lin
P P Y 9 P 9 9 9 P g•
-Limited modifications to intake structures are required to install—usually just anchoring points for the sample -Samples a smaller portion of the spatial variability, because pump inlets are generally smaller than net
pipe.
openings.
- In-line flow metering offers greater precision in measuring the volumes of flow sampled.
-Some potential for mechanical damage. However, correctly designed systems can offer<5%damage or
destruction of eggs and larvae.
' -Fixed pipe allows precise control over water depth and orientation to intake flows.
-Less potential for extrusion than nets, because the filtering net sits in a buffering tank.
-Lower potential for missed samples due to severe weather.
-Allows technicians to observe sample collections and minimize potential for invalid samples(e.g.,use of 330-
um nets increases potential for net occlusion and frequent net change outs may be required during certain times
of the year).
1
I
I
1
2 A general rule of thumb,as described in EPRI 2014,states that the total effective open area of the netting(percent open area x area of netting)should be at least twice the area of the net mouth opening. Others have suggested a net length to mouth opening diameter ratio of three or more.

111 Entrainment Characterization Study Plan
Marshall Steam Station
While all sampling techniques are biased to some degree, the design for the Marshall CWIS is
ideal for pumped entrainment sampling. Potential disadvantages to pumped entrainment
sampling at intake structures (compared to sampling at the discharge) have been studied by
' EPRI (2014). An explanation of how these potential disadvantages are minimized at Roxboro is
provided in Table 5-2. Despite some potential disadvantages, pumped samples collected at the
intake structure remain a better option than sampling at the discharge because:(1) organisms
will be less damaged compared to those passing through the cooling water system to the
discharge (resulting in a higher probability of taxonomic identification); (2) access to the intake
structure is easier logistically; (3) lower velocities at the intake structure will result in less
extrusion of larvae and/or damage to nets; and (4) safety issues and inclement weather will not
be major factors resulting in lost sampling dates.
Table 5-2. Potential Disadvantages of Pumped Ichthyoplankton Sampling at Marshall
Steam Station
Potential Disadvantage
(as described in EPRI 2014) Sampling at Marshall
Non-random vertical distribution may Sampler at Marshall will be designed to sample near surface,
require depth-stratified sampling depth, and near bottom simultaneously(see Section 6)
Low water velocities at some CWIS'
my increase likelihood of active gea take velocities at Marshall are less than 1.0 feet per second (fps)at
avoidance for more motile larval an a point of sampling and some avoidance may occur.
juvenile stage - -
Potential uncertainty as to whether Only the late larvae and early juvenile fish might escape entrainment
organisms were destined to be from downstream of the bar racks where velocities are greater than
entrained 0.5 fps.
The recommended approach for Marshall (described in greater detail in Section 6) is to pump
' water from the CWIS between the trash racks and the fixed panel screens to an entrainment
sampler. The sampling system will utilize a fixed pipe with three orifices that will allow
simultaneous withdrawal from near surface (- El. 7453), mid-water (- El. 737.5), and near
bottom (- El. 730). Two identical pipes will be installed at Unit 3 and Unit 4 so that if one unit is
not in operation on any given sampling date, samples can be taken from the other. These
locations are near the mid-point of the intake structure and will sample the Units that have the
highest capacity factors resulting in samples that are representative of plant operation (see
Table 2-1). Both pipes will be equipped with quick-connect couplings to allow either pipe to be
sampled utilizing a single pump. The quick-connect couplings will also allow above deck piping
' to be removed between sampling events to facilitate station personnel cleaning the fixed panel
screens.
111 RECEIVEDINCDEQIDWR
3 Elevations in this document refer to Mean Sea Level MAY 31 2016
Water Quality
Permitting Section
' Duke Energy 114

1
Entrainment Characterization Study Plan
Marshall Steam Station
' Entrainment sampling surveys will be conducted twice per month between March 1 and October
31 in 2016 and 2017. This period is expected to correspond to when fish eggs and larvae would
be present in Lake Norman based on spawning characteristics of the species most likely to be
entrained (see Appendix A). However, the sampling program will be run adaptively in response
to entrainment densities. For example, if the densities of entrainable organisms remain high in
' the October 2016 sampling, additional sampling events will be added to the program in 2016
and the sampling period in 2017 will be extended accordingly. Similarly, if densities of
entrainable organisms are high in early March 2016 when the program is initiated, then the 2017
' sampling plan will be expanded to begin earlier (e.g., February). In this way, the adaptive
management plan will provide the greatest potential to collect representative samples
throughout the entrainment season.
Each sample collection event will be conducted over a 24-hour period with sample sets
collected in four 6-hour intervals. The sample frequency selected for this entrainment study will
provide fish taxa, density distribution, and seasonal/diel variation data over a two year period.
The approach for development of the specific entrainment characterizations required in
' §122.21(rX9) based on the collected data is summarized in Table 5-3.
1
1
1
' Duke Energy 1 15

IEntrainment Characterization Study Plan L)�
Marshall Steam Station r
I
Table 5-3. Summaryof Approach for Development of§122.21(r)(9) Required Entrainment
pp P
Characterizations
I
122.21(r)(9) Requirement Basis for Meeting the Requirement
Two years of data and annual Entrainment samples will be collected during 2016(Year 1)and
variation 2017(Year 2)
Seasonal variation Entrainment samples will be collected twice per month during March
I through October each year
Diel variation Each 24-hour sampling event will be split into four 6-hour sampling
periods to capture diel variation
I Variation related to climate and Weather information and water temperature will be collected during
weather each sampling event to evaluate differences in entrainment rates
based on these factors
I
Year 1 and Year 2 data will be analyzed to determine species/life
Variation related to spawning, stage variations over time along with spawning and feeding variation;
feeding and water column Entrainment samples will be collected at three depths(near surface,
I migrations mid-depth, and near bottom) to account for depth variability by
species/life stage for water column migrations
he resolution of taxonomic and life stage designations will be
Identification of lowest taxon onitored through regular evaluations of catch data with the goal of "
I
;possible clueing percent of unidentified organisms and increasing resolution
of genera and higher taxonomic designations
Data must be representative of each Sampling in Unit 3 and Unit 4 are expected to be representative of
intake the total CWIS
ow t'e oda tori o he in ake in e It
ampling of near surface, mid-depth and near bottom at downstream
water body are accounted for f the bar racks assumed to be best method for accounting for intake
location1 `
Document flow associated with the Facility will monitor flows for period of sampling for use in the final
data collections report produced after sampling
IMethods in which latent mortality will Latent mortality will not be evaluated as a part of the study and
be identified therefore methods are not provided
I Data must be appropriate for a Data will be expressed as taxon and life stage specific densities
quantitative survey which can be multiplied by flow to support quantification of
entrainment
I
6 Entrainment Characterization Study Plan
1 6.1 Introduction
I This section of the ECSP provides methods, materials, and procedures for entrainment sample
collection and processing. Any failures at the sampling or laboratory analysis stage are often
uncorrectable because design-specified sampling times cannot be repeated once they have
Ipassed. Therefore, Standard Operating Procedures (SOPs) including a Quality Assurance (QA)
Plan will be developed by Normandeau, who will be performing the field studies for the
I
IDuke Energy 116

Entrainment Characterization Study Plan
Marshall Steam Station
entrainment sample collection and processing based on the ECSP and its preferred methods,
datasheets and equipment to eliminate, reduce, and/or quantify those errors.
Adherence to sample collection and lab analysis SOPs will be observed and documented twice
during sampling and once in the laboratory during each year of sampling as described below.
' Quality Control (QC) procedures such as re-inspection of sorted samples in the laboratory and
documented training and adherence to the project SOP (field and lab) will be ongoing
' procedures controlled by project personnel. A QA officer, who is independent of those
individuals collecting and generating the data during the study and has experience in performing
QA/QC programs for aquatic monitoring surveys, will conduct technical assessments/audits, to
' ensure that QC procedures are being followed and documented, and proper procedures are
used for data collection. The specific requirements are to be developed and included in the
SOP, will incorporate a checklist of items to be inspected based on the SOPs, and will include
observations relevant to performance of sampling that may not be covered by the SOP. Careful
attention will be paid to the initiation of the study when staff may be less familiar with the SOPs.
1 6.2 Sample Collection
Entrainment samples will be collected twice per month from March 1 through October 31 in
' 2016 and 2017 (16 sampling events in each year). This frequency should be sufficient to
capture the annual, seasonal, and diel variability in entrainment with acceptable confidence
levels (inferred from EPRI 2014). Based on life history data of species likely to be entrained at
Marshall (see Appendix A), this window should collect the vast majority of entrainable-sized
organisms present in Lake Norman. The proposed monitoring program will start in March to
collect the earliest entrainable life stages (e.g., eggs) and will continue through the end of
111 October when spawning activity is expected to be completed. If no circulating pumps are
scheduled to operate during the specified sampling period, a request will be made to turn on a
' circulating water pump for the duration of sampling in order to get representative density
measurements. This twice per month sampling frequency should be sufficient to adequately
describe seasonal patterns in entrainment as requested in the final §316(b) rule for existing
' facilities.
At each 24-hour sampling event, the Unit 3 or Unit 4 intake will be sampled (depending upon
which unit is in operation) within the following discrete 6-hour time intervals: 2100-0300 (night),
0300-0900 (morning), 0900-1500 (day) and 1500-2100 hours (evening). During each 24-hour
' sampling event, collections will be taken within each 6-hour period resulting in four samples on
each sampling day. In the corpuscular periods, target sample collection times will be 1 hour
' preceding and 1 hour following sunrise and sunset. A total of 64 samples will be collected during
each year over the entrainment season for a program total of 128 for the two years of study
(Table 6-1)
Duke Energy i 17

Entrainment Characterization Study Plan
Marshall Steam Station
I
Table 6-1. Entrainment Sampling Details
Details
' Units to be Sampled Unit 3 or Unit 4
Sampling Events(Days) Thirty-two(32)sampling events; twice per month; March 1 and
October 31, 2016 and between March 1 and October 31, 2017
Daily Collection Schedule Samples collected within every six hours in a 24-hr period (4
collections/24-hr period)
Targeted Organisms Fish eggs, larvae, and juveniles
' Depths Depth integrated sample using selective withdrawal from near
surface, mid-depth, and near bottom.
Sample Duration Approximate 2-hour samples collected within each 6-hour sampling
interval.
Number of Samples per Sampling Four samples per sampling day
' Event(Day)
Total Number of Samples Sixteen (16) sampling events/year x 4 samples/sampling event
(days)x 2 years= 128 samples
1 6.2.1 Location
Entrainment samples will be collected from either the southern-most pump bay of Unit 3 or
northernmost pump bay of Unit 4, which are immediately adjacent to one another. Each intake
pipe will be installed through the intake deck grate midway between the bar racks and the fixed
screens and in the middle of the intake bays. If both pumps are operational on a given sampling
date, the Unit 3 pump will be selected preferentially, because it is more centrally located within
the CWIS. Samples will be collected from between the trash racks and the flat panel screens at
' near surface (- El. 745.0), mid-water (- El. 737.5), and near bottom (- El. 730.0) depths (See
Figures 6-1 and 6-2). Piping inlets at near surface, mid-depth, and near bottom will be sized to
allow concurrent equal flow from each depth, thereby contributing to a single aggregate sample.
The pump and buffering tank will be located on the concrete deck between the two units (Figure
6-3). Changes or variations in the sampling location over the duration of the 2-year study will
' require Duke Energy notification and approval.
I
1
1
I
Duke Energy 118

= = N IIIIIII MI I MI M E — — — NM M MO — M i ME
Entrainment Characterization Study Plan FN
Marshall Steam Station
Approximate Sampling
Locations
-- 1 Hr:
4-1
'.
1 '• �: `'✓
b "7710.;
-e
1 o o ..
tNi ,.,..
.,. v .. -5;'
- - -...Mi=3-. .r 4 -,3 :' 1111.1! 111111.111 .
' - _..=.._......_
s
OEM-
�. t. .. .. .. ..I'•i . -'.4 1._. .. t.. .. . ..
•
.. 11.2'
TRASH 1— FIXED-PANEL BULKHEAD I CIRCULATING (TYP.)
RACK(TYP.)— SCREEN(TYP.) GATE(TYP.) WATER PUMP(TYP.)
UNIT 4 UNIT 3 —— UNIT 2 — UNIT 1
0 15' 30'
Figure 6-1. Plan View of the Marshall Steam Station's Cooling Water Intake Structure with Approximate Sampling Locations
-Sampler not to Scale (Image Modified from: Alden 2012)
•
Duke Energy 119

IEntrainment Characterization Study Plan EN
Steam Station
I is
hcc
' GATE
TOP DECK EL.770A BULKHEAD
GUIDES
LJj 1 - — \1 ( PULL POND Fa 7e0.0'
\\
1\ \
- \, \\ \ -TRASH RACK
Near-surface inlet(745 ft• \\
dr
I
\C \ EXPECTED LOW WATER EL.7500
\\ \
\ \ z Mit DRAWDOWN EL.7458
I
I Mid-depth Inlet (737 5 ft) ! I i \\\ \
\
\\ \
\\ \
I
Near-bottom Inlet (730 ft) \�11
\\ •
��
Direction of Flow
' FIXED
SCREEN
/ GUIDES \\ \
111 , 1j W.a..raiz \tk `\
�• 0 J I
Figure 6.2. Section View of the Marshall Steam Station's Cooling Water Intake Structure
I with Approximate Location of Sample Inlets at Three Depths — Sample not Shown to
Scale (Image Modified from: Alden 2012)
I
I
I
I
I
I
I
I
Duke Energy 120

1 Entrainment Characterization Study Plan
Marshall Steam Station F)�
1
. N., 1 44 ,
. 1 y.
I r
4 ,,,IV
il
1"11 - ..4,'i';:::.'4 t'.':: 4 1' : ' 'it 44,:-'--1:-4 ' '
li
Location of
Sampler and ' ,,,
Pump (approx.) :.
11_::________,:,_- , ,,
i
I
I
• Duplicate
I
Temporary
lexible Lines
from Sample ,,
I
Locations to Pq�
sampler (approx.) `leo
1
I . •
I9 .. - .tel': fid .
I
Figure 6.3. Deck Level View Showing Approximate Locations of the Sampling Gear
(Image Modified from: Bing Maps)
During each six-hour sampling period, one sample, aggregated by depth, with a target volume
1 of 100 m3 will be collected such that on each sampling day, four discrete 6-hour samples will be
collected at Marshall. The volume sampled will be estimated using an in-line flowmeter.
1 Depending upon pump flow rates, this sample will require approximately 2 hours to collect with
IDuke Energy 121

' Entrainment Characterization Study Plan
Marshall Steam Station r L��
additional time required to wash down nets and prepare the samples for shipping. These
samples will be processed discretely to investigate diel variability in ichthyoplankton composition
' and abundance.
Withdrawn water from the PVC sampler will be filtered through 330 micron (pm) plankton nets
' suspended in a water-filled tank to reduce velocity and turbulence and prevent extrusion of
larvae through the mesh. A larger mesh (e.g., 505-pm) may be used if net clogging precludes
sampling with 330- pm mesh. An example ichthyoplankton sampler is shown in Figure 6-4.
IThe net mouth will be suspended above the water line in the tank to prevent overflow and loss
of organisms in the event of tank overflow. The net will be washed down to concentrate
' ichthyoplankton into the attached collection cup once 100 m3 of pumped water is filtered. If high
debris buildup leads to net dogging then more frequent net washdowns may be required. A
second net and collection cup will be on-hand in the event a collection is required prior to
I
pumping 100 m3 of water. In such an event, the net and collection cup will be replaced with the
backup net before rinsing and preserving the collected samples, which will minimize lost
sampling time. The collection cup will be carefully rinsed into sample jars with preprinted labels
'
and preserved in 5-10%formalin solution.
Total sample volume, total sample duration, intake water temperature, dissolved oxygen, pH,
'
and conductivity will be recorded on pre-printed field data sheets. Samples will be transported
back to the laboratory for analysis under the Chain of Custody to be described in the SOP.
i
r
I
JOINT MUST SWIVEL 4—
APPROX.90"
(POSSIBLY MORE) SAMPLE FLOW RATE-240 gpm
/ 3'0 ADAPTER SOCKET
' WOODEN CRADLES(lyp.2) -- \ _I. I 3'O OVERFLOW DRAIN
-. �■�\1\� SI--'-�f i
I
STAINLESS STEEL BANDS(typ.2)Jr
_ \,\ 330P ICHTHYO-NET
/
I
INLINE \ / I
FLOW TOTALIZING 3'0 PVC VALVE /
METER , 3'0 PVC
(PVC SADDLE MOUNT) , � `—"(PASSIVE DISCHARGE)
330p t- 110 gal
I COD END -1. U POLYETHYLENE I { l
' 1 BUCKET . TANK h^'
FLOW IN MI — ./ l ' l� _
3i'OPVC,NIV61. �'> k 3(
' .I/ 1/{�CNNQOE THROdG6 N )
ri
3'O RADIAL FLEX HOSE-" `3'0 QUICK-CONNECT 3'O PVC VALVE
NOT TO SCALE
IFigure 6.4. Example Entrainment Pump Sampling System Configuration
I
I
Duke Energy 122

Entrainment Characterization Study Plan
Marshall Steam Station
1 6.3 Sample Sorting and Processing
Upon arrival in the laboratory, all ichthyoplankton samples will be logged on an lchthyoplankton
Sample Control Sheet/Sorting Form. Because Lake Norman is a freshwater reservoir, we do not
anticipate shellfish larvae, as defined as commercially important crustaceans or bivalves, will be
' present in the samples. Therefore, procedures for identifying shellfish larvae are not presented.
Before sorting, ichthyoplankton samples will be rinsed using a U.S. Standard Sieve with a mesh
size opening of less than or equal to 330 pm to remove excess detritus and formalin. All fish
eggs and larvae retained on the sieve will be hand-sorted from the debris with the aid of an
illuminated magnifier. Samples that are estimated to contain more than 400 fish eggs and larvae
' (all taxa combined) will be split with a plankton splitter and to a subsample quota of about 200
eggs and larvae combined and then analyzed. The number of eggs and larvae present in the
sample will be recorded. If possible, long-dead and/or non-viable eggs will be identified using
appropriate and well-defined techniques identified in the SOP and categorized in the database
accordingly. For example, the SOP may require that eggs collected live be whole, show signs of
fertilization and not be covered with fungus at the time of their entrainment. lchthyoplankton
from each sample will be placed in individually labeled vials and preserved in 5 to 10 percent
formalin prior to taxonomic analysis. Examples of organisms identified as long-dead or non-
viable will be stored in separate vials.
Fish eggs, larvae, and juveniles will be identified using a dissecting scope equipped with a
polarizing lens. Identifications will be made to the lowest practical taxonomic level using current
references and taxonomic keys (e.g., Auer 1982, Wallus et al. 1990, Kay et al. 1994, Simon and
Wallus 2004, EPRI Larval Fish Identification Key http://www.larvalfishid.com/). Larvae and
' juveniles will be categorized as follows:
• Yolk-sac larvae: Phase of development from the moment of hatching to complete
absorption of the yolk;
• Post yolk-sac larvae: Phase of development from complete absorption of yolk to
development of the full complement of adult fin rays and absorption of finfold;
• Juvenile: Complete fin ray development and finfold absorption;
• Eggs: Are required to be whole, show signs of fertilization, and not be covered with
fungus to be considered live.
Ichthyoplankton larval life stage will be identified as "larvae" if they are damaged to the point
' that they cannot be confidently classified as yolk-sac or post yolk-sac.
For each diel (6-hour)sample, the following morphometric data will be collected:
I • Up to 10 yolk-sac, post yolk-sac and "larvae" of each fish species will be measured for
total length, greatest soft tissue body depth, and head capsule depth to the nearest 0.1
mm. Among dorso-ventrally compressed organisms whose body or head capsule width
' exceeds the body or head capsule depth, soft tissue body and head capsule width will
also be measured to the nearest 0.1 mm.
Duke Energy 123

Entrainment Characterization Study Plan
Marshall Steam Station
• Up to 10 eggs of each taxon will be measured for minimum and maximum diameter to
the nearest 0.1 mm.
1 Only whole organisms will be subject to morphometric evaluations. If more than 10 eggs or
larvae are present, a random subset of each species and life stage will be measured. Length
measurements will be performed with a calibrated ocular micrometer or other calibrated tool
(e.g., ImageToolTm Software).
Organism identification will be cross-checked using the QC procedure described below.
6.4 Data Management
' Field and laboratory data will be recorded on forms compatible for computer entry and data
processing activities will be recorded on log sheets for each batch of data. A digital image of the
datasheet will be taken in the field prior to the datasheets leaving the site. Data sheets will be
inspected for completeness prior to data entry. The data will be entered using a double data
entry software, a feature that ensures entry errors will be caught and corrected as the operators
key the data. Using this procedure, data sets are entered twice in succession and the software
compares the first and second entries and identifies any discrepancies. Discrepancies must be
resolved before the second data entry can continue. Keyed data sets will be error checked. If
any errors are encountered, they will be corrected in the database. Once the database is
cleared of errors, the data file compared to the data sheets will be audited to ensure an Average
' Outgoing Quality Limit (AOQL) of 1 percent (t 99 percent accuracy). The data set will then be
ready for the production of summary tables, which will be proofed to confirm that the summary
program worked properly. The data editor will sign and date each proofed summary table and
include notations as to which values were verified.
6.5 Entrainment Data Analysis
Data analysis will be performed using the QAQC'ed database and will include summaries of
proposed vs actual samples collected, sample volumes, entrainment densities, morphometric
measurements, and water quality parameters. Generally, minimum, average or median, and
maximum values will be provided by sample event or month. Collection densities, expressed as
number per 100 m3, will be calculated from entrainment catch data for each taxon and life stage
by month of sampling, sample event (i.e., including all samples collected within a 24-hour
period), and by six hour diel intervals (e.g., 2100-0300 [night], 0300-0900 [morning], 0900-1500
[day] and 1500-2100 hours [evening]) across all sampling events. Average concentration of
' organisms per unit volume in the he' stratum (i.e., month, sample event or six hour interval),II, ,
will be calculated as:
1
Duke Energy 124

Entrainment Characterization Study Plan 1.01Marshall Steam Station
1 n
Xh = hi
where:
nh = the number of samples in the hal stratum
xh, is the ith observation in the hth stratum.
Next, these densities will used to estimate the total entrainment at cooling water intake based
on design and actual intake flows. The estimated total entrainment will be calculated in the
following manner. First, the average concentration of organisms per unit volume in the hth month
will calculated using the equation provided above. The total number entrained (E) during the
sampled months will then then calculated as:
E_ �Vn xn
n=t
where:
H total number of months sampled
Vh = volume of water withdrawn by the station in the hth stratum.
1
6.6 Field and Laboratory Audits
Prior to the first scheduled sampling, an experienced senior staff member will accompany and
train field personnel, including: protocols for site access and contact with facility personnel;
safety requirements as contained in the Health and Safety Plan, implementation of the field
SOP including the operation of the pump samplers, sample collection, sample preservation,
proper datasheet documentation, chain of custody, and shipping. At this time, a readiness
review will be conducted to ensure that trained personnel, required equipment, and procedural
controls are in place. In addition, equipment will be tested to ensure its proper operation.
At the initiation of sampling, two trained QA staff members (one each from Normandeau and
HDR) will conduct an independent QA audit to ensure that the SOP is being implemented
correctly. Results of the audit will be summarized in a technical memo. This memo will
categorize deviations from the SOP into three categories: (1) those that do not affect the quality
of the data, (2) those that may affect the quality of the data, and (3) those that definitely affect
the quality of the data. Variances from approved procedures will be documented and corrected,
either by modifying the SOP to address systematic problems or by testing and/or retraining staff,
as necessary. Any changes to the SOP will be discussed with and agreed upon by Duke Energy
' representatives before implemented in the field. Partway through the sampling program a
' Duke Energy 125

Entrainment Characterization Study Plan L1�
Marshall Steam Station rJ
trained QA staff member from HDR will conduct an independent QA audit following the same
procedures to provide on-site training, to observe sampling activities, and to verify that the
project's SOP is being followed. In addition, senior staff will observe initial laboratory and data
management activities to verify the same.
Implementation of the laboratory SOP will be overseen by a senior Normandeau laboratory
manager who will also ensure that staff technicians have been properly trained. Once during
each year of sampling, an independent QA trained HDR employee will conduct an audit to
ensure that the SOP is followed. These audits will include safety, sampling procedures, sample
processing, and identification. Audit reports will be prepared and any substantial shortcomings
identified will be addressed as soon as practical. Samples from the first collection event will be
analyzed prior to the initiation of the second event to ensure that organisms are being collected
with limited damage to allow identification.
6.7 Laboratory Quality Control
Quality control methods for split, sort and identification of ichthyoplankton will be checked using
a continuous sampling plan (CSP) to assure an AOQL of 10 percent (n90 percent accuracy).
Identification checks will be inspected using a QC procedure derived from MIL-STD (military-
standard) 1235B (single and multiple level continuous sampling procedures and tables for
inspection by attributes). Detailed methods for quality control will be provided in the SOP
developed by Normandeau. The QC checks will be recorded on appropriate datasheets and
' these records will be maintained for review. Samples will be stored for a minimum of three years
after the end of the project or longer if Duke Energy requests additional storage time.
1 6.8 Reporting
During the study, monthly progress reports will outline the status of the on-going sampling and
laboratory processing. At the end of the first year of study, a report describing preliminary
results of testing will be provided to Duke. This first-year report will include entrainment
estimates by month and diel period using design intake flow. At the completion of the study, a
comprehensive report describing all aspects of the study program (facility description, study
design, sampling methods, data analysis methods, and results) will be generated. The final
report will include all tables, figures, photographs and engineering drawings as necessary to
fully document the evaluations conducted. Included will be estimates of entrainment by species,
life stage, month, and diel period under design and actual intake flows. The report will be
organized with supporting information and detail in attached appendices, as needed.
6.9 Safety Policy
All work performed under the direction of Duke Energy on Duke Energy properties and/or on
properties owned or operated by third parties (i.e., not owned or operated by the contractor or
' Duke Energy) will be performed using safe work practices that are at least equivalent to those
required for Duke Energy personnel and of any third party owner or operator. At a minimum, all
contractors are expected to be aware of, and adhere to, Duke Energy's Corporate Safety Policy,
and other location-specific safety policies and procedures.
' Duke Enengy 126

Entrainment Charaderization Study Plan
Marshall Steam Station
1 7 Evaluation of the Skimmer Wall to Reduce
Entrainment
7.1 Introduction
Marshall utilizes a deep skimmer wall intended to selectively withdraw cooler water from below
the Lake Norman thermocline (approximately 60 to 80 feet deep). The original intent of this
structure was to take advantage of the cooler water for more efficient condenser cooling and to
reduce discharge temperatures to meet thermal limits. However, this skimmer wall at Marshall
forms a barrier to ichthyoplankton and is believed to result in substantial reductions in
entrainment.
Historical studies were undertaken at Marshall to demonstrate the skimmer wall reduces the
passage of ichthyoplankton (as described below). These studies showed that the skimmer wall
could substantially reduce ichthyoplankton available for entrainment. That said, Duke Energy
' has elected to gather additional data to supplement these historical studies for several reasons.
First, species composition in Lake Norman may have changed since the historical studies were
conducted in the late 1970s and it is important to demonstrate that the skimmer wall remains
' effective with the species present currently in the reservoir. Second, monitoring techniques have
improved and may provide a better estimate in entrainment reduction that is achieved by the
skimmer wall. Third, new studies will provide a better estimate in entrainment reduction afforded
' by the skimmer wall.
7.2 Historical Studies
The effectiveness of the skimmer wall at Marshall was determined by measuring densities of
ichthyoplankton upstream of the skimmer in Lake Norman and comparing them to towed
ichthyoplankton sampling downstream of the skimmer wall. In addition, these densities were
compared to larval fish collections made by the North Carolina Wildlife Resources Commission
(NCWRC) in the intake, discharge and two control coves (Miller and DeMont 1974; as cited in
Olmsted and Adair 1981).
7.2.1 Methods and Materials
Entrainment samples were collected from March through August 1976 using a 5.1-centimeter
(cm) (2-inch) diameter gate valves in the waterbox upstream of the main condenser at Unit 1
' and Unit 4 (as described in Section 3). When the valve was opened water passed through a
794-pm mesh plankton net suspended in a 208-liter drum filled to minimize mechanical damage
to entrained ichthyoplankton. Water flow rates, which approximated 16 cubic meters per hour,
were determined by measuring the time required to fill the drum. Twenty-four, hour samples
were collected three times per week throughout the sampling period.
Larval fish densities in Lake Norman were estimated by towing a 0.9-m circular net with 3-foot
hoop net with 794-pm mesh. Night trawls were replicated at the surface and 5 meter (16 foot)
depths upstream of the skimmer wall. Densities (number/1,000 m3) at surface and 16 foot
Duke Energy 127

IEntrainment Characterization Study Plan L��
Marshall Steam Station 1
Idepths were averaged to get a monthly mean density upstream of the skimmer wall. All trawling
was done at night with replicates at surface and 5-meter depths outside the skimmer wall
I Catches were expressed as number per 1,000 m3 and densities at the surface and 5-meter
depth were averaged to give monthly mean density lakeside of the skimmer wall. Trawling was
performed March through August 1975.
IIn addition, NCWRC collected eggs and larvae March through July 1971 using a 28.5-cm
diameter circular net with 570-pm mesh at the intake, discharge, and two control coves. The net
I was towed for 15 minutes at each station from the surface to 2 meters (6.5 feet). Crappie
(Pomoxis spp.) s15 mm length were considered larvae and included in the analysis. For all
other taxa, organisms 5. 21 mm were considered larvae. Organisms larger than these criteria
Iwere considered juveniles and not included in the analysis.
7.2.2 Results and Discussion
IEntrainment was low at Marshall. Monthly catch densities ranged from 0.0 (June) to 0.5 (April)
larvae/1,000 m3 (Table 7-1). Entrainment samples were limited to crappie (Pomoxis spp.; 5
I larvae); shad (Dorosoma spp.; 2 larvae); and Channel Catfish (; 1 larva). These densities were
compared to mean monthly estimates in the top 5 m (16 feet) upstream of the skimmer wall that
ranged from 15.0 (April) to 677.2 (June) larvae/1,000 m3 (Table 7-1). Ichthyoplankton upstream
111
of the skimmer wall was dominated by Yellow Perch. The finding that the skimmer wall is
reducing the available ichthyoplankton for entrainment was bolstered by Miller and DeMont
(1974; as cited in Olmsted and Adair 1981) who observed densities of larvae in the Marshall's
1 intake cove to be 5 percent of those in upstream and downstream control areas.
Table 7-1. Mean Monthly Densities (larvae/1,000 m3) of Larval Fish in the upper 5 meters
I (16 feet) of the water column upstream of the Marshall Steam Station skimmer wall, April
through September 1975
I . Pomoxis Lepomis ,earEntrainment
Month Clupeidae Percidae spp. spp. Other Total Total'
April 0.5 13.3 0.5 0.0 1.2 15.5 0.49
I
May 269.1 4.5 8.5 1.6 6.0 289.7 0.09
June 669.1 0.2 0.0 4.3 3.6 677.2 0.00
July 623.2 0.0 0.0 0.0 0.9 624.1 0.24
1 August 661.2 0.0 0.0 7.4 0.0 668.6 0.28
September 37.4 0.0 0.0 0.0 0.0 37.4 --2
Mean densities (larvae/1,000 m3) of larval fish of all taxa collected in entrainment samples
collected in the same months (1976) are shown in last column. Modified from Olmsted and Adair
1981.
I 2-no data collected
It appears that the data from the 1970s support the hypothesis that the skimmer wall at Marshall
reduces entrainment; however, these studies are not without potential criticism. First,
Ientrainment was sampled using 794-pm mesh nets, which may not have collected smaller eggs
and larvae. The current standard for entrainment monitoring is around 330-pm or slightly larger
1 in locations with high debris loading. In addition, the water volume sampled was estimated using
IDuke Energy 128

Entrainment Characterization Study Plan ���
Marshall Steam Station
estimated flow rates and sample times rather than measured by gage. Second, the
ichthyoplankton entrainment data were collected in a different year than the net tows upstream
of the skimmer wall. Inter-annual spatial and temporal variability in ichthyoplankton density as a
result of larval production, dispersal, and mortality is a common phenomenon and some of the
observed reduction in larval density in entrainment sampling compared to upstream density may
have been a result of this variability. Because of these shortcomings, caution should be used
with these data when quantifying the reduction in entrainment that the skimmer wall can
achieve. Instead, these data show that the potential for high entrainment reduction exists and
further study is warranted.
' 7.3 Proposed New Skimmer Wall Evaluations
Ichthyoplankton sampling of Lake Norman will occur upstream of the skimmer wall concurrent
with entrainment sampling at the CWIS using a Tucker trawl as described below. Entrainment
sampling at the CWIS is described in detail in the ECSP (Section 6). Samples collected during
the skimmer wall evaluation will be sent to Normandeau' s laboratory for processing. Sorting,
' identification, and measuring of eggs and larvae will follow the same procedures as those
outline in Section 6 of the ECSP and subject to the same data management and quality control.
7.3.1 Sampling Location and Frequency
lchthyoplankton samples will be collected at two perpendicular transects located on either side
of the skimmer wall (i.e., upstream and downstream) (Figure 7-1). Samples will be collected
monthly from March through October which is anticipated to coincide with the peak periods of
potential entrainment based upon historical data and the life histories of the fish species present
in Lake Norman. Sampling will occur during the same weeks as entrainment sampling.
During each of the eight survey events, ichthyoplankton samples will be collected during the
nighttime period. Four shallow water samples will be collected at approximately 10 feet deep.
Preliminary study designs explored the potential to collect samples at 10 feet above the lake
bottom (approximately 75 feet deep). However during equipment testing, hydroaccoustic
surveys of the area identified high levels of debris at this depth raising safety concerns;
specifically that the debris could snag nets and potentially causing small craft to capsize.
Therefore, deep water sampling was deemed infeasible at present and dropped from the study
design. The study design is summarized in Table 7-2.
Duke Energy 129

Entrainment Characterization Study Plan
Marshall Steam Station
I
Table 7-2. Study Design Matrix
Total Number
Locations Depth Samples Per Event Number of Events of Samples
8 8
our(4)upstream One per month
UpstreamF 10 eet (i.e., March
the skimmer wall through October) I 64
Four(4) 4011 One per month
Downstream 10 feet downstream of (i.e., March
skimmer wall through October)
1
I
1
Duke Energy 130

NM E all S M — = i M M M S O MO I M N i i
Entrainment Characterization Study Plan FYZ
Marshall Steam Station
Skimmer • ;. ' �.
ff •
Wall ;,� • +,.;+
".• 1.11*" rF
z Sr k
Downstream Transect Upstream Transect s�'':; :�s, ;,
T Intake -------_ --~
Cove + + ,`
%,. :...t. ,* .14%* :'".N.,., .......
Intake ._ . �I:, ': j, 4
Canal j 4,4r . ..
Existing r "�r . • a
C. /
1 1
-
•
I
"
.* r •
MARSHALL
STEAM
�� \. : fix" r ;,
STATION "
•Discharge . R''
A
• ,. Canal
'"` -�, f :.,,n t"
esti
Su . f•i ,UitlUal Glu li•, G•.E Yr, [u l••.I, 111 Ua,U]U�.Af X,4•,lay.. VII a• I I.'If.N IC•V vwiti vtuua..u.I tli. VIS U4.•i Luui uuty
Figure 7-1. Approximate Sampling Locations at the Skimmer Wall of Marshall Steam Station (Source: Image Modified from:
Google Earth)
Duke Energy 131

Entrainment Characterization Study Plan ��
Marshall Steam Station
1 7.3.2 Sampling Gear Specifications and Sampling Protocol
I The ichthyoplankton surveys will be conducted using two boats each piloted and manned by a
2-person crew. The samples will be collected using a 1.0-m square mouth opening (when under
tow) by 8.0-m long Tucker trawl with 333-pm net mesh and fitted with a 4.0-inch diameter PVC
I cod-end bucket (example Tucker trawl net shown in Figure 7-2). The Tucker trawl will be
deployed from a boom off the side of the vessel with the boat in forward gear and at idle speed.
Optimum tow duration will be 10 minutes. Tucker trawls will be conducted against the prevailing
I current, which is presumed to be toward the CWIS, at a speed through the water of 90-110
cm/sec (2.95-3.6 fps). Tow velocities will be adjusted using a flowmeter with deck readout.
I
I .
1 - 6
I
� 1•
I
1
' Figure 7-2. Example Tucker Trawl Net (Source: www.algalita.org)
The Tucker trawl will be equipped with a double-trip release mechanism (DTRM) to permit
Iopening and closing the net during sampling to collect depth discrete samples at 10 or 75 feet
below water surface. A calibrated digital flowmeter will be mounted in the net mouth and is used
in the calculation of sample volume. The digital flowmeter reading will be recorded on the field
Isampling data sheet prior to and immediately upon retrieval of the net. Target net depth will be
set and monitored using a tow cable (wire) length to wire angle relationship; these values will be
I recorded on the field datasheet. Boat tow speed will be recorded for each sample on the field
data sheet. The tow duration will be recorded with a stopwatch from the moment the Tucker
I
IDuke Energy 132

Entrainment Charaderization Study Plan
Marshall Steam Station
trawl is opened by activating the DTRM to the time the net is closed by activation of the DTRM a
second time to close the net. Actual tow times will be recorded on the field data sheet.
At the end of the 10 minute tow duration, the Tucker trawl will be retrieved using the vessel's
hydraulic winch, lifted vertically from the water, set down in the wooden cradle on the gunwale
of the vessel, and secured. The mechanical flowmeter will then be checked for the number of
revolutions to verify tow status. If there was damage to the net, loss of sample, or the net did not
fish properly, the tow will be deemed invalid and the tow repeated. If the tow is valid, then the
' net will be washed down from the outside and the sample concentrated into the removable cod-
end bucket. The contents of the bucket will then be poured into a labeled sample jar and
preserved with a 5-10% formalin solution. All preserved ichthyoplankton will be returned to the
' laboratory for later sorting and analysis as described in Section 6.
' 7.3.3 Data Management and Analysis
The database used for the entrainment data and described in Section 6.4 will also include data
collected during the skimmer wall efficacy testing, including: sample volumes, ichthyoplankton
' densities, and water quality parameters from each side of the skimmer wall and at both depths.
Using the methods described in Section 6.5, collection densities, expressed as number per
100 m3 will be calculated from the trawl data for each taxon and life stage by month of sampling
and depth.
' Estimated reductions in ichthyoplankton density attributable to the skimmer wall, for each taxon
and life stage, will be calculated as:
(Density outside skimmer wall—Density inside skimmer wall)
Density outside skimmer wall
1
1
RECEIVEDINCDEQIDWR
' MAY 31 2016
Water Quality
Permitting Section
Duke Energy 133

Entrainment Characterization Study Plan
Marshall Steam Station
8 References
Alden Research Laboratory, Inc. (Alden). 2012. Generating Station Assessment Draft 316(b)
Rule Compliance Options — Marshall Steam Station. Prepared for Duke Energy and the
Electric Power Research Institute.
Auer, N.A. (ed.). 1982. Identification of Larval Fishes of the Great Lakes Basin with Emphasis
on the Lake Michigan Drainage. Great Lakes Fisheries Committee Special Publication
1 82-3, Ann Arbor, Michigan. 744 pp.
Electric Power Research Institute (EPRI). 2014. Entrainment Abundance Monitoring Technical
' Support Document: Updated for the New Clean Water Act §316(b) Rule. 3002001425.
EPRI, Palo Alto, CA. 156 pp.
Kay, L.K., R. Wallus, and B.L. Yeager. 1994. Reproductive Biology and Early Life History of
Fishes in the Ohio Drainage. Volume 2: Catostomidae. Tennessee Valley Authority,
Chattanooga, TN, USA.
' North Carolina Department of Environment and Natural Resources (NCDENR). 2003.
Basinwide assessment report, Catawba River Basin. Division of Water Quality, Water
Quality Section, Environmental Sciences Branch. June 2003.
North Carolina Department of Environment and Natural Resources (NCDENR) and North
' Carolina Wildlife Resources Commission (NCWRC). 2001. Catawba River Basin Natural
Resources Plan. October 2001.
Simon, T.P. and R. Wallus. 2004. Reproductive Biology and Early Life History of Fishes in the
Ohio River Drainage, Volume 3: Ictaluridae—Catfish and Madtoms. CRC Press. 204 pp.
U.S. Fish and Wildlife Service (USFWS). 2015. Information for Planning and Conservation
(IPaC) Report for Marshall Steam Station (Lake Norman). Generated November 05,
2015.
. 1996. Recovery Plan for Carolina Healsplitter (Lasmigona decorate). U.S. Fish and
Wildlife Service, Southeast Region, Atlanta, GA. 30 pp.
' Wallus, R., B.L. Yeager, and T.P. Simon. 1990. Reproductive Biology of Early Life History
Fishes in the Ohio River Drainage, Volume 1: Acipenseridae through Esocidae.
Tennessee Valley Authority, Chattanooga, TN. 273 pp.
1
' Duke Energy l 34

Entrainment Characterization Study Plan �L
Marshall Steam Station r
APPENDIX A — Select Species Spawning and Early
' Life History Data
Sampling for entrainment year-round at Marshall is expected to be a poor allocation of
resources, since few if any eggs or larvae are likely to be present in Lake Norman during the
winter months. Several species were identified as likely to be present in entrainment samples
and life history information for these species is summarized here and supports a sampling
period of March through October. It is important to note that low densities of entrainable
organisms at the outset and conclusion of the sampling period are unlikely to change the
estimates of entrainment in any meaningful way and thus do not warrant the additional sampling
costs necessary to extend the sampling season.
1
1
1
1
1
1
1
Duke Energy 135

Marshall Steam Station
I
Table A-1. Life Histories of Selected Species Present Near Marshall Steam Station
1 NCDENR Young-of-
Species Spawning Period Spawning Habitat Nest Structure Eggs Fecundity Rates the-Year Cut-off Larvae Size References
Lengths
IFamily Centratchidae
pr
, Iwo --' ,...,.
� emersal and
' :
. , ,... .
' . .
4.
adhesive •r ;
• Spring Shallow calm waters near 3,000 to 188,000 ' g<75 mm 1, 2, 3
Black Crappie(Pomoxi Depression in sand
nigromaculatus) ater Temperature 64.4-68°F 19etationorcoer. � gravels. ;
• verage di
I0.93 mm.
, . w Saucer-shaped Adhesive
I
Spnrig and Early Summer depressions in sand or
Shallow waters with sand and ravel typically one to two Upto 60,000 eggs <50 mm 1, 3,4
Bluegill (Lepomis macrochirus) °II
gravels. g tYp y Average diameter: gg
Water Temperatures 70 75 F feet in diameter and a few 1.2 to di mm
lab
11 -inches deep. _ ,... .
ISaucer-shaped Demersal and -
Spring and Early Summer depressions in sand or adhesive 111
Green Sunfish (Lepomis cyanellus) hallow waters with sari an gravel typically one to two up to 50,000 eggs <50 mm 4 mm 1, 2, 3, 4,5
I 10
Water Temperatures 60-80°F gravels. feet in diameter and a few Average diameter:
inches deep. 1.0 to 1.4 mm ... .iiii
Early Sprang �� Shallow water on bottoms Circular area 2 to 3 feet in Adhesive a
I
Largemouth Bass (Micropterus diameter with clean sand or
composed of sand, gravel,or 5,000 to 43,000 egg - <100 mm 6-6.5 m ` 1,2, 3,5
salmoides) Water Temperatures 60 75°F pebbles near cover. fine gravel clear of organic Average diameter:
debris and silt. 1.49 to 1.67 mm
pemersal and illilli_ -
I
Redear Sunfish (Lepomis Spring Shallow water on firm 'Depressions in sand to soft adhesive {
substrates often in locations mud constructed in areas 2,000 to 10,000 eggs <50 mm11
microlophus) 1, 7,2,5
Water Temperatures 68-70 °F exposed to the sun. containing aquatic plants.; Average diameter:
1.0 to 1.5 mm . ., - _ .
Late Spring and Early Summer Shallow water on bottoms 41111 SIMEMIPMEMININIIINIIIIMINEE
Spotted Bass(Micropterus composed of sand,gravel,or Shallow nest over rock and Demersal and up to 5,000 eggs <100 mm N/A • LIP.I.111W1, 11, 12
punctulatus) Water Temperatures 65-75°F pebbles near cover. gravels. adhesive
IAll Late Spring and Early Summer Shallow waters on loose siltAdhesive
Warmouth(Lepomis gulosus) next to stumps or underwate Depressions in san"san Ai Up to 63,000 eggs <50 mmi 2.3 to 2.8 mm 1, 11, 12
Water Temperatures 69.98-77.9°F logs,or clumps of vegetations gravel. verage diamete
1.0 to 1.1 mm i
II ',� Shallow water on bottoms Saucer-shaped el II 1 i
Redbreast Sunfish (Lepotitts I Spring and Early Summer composed of sand,gravel,or depressions in gravel or Demersal and up to 14,000 eggs <50 mm r. i 4.6-5.1 mm 1, 6, 11, 12
Iauritus) pebbles near cover. silt. adhesive
Family Clupeidae
I
(
Shad Demersal and
Gizzard ShDorosoma Spring and Early Summer adhesive
cepad (Dm) Shallow water Open water Up to 50,000 eggs <100 mm 6-20 mm 1, 2
Water Temperatures 62.6-73.4°F Diameter range:
I0.75to1.1mm
Spring and Early Summer
Threadfin Shad (Dorosoma Eggs scattered over plants orlor and <100 mm 4.9-5.5 mm 1, 3, 4,6
petenense) Water Temperatures 60-80 °F
loose sediments. Open wat adhesive 2,000 to 24,000 eggs

MI
NCDENR Young-of-
I Species Spawning Period Spawning Habitat Nest Structure Eggs Fecundity Rates the-Year Cut-off Larvae Size References
Lengths
Demersal and
I
Sprang and Early Summer Shallow waters and are adhesive
Alewife(Alosa pseudoharengus) Shallow water known to broadcast egg 10,000 to 360,000 eggs.: <50 mm 8, 9, 13
Water Temperatures 50-80.6°F over any substrate. Average diameter: 4 lour
0.95 to 1.27 mm
IFamily lctaluridae
Streams or reservoirs bottoms Depressions in soft 900 to 1,350 eggs/kg of <100 mm 4 tAILo 9 mm 1, 10, 11
Blue Catfish(lctalurus furcatus) Late Spring and Early Summer Adhesive •
I
with cover. sediment bodyweight
i
...t. . .. Eggs are deposited in t
ate Spring and Early Summer. crevices with hollow wood
Adhesive
Channel Catfish (lctaluru : Streams or reservoirs bottom ;debris and undercut banks 1r21
,000 eg <100 mm 1, 4, 6, 7
I
punctatus) with cover. Average diameter
ter Temperatures 70-85(°F). Nest can be made directly 2.4-3.0 mm
in mud bottoms.
I Early Summer Eggs are deposited in
crevices of hollow woody 1111111*
4111
White Catfish (Ameiurus catus) Stream or reservoir bottoms debris and undercut Adhesive Up to 4,000 eggs <75 mm 1 N/A t 1111 1
Water Temperatures banks. Nest can be made -m.
I65-75°F directly in mud bottoms.
Family Percidae
I Adhesive and -- - , 11E4
Late Winter or Early Spring demersal
Yellow Perch Shallow waters with moderate Eggs are deposited over
I (Perca flavescens)
submerged plants, logs, 23,000 egg <80 mm 1C4 to 7 mm
vegetation. gravel, and rocks. Average diamete
Water Temperatures 44.6-51.8 °F
1.9 to 2.8 mm
I Family Moronidae
1111111
Late Spring and Early SummerAdhesive `
I White Perch(Morone americana) Shallow water . Eggs are deposited over
sands and gravel. Average diameter: 20,000 to 150,000 eggs r <75 mm 3 to 4 mm
Water Temperatures 64.4-68°F 0.7 to 0.8 mm . .-
1
1
1
RECEIVEDINCDEQIDVVR
MAY 31 2016
I
Water Q Section
Permitting
1 Duke Energy 137

Entrainment Characterization Study Plan
Marshall Steam Station
1 Life History References
' 1) Rohde, F.C., R.G. Amdt, D.L. Lindquist, and J.F. PameIl. 1994. Freshwater fishes of
the Carolinas, Virginia, Maryland, & Delaware. The University of North Carolina Press.
Chapel Hill, NC
2) Fletcher, D. E., E.E. Dakin, B.A. Porter, and J.C. Avise. 2004. Spawning behavior and
genetic parentage in the pirate perch, a fish with an enigmatic reproductive morphology.
' Copeia 2004. Volume (1): 1-10.
3) Boschung, H.T. Jr., and R.L. Mayden. 2004. Fishes of Alabama. Smithsonian Institution,
' Washington, D.C.
4) Emig, J.W. 1966. Largemouth bass. In: A. Calhoun (ed.). Inland. Fisheries Management.
' State of California, Department of Fish and Game.
5) Pitlo, J., D. Dieterman, and G. Jones. 2004. Largemouth bass (Micropterus salmoides). In:
' Pitlo, J.M. Jr. and J.L. Rasmussen (eds.). UMRRC Fisheries Compendium. 3rd Edition.
Upper Mississippi River Conservation Committee. Rock Island, Illinois. January 2004. pp.
169-173.
6) Carlander, K.D. 1977. Handbook of freshwater fishery biology. Vol. 2. The Iowa State
University Press, Ames, IA. 431 pp.
' 7) Adams, J.C., and R.V. Kilambi. 1979. Maturation and fecundity of redear sunfish. Arkansas
Acad. Scie. Proc. 33:13-16.
8) Nigro, A.A. and J.J. Ney. 1982. Reproduction and early-life history accommodations of
landlocked alewives in a southern range extension. Transactions of the American Fisheries
' Society 111:559-569.
9) Pardue, G.B. 1983. Habitat suitability index models: alewife and blueback herring. Fish and
' Wildlife Service, U.S. Department of the Interior. Washington, D.C. FWS/OBS-82/10.58.
September 1983.
' 10) Graham, K. 1999. A review of the biology and management of blue catfish. Pages 37-49 in
E.R. Irwin, W.A Hubert, C.F. Rabini, H.L. Schramm, Jr., and T. Coon. editors. Catfish 2000:
proceedings of the international ictalurid symposium. American Fisheries Society
Symposium 24, Bethesda, Maryland.
11) Hendrickson, Dean A., and Adam E. Cohen. 2015. "Fishes of Texas Project Database
' (Version 2.0)" doi:10.17603/C3WC70.
12) Ross, S. T. 2001. The Inland Fishes of Mississippi. University Press of Mississippi,
' Jackson.
13) Myers, P., R. Espinosa, C. S. Parr, T. Jones, G. S. Hammond, and T. A. Dewey. 2016. The
Animal Diversity Web (online).
Duke Energy 138

tEntrainment Characterization Study Plan LYZ
Marshall Steam Station 1
I APPENDIX B — Response to Informal Review
111 Comments
While not required to be peer reviewed under the Rule, Duke Energy engaged subject matter
I experts to informally review this Entrainment Characterization Study Plan. The purpose of the
informal review was to afford the Biology Peer Reviewer the opportunity to evaluate the
entrainment study objectives and methodology, and to comment if the proposed methods do not
I meet industry standards. Duke Energy's intent was to ensure that if data were collected as
detailed in the ECSP that the data would be sufficient for the intended use in the Best
Technology Available (BTA) determination process required in §122.21(r)(10)-(12), and would
not be questioned at a later time.
In order to help focus the review, charge questions were developed (Table B-1). The primary
I
goal was to develop a study that meets the objectives of the Rule-required Entrainment
Characterization Study.
ITable B-1. Directed Charge Questions
Question Entrainment Characterization Study Response Comments(if any)
Number Plan
Will the proposed sampling depth(s) Yes/N
1) and location provide for a
representative sample of the water
Icolumn?
Considering fish and shellfish known o . Wig
expected to be in the source waterbody'
I 2) will the proposed sampling peri
(months) provide the ability t
understand seasonal variations i
I it. entrainment? ..
Is the sampling equipment proposed Yes/Nom
113) appropriate to collect entrainable
organisms at this type of intake
I structure?
Does the plan lay out QA/QC Yes/No
4) requirements clearly? Are these
I
requirements adequate?
Identifying eggs and larvae to species is Yes/No
often difficult and sometime impossible., "
I Does the sampling plan provide.
5) sufficient measures to preserve;
organism integrity and support'.
identification to the lowest taxon;
practicable?
I Does the study design meet th
6) requirements of the Rule at 40 CF
122.21(r)(9)?
I
I
IDuke Energy 139

Entrainment Characterization Study Plan
Marshall Steam Station FYZ
Question Entrainment Characterization Study Response Comments(if any)
Number Plan
' Will the study design provide sufficient Yes/No..
data to support a benefits analysis of
7) entrainment reducing technologies
required to be evaluated by the Rule
' including biological performance as
required in 40 CFR 122.21(rx11)?
[ Are there any deficiencies in the stud
' '
plan that might prevent you or othe '
8) (e.g., Regulators) from understandi
what is being proposed for sampling?
so,what needs to be added or clarified?
After receipt of the peer reviewer's comments, the following responses were developed and the
ECSPs were updated to reflect those changes. Comments were divided into four categories as
follows:
• Category 1: Comments that are clearly applicable (i.e., relevant under the charge and
improve the quality of the work product). These comments were incorporated into the
ECSP.
• Category 2: Comments that represent a misunderstanding by Informal Reviewers. These
comments were not incorporated into the ECSP.
• Category 3: Comments that are minor and do not materially change or lend additional
value to the ECSP (e.g., comments that were meant as "FYI", or meant as preferential
suggestions, or are beyond the scope of the charge). These comments may or may not
' have been incorporated into the ECSP at the discretion of the Report Originator.
• Category 4: Major Peer Reviewer comments that the Report Originators do not agree
with and choose not to incorporate into the ECSP.
' Below are the site-specific responses to comments received on the Marshall ECSP.
1
1
1
t
1
1
Duke Energy 140

111
Entrainment Characterization Study Plan
Marshall Steam Station
ITable 8-2. Peer Reviewer Responses to Directed Charge Questions
I
Category Charge No. Comment Response and Resolution
The depth distribution of the three openings on the in-water sampler is appropriate to adequately integrate .. .... _
3 1 any vertical variation in distribution of eggs or larvae(but see comments regarding potential effects of No response needed
I
orifice size on sampling efficiency in response to question 3).
1 I agree that the proposed locations in the intake bays for Units 3 and 4 should be representative for CWIS.j1.No response needed
It is unclear to me from the information provided where exactly the sampling pipe orifices will be situated in tcalioECSP was updated to better explain the location of sampling pipe.
I
relation to the walls or other physical structure of the bays. Just as water velocity in a stream can vary in
different places in response to physical structures, and is always slower near the bottom and sides than in _ o the extent possible the entrainment sampling pipe will be placed in an area of the intake that is well mixed and
the middle of the stream and water column, so it seems possible that flow may be lower in localized areas typical of hydraulic conditions within the intake. That is,away from structures that could be causing vortices(highly
near the sides of the intake or around any structures(e.g., the wall or structure on which the sampling pipe turbulent areas)or in eddies or other areas of stalled flows.
I
• is mounted)due to eddies, turbulence or other hydrodynamic processes. It will be important to ensure that
the intake flow immediately adjacent to each orifice opening is representative of flow across the mouth of At Marshall, the pipes will be located midway between the bar racks and the fixed screens at the center of an intake
the intake in general. If this cannot be determined with confidence from first principles or historical data,a bay in Units 3 and 4.
I
few strategically located measurements with a current meter may suffice.
1 r 1 How will the protocol be modified if the water surface elevation drops below the upper sampler inlet? The ECSP reviewed included incorrect inlet elevations.The inlet elevations have been updated: upper inlet(745 ft),
middle inlet(737.5 ft),and bottom inlet(730 ft). Figure 6-2 was updated to include these new elevations.
t Figure 2 in the ECSED document shows a schematic of the CWIS, but does not show the in-water No action taken because Figures 2-3 and 2-4 are used to describe the cooling water intake structure. In section 6,
2 1 sampling pipe or the approximate location of the three sample inlets; it should show these features as in specifically Figure 6-1, depicts the location of the in-water sampling pipe and Figure 6-2 shows the three sampling inlet
figure 6.2 of the ECSP document. locations.
The proposed sampling period and frequency are appropriate to encompass the periods when fish eggs
2 and larvae are likely to be present, and to provide information on seasonal patterns of entrainment(but see No response needed
concerns regarding the need for sample replication in response to question 6).
I
lt, It isn't clear why life history information is provided for some species and not others(e.g., no information is
provided for several of the species that were present in 1976 entrainment samples). Detailed life history
information of the type provided is reasonable for the primary entrainment candidates, but at least limited,
key information ought to be provided on the other species present in the reservoir. Consider providing a
1 2 complete species list in table form that gives the spawning period for each species,the kind of eggs and This is an excellent idea. This was added to this and other ECSPs. See revised Appendix A.
larvae they have(e.g., adhesive eggs, pelagic larvae)the size of their eggs and larvae, and other key
relevant information, along with the source reference(s),to provide some evidence that the sampling period
is sufficient for all of them. That would be more valuable than extensive narratives on each species that
include information that isn't particularly relevant.
___ 2 Section 9.4 on White Perch does not indicate the times or temperatures at which they spawn. This information was added to the ECSP.:.
Pump sampling has been used in a variety of settings to sample zooplankton and fish early life history •
1 stages. However, sampling efficiency of pump sampling is likely to be different(likely lower)than traditional
net sampling approaches. Traditional net deployment and retrieval methods may not be feasible in the
intake bays, but if a judicious number of samples could be collected with another method,at a time when
I
densities are relatively high),that would go a long way toward addressing any concerns about if, or how See Pumped Sampler White Paper(Appendix C).
I much,this method underestimates actual entrainment. If such comparisons are already available from
other studies, so much the better; note them. Another approach that might allow a useful comparison
would be to collect samples using a Schindler-Patalas trap. However,this gear samples a very small
- volume of water, so even at peak densities it may not be feasible.
Orifices were sized to achieve equivalent flow rates through each port while minimizing friction losses to reduce the
Orifices will vary in size to allow equal flow from each depth. It is important that flow(and therefore amount of suction lift required.Calculations were performed for each facility to account for differing standpipe and
contribution to the total sample volume)be equal through all three openings, but orifice size and flow piping dimensions; as well as reservoir, intake deck, and inlet port elevations.Calculations were based on friction
I
velocity at the orifice opening are both likely to affect sampling efficiency. If each orifice samples the same losses(converging flow, piping material and length,orifice size)and elevations to determine total head and pump flow.
amount of water, but samples organisms with different efficiencies, then the combined sample of entrained Details will be provided in the final report.
•
organisms will not reflect equal contributions from all three depths. Volumes sampled and orifice intake velocities are both important considerations and will factor into the design of the
sampling pipe.
RECEIVED/NCDEQ/DWR

mi
Category Charge No. Comment
I Sizing of the sampling pipe orifices will be described in greater detail in the final reports.
It would be difficult to simulate the flow field conditions from a power plant in a laboratory setting. Presumably one
would want to test at the approach velocities similar to levels in the field.This would require a test flume with flow
What sizes will the orifices be, and what will the flow velocity be at the opening? Both could affect capture control to achieve the desired test velocities. Since one would be removing organisms and water while sampling, one
I
efficiency,given escape behaviorsracae rheotaxisededhat the d uy larval fish. If the organisms
arefy sml it would have to develop a method to replace the removed organisms and water. If you use a recirculating flume, so only
4 3 may not matter. But some assurance is needed that the setup will sample the organisms equally, not just the volume of water removed would need to be replaced,then that replacement water would have to seeded with
the water. If necessary,test runs could be done in the lab, sampling larvae of a known density out of a organisms of a known number. The test pump would be removing 240 gallons per minute, so your total capacity in your
I tank. flume would likely need to be at least 5x this volume,if not much higher(otherwise the pump sampler would be
inducing flows not the circulating pump on the flume). If you want to test different species, life stages and/or organism
sizes,then several tests would be required. What at first glace appears to be a simple test, is actually quite
complicated.
I If a 330-pm mesh net works without clogging,that would be ideal; but if problems occur then use of a-500.
3 pm mesh net would be acceptable for larval fish collections. That mesh size is often used in larval fisho response needed
1. collections.
IManyof the species in the lake have eggs that are smaller than the mesh size to be used in the collecting We disagree with the statement that many of the species in Lake Norman have average egg diameter smaller than the
3 mesh size(0.333 mm). Most species average diameter range from 0.7 mm to greater than 1.5 mm. Therefore, no
net. Some rationale is needed for why this isn't an issue.
additional action is required.
I
The proposal notes that"properly designed and operated pumped systems have shown collection Systems using trash pumps with recessed impellers routinely collect greater than 95 percent of fish eggs and larvae.
efficiencies of 95 percent or greater for fish eggs and larvae with little or no organism damage(EPRI Unfortunately neither EPRI 2005 or EPRI 2014 can provide much greater detail. Both state, "Studies of properly
3 2005 ." However, it wasn't clear how this was measured and if it directly applies to this situation(I believe
also indicated that an updated version of this EPRI document is available). It would be helpful designed and operated pumped systems have revealed little damage or destruction of entrained organisms with
if you could elaborate a bit. collection of greater than 95 percent of fish eggs and larvae being routinely achieved."
,;,agree that collection at the intake structure is preferable to sampling organisms after passage through the
3 3Imp oling system. No response needed •
3 •
' 41111.11111Pgillir The lab and field SOP and audit plans are generally sound. No response needed
However, an Average Outgoing Quality Limit(AOQL)of 1% (?99%accuracy)strikes me as rather liberal
I
4 for data entry. It seems to me that an error rate of one error per 100 entries is too high. How does it Typically the average outgoing quality(AOQ)is better than the AOQL. Both this and the AOQL for larval identification
agimikr,„ compare to the observed error rate on similar work(I expect actual accuracy is typically better than that)? If are industrystandard for entrainment sampling.
it is feasible to commit to a lower error rate that would be preferred.
Likewise, an AOQL of 10 percent for organism identification seems pretty high to me(I expect the error rate The sampling plans implemented under our proposed QC procedures have a specified average outgoing quality limit
will be lower than that for experienced personnel). I recognize that identifying fish eggs larvae is trick (AOQL)of 10 percent,which represents the maximum fraction of all items(e.g., measurements,taxonomic
p p ) g ty g y' identifications or counts)that could be defective as a worst case. A defective item could be a measurement or count
so I fully expect some individuals to end up in broader categories(e.g., unidentified shad or unidentified that falls outside of a specified tolerance limit(e.g., plus or minus 1 to 10 percent). Typically the average outgoing
4 P yp Y 9 9 9
larvae)-I don't consider that an identification error. But I would expect organisms identifiable to a given quality(AOQ)is better than the AOQL. Items are inspected using a QC procedure derived from MIL STD (military
taxonomic level to be correctly classified more than 90%of the time. Again,what is the observed error rate standard) 1235B(single and multiple level continuous sampling procedures and tables for inspection by attributes)to
on similar analyses? Maybe my expectations are too high. I
the 10 percent AOQL . Both this and the AOQL for data entry are industry standard for entrainment sampling.
II The data security and chain-of-custody plan is good, but one can never be too careful. Data remain
1 4 vulnerable to loss during the period when they exist only on one hardcopy datasheet, particularly while still This is a good idea.We will add words that a digital image of the datasheet will be taken in the field prior to the
in the field. You might consider taking a picture of each datasheet when completed,to have an electronic datasheets leaving the site.
backup until the datasheet can be scanned or entered into a computer.
I
Given that regulatory compliance is sometimes the subject of litigation, I think that retaining samples for Revised to indicate organisms will be held a minimum of three years after the end of the project or longer if Duke
4 only three years is not sufficient. My sense is that something on the order of seven years would be a better Energy requests additional storage time.
safeguard.
i.
1
5 Adequate information is provided to document that the specific pump(s)to be used are of the type that will No response needed
IIIIIIIIIIIIII not cause organism damage as noted in EPRI (2005).
3... 5 Proposed preservation methods will fix organisms in a manner that will maintain their morphological No response needed
i • integrity for identification purposes.
RECElVEDiNCDEQIDWR
1 a Attribution of comments to reviewers and references to reviewers have been removed from this document consistent with peer review standards. MAY 3 1 2016
Water Quality
Permitting Section
IDuke Energy 142

■ Category Charge No. Comment
o
.
The proposal indicates that"To the extent practicable,long-dead, moribund, and/or non-viable eggs will be
identified using appropriate and well-defined techniques developed in the SOP and categorized in the
database accordingly.When estimates of entrainment are generated, moribund, dead,and non-viable
individuals will not be included." The discrimination of dead,moribund or non-viable eggs from live eggs is Agreed.We will want to look at the data inclusive and exclusive of our two categories.At present there are few reliable
I
critical step becauseenfit directly ain samplest,
estimates. Differences between live and dead methods that are not time consumptive or expensive to implement. Here we are thinking of excluding only the most
individuals are often fairly obvious in fresh samples, but can be markedly reduced after preservation. s` f obvious categories of organism. For example,we might require
Because any error in this process will bias entrainment estimates downward I expect this step would P 9 eggs be whole,show signs of fertilization, and not be
receive heightened scrutiny. Therefore the methodology should be fully explained, and be pretty iron-clad, covered with fungus.
I Even if it is, I certainly recommend retaining both groups of eggs in separate vials,and be prepared to
provide entrainment estimates based on both groups combined as a conservative measure of entrainment
fif necessary.
he final Rule does not require replication nor is there an obligation to provide confidence intervals or bounds around
e entrainment estimates generated.The study must be sufficient to show diel,monthly,and annual variation,which
I is study plan addresses.
i
'I .The Rule requires"sufficient data to characterize annual,seasonal,and diet variations in entrainment, e interpret the Rule as requiring sufficient sampling to collect data over the range of conditions that are likely to occur
#including but not limited to variations related to climate and weather differences,spawning,feeding,an. ' nd to prevent bias through selective sampling. For example,you could not propose to sample only during the day,
water column migration." The proposed sampling plan calls for collecting a single,large sample in each •:cause you would miss any density differences due to diel variability.You could not propose to sample only on sunny
sampling period. I believe that this collection plan will provide data representative of the entrainment at th •ays, because you would miss any density differences due to weather.You could not propose to sample only from near
I intakes,but determining if apparent patterns or differences are real(as the requirements seem to call for) ': e bottom of the intake,because you would miss any density differences due to vertical stratification in the water
requires some measure of variability in the estimates. That requires replicates. The number of samples •lumn.
collected over the course of the project will be sufficient to detect annual variation between the two years
but seasonal,diel and weather effects would be confounded with each other. Additional replication is e believe that the way in which these data will be used do not justify extensive replication. Relationships between ,
I necessary in order to determine if any of these factors affect entrainment. For example, one could not - eather,climate,spawning,and feeding (as a few examples)and entrainment rates are not going to change the
separate weather effects from temporal differences in this sampling design. If it is necessary to be able to etermination of best technology available for entrainment reduction or the outcome of any social cost/social benefit i'.
'show whether or not there are effects of weather,enough samples will be needed to use weather variable:; Iculations. In addition,the study plan includes some replication. Each sampling event is divided into four independent
(e.g.,water temperature,cloud cover)as covariates to test for effects. - - ..mples based on time of collection.
I
In addition sampling events occur twice in each month. If necessary,confidence intervals can be generated base•on
. ese 8 samples within a month. Determination of whether confidence intervals are beneficial can be made at the -.,.u•
.1 the •r•.ram.
I .
That said, I think this issue could be addressed with a modicum of additional effort. The simplest approach
would be to divide each 2-hour sample into at least three,preferably four, samples collected immediately =°We disagree that splitting each 2-hour sample into smaller sub-samples requires only a small additional effort.While it
S' one after the other. The collection cup(or entire net)could be swapped out after a sample and processed is true that the extra effort in the field would be minimal(extra net wash-downs, extra datasheets to fill out),the effort
I - while the next sample is being collected. This replication would allow straightforward statistical analysis to j'(and associated labor costs)in the laboratory would increase proportionally. So splitting the 2-hour samples into four
determine if these factors(or their interaction)affects entrainment. The individual samples could still 1,sub-samples will quadruple the lab costs.
'` combined into one composite sample if warranted, but the inverse isn't true.
Replication is important for density estimates,but it would not be necessary to have morphometric
itia
s measurements on a full compliment of individuals from each replicate;one pooled sample for each six-ho reed
. period would suffice. A total of up to 10 individuals of each taxon could be drawn at random from all the ,_`
individuals collected in all replicates combined within one six-hour sampling period.
Any vertical migration should be adequately integrated by the multi-depth sampling scheme,and the
temporal pattern of sampling should detect the seasonality of spawning. It is unclear to me what"feeding We agree.Our interpretation of EPA's request is that sampling encompasses the range of conditions and fish behaviors
3 6 variation" refers to or how sampling will assess it,but it is also unclear to me how that is relevant for this likely to occur in a typical entrainment season.That is, if feeding behavior impacts vertical migrations in the water
' assessment. If this refers to changes in feeding behavior over the diel period or seasonally that would column,then you would need to sample during periods that include those in which larvae are feeding.
increase or decrease vulnerability to entrainment,then I believe such effects would be adequately captured
by the proposed sampling scheme.
ifin
I 3 7 With minor modifications as noted in responses to other questions,this study design should provide a No response needs
sound basis to support the required benefits analysis.
In general the proposal is clearly written and understandable,with only minor exceptions. Some points to
3 ;I 8 be added or elaborated upon,or deficiencies in the design, have been noted in response to other No response needed
Iquestions.
I
IDuke Energy 143

Category Charge No. Comment Response and Resolution
/11P lir 'WM
Ii
Even if flow through the net is substantially lethan 100%each tow will sample vera)hundred cubicmeters of water,so the proposed sampling plan should provide samples that are quite representative of fi
larval densities in each location on each sampling date. However,as noted in response to question 6,
differences cannot be demonstrated statistically,and therefore meaningfully,without replicates. With only •
I .I
single deep sample on each sampling date for each site,it will not be possible to say with any confidence
~that larval densities at those deep sites are(or are not)different from each other,from the shallow sample :ased on Best Professional Judgment and historical thermal data from the area near the skimmer wall,we anticipate
j or from the corresponding entrainment samples. As noted above for the entrainment sampling,some 'ttle to no ichthyoplankton at 60 ft(i.e.,the deep water location),which is below the thermocline when entrainment is a
replication is necessary. Samples from different dates are not replicates,as seasonal or weather-related encem(Lake Norman typically turns over in November and through the winter months).Earlier drafts of the skimmer
I 'differences would confound any comparison. The text in Section 7.3.1 indicates that two shallow samplesall study plan included no deep water samples.Deep water samples were introduced to demonstrate no
i and one deep sample will be collected during each sampling event(lines 9-10,p.26),but Table 7.2 hthyoplankton at this depth.We would prefer to focus our resources to sampling in the upper water column where we
;

Entrainment Characterization Study Plan
Marshall Steam Station 1-
)1
APPENDIX C - Comparison of Pumps and Nets for
' Sampling Ichthyoplankton
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
' Duke Energy 145

Cha
1 arshall ent Steam Station tion Study Plan 1 ��
1 Comparison of Pumps and Nets for Sampling Ichthyoplankton
1
Prepared by:
1 FY2
440 S. Church Street, Suite 900
1 Charlotte, NC 28202-2075
February 19, 2016
Introduction
As a part of Duke Energy compliance projects associated with the 2014 Clean Water Act
§316(b) rule for existing facilities (Final Rule), the company submitted draft Entrainment
1 Characterization Study Plans (ECSPs) to informal review by subject matter experts. In the
comments received on the draft ECSPs and during discussions at a peer review kick off
meetings, there was a concern among the biological reviewers that the proposed pumped
1 ichthyoplankton sampling method could impart a systematic bias compared to nets for
estimating power plant entrainment. In particular, there was a concern that pumped samples
could underestimate entrainment and that additional gear efficiency testing could potentially be
1 undertaken periodically throughout the entrainment sampling period to determine if a bias exists
and quantify the magnitude of difference if it exists. As a preliminary step, HDR conducted a
literature review. The sections that follow provide background, the methods and results for the
literature review, and conclusions.
Background
1 Two primary methods have been historically used to estimate ichthyoplankton entrainment at
power plant intakes: streamed/towed nets and pumped samplers. Traditional ichthyoplankton
1 nets can be used to filter water as it enters the intake or exits the discharge and collect
organisms. Alternatively, pumps can be used to convey water to a fine-mesh net onshore.
Onshore nets are typically suspended in a buffering tank to minimize damage and extrusion of
eggs and larvae.
Each method has advantages and disadvantages and a comparison of the two methods are
summarized in Table C-1. The primary advantages of utilizing pumps include ability to meter
precise sample volumes, longer sample collection times, reduction in the potential to miss
1 samples due to inclement weather or other events, and increased ability for technicians to safely
observe net filtering and other aspects of the data collection. Their versatility includes being
utilized in fresh, estuarine, and marine water environments. Properly designed and operated
systems can be accurate and effective. While no sampling method is perfect, pumped samplers
Held at HDR offices in Charlotte,NC,January 28-29,2016.
1
1 Duke Energy 146

Entrainment Characterization Study Plan
Marshall Steam Station
were determined to offer the best, most cost-effective, and consistent sampling method for Duke
Energy and therefore were included in the draft ECSPs.
Pumped samplers have been used in sampling plankton as far back as 1887 (Gibbons and
Fraser 1937; Aron 1958; both as cited in Taggart and Leggett 1984) and have a long history of
successful application for collecting ichthyoplankton samples at power plants (Bowles and
Merriner 1978; as cited in EPRI 2014). Pumped samplers are among the preferred gear types
accepted by EPA and have been used extensively to successfully monitor entrainment at power
plant intakes for decades.
At present, HDR is aware of on-going or state approved plans for pumped entrainment sampling
at power plants to support the Final Rule in at least seven states: Florida, Wisconsin, New
York6, Pennsylvania, New Hampshire', Massachusetts, and Virginia. If this list of states is
expanded to include sampling under the remanded Phase II Rule, then the list of states would
be expanded to include Connecticut, Louisiana, Texas, Ohio, Michigan, New Jersey, and
Maryland. In addition, there may be other states that HDR is unaware of, that have also
approved this approach. Still other States could be added to the list as more power generating
companies begin entrainment studies over the next few years under the Final §316(b) Rule.
Table C-1. Advantages and Disadvantages of Hoop Nets and Pumped Samplers for
Estimating Ichthyoplankton Density in Cooling Water Intake Structures (some
information adapted from EPRI 2014)
Gear Type Advantages Disadvantages
Hoop Nets - Large volumes are sampled quickly(less - Can be difficult to deploy and retrieve in the
Deployed in the manpower required for the same number of confined space of intake structures—
Intake pumped samples). precludes the use of some net types(e.g.,
- If net frames are not used,then there is standard bongo, neuston nets,or Tucker
limited to no modifications to the intake trawls).
required for deployment. -Depending on deployment method, may
-No potential for mechanical damage require modifications to intake structures
(e.g., frame mounted nets in frame guides).
associated with pump passage.
- Less precise flow metering than pumped
samplers.
-Large volumes are sampled quickly—
capturing less temporal variability in a single
sample as compared to pumped samples.
- Relatively small nets needed to fit in the
intake structure offer a small spatial sample.
Multiple nets can be used to increase
sampled area at the cost of additional
samples to be processed in the laboratory.
6 New York State applies a more stringent standard than the Final Rule under the New York State Department of Environmental
Conservation Policy CP-52 for Best Technology Available(BTA)for Cooling Water Intake Structures.
The states of New Hampshire and Massachusetts do not have delegated authority to issue NPDES permits and are administered
by EPA Region 1.
1
Duke Energy 147

' Entrainment Characterization Study Plan FIN
Steam Station
Gear Type Advantages Disadvantages
- Tow speeds in the range of 1-2 meters per
' second(commonly used during
ichthyoplankton sampling)is above the
intake velocities at the majority of intakes.
' - Some active avoidance possible by larger
motile life stage (e.g.,late larvae and early
juvenile). Larger hoops can be used to
decrease potential for avoidance, but would
' require a larger deployment area, since
length is proportional to opening diameter in
properly sized nets8.
- Greater potential for extrusion than pumped
samples(no buffering tank).
- Boat deployed nets are subject to weather
delays and associated safety concerns.
' - May be restricted to relatively deep areas that
are free of floating debris, submerged
snags, and other obstructions
Pumped - durations are typically longer Some active avoidance possible by larger
Samplers in the increasing the potential to capture temporal motile life stages(e.g., late larvae and early
Intake variability in ichthyoplankton densities not juvenile).
observed in net samples. Improperly designed samplers can lead to
Limited modifications to intake or discharge damage to organisms during sampling.
structures are required to install—usually
just anchoring points for the sample pipe. -
Samples a smaller portion of the spatial
I variability, because pump inlets are
- In-line flow metering offers greater precision generally smaller than net openings.
in measuring the volumes of flow sampled.
-Some potential for mechanical damage.
However, correctly designed systems can
offer<5%damage or destruction of eggs
and larvae.
- Fixed pipe allows precise control over water
depth and orientation to intake flows.
-Less potential for extrusion than nets,
because the filtering net sits in a buffering
tank.
- Lower potential for missed samples due to
severe weather.
-Allows technicians to observe sample
collections and minimize potential for invalid
samples(e.g., use of 330-urn nets
increases potential for net occlusion and
frequent net change outs may be required
during certain times of the year).
8 A general rule of thumb, as described in EPRI 2014, states that the total effective open area of the netting (percent open area x
area of netting)should be at least twice the area of the net mouth opening. Others have suggested a net length to mouth opening
diameter ratio of three or more.
1
Duke Energy 148

Entrainment
' Marshall Steam Station Study Plan
' Literature Review
' Methods
HDR conducted a literature review to assemble and assess the available data on the
' comparative effectiveness of pumped entrainment samplers and nets to estimate power plant
entrainment. Available data were identified via online electronic searches for published articles
and government and industry reports. The electronic literature search used Google, Google
' Scholar, and the ProQuest Aquatic Science Collection database accessed through the
University of Massachusetts, Amherst. In addition, we consulted the EPRI technical support
document on entrainment abundance monitoring (EPRI 2014) to identify relevant literature. The
' literature search revealed a short list of relevant citations and abstracts. Several of these were
available in HDR's corporate library and some were obtained from the publishers. For some of
the industry gray literature and less common symposium papers, we relied on abstracts or
' summaries presented in other documents (primarily EPRI 2014).
Below is an annotated bibliography of several of the key references that were reviewed. To the
extent practical, these studies were critiqued and the results and findings put in the context of
the goals of entrainment monitoring at the Duke Energy facilities.
Annotated Bibliography
Cada, G.F. and J.M. Loar. 1982. Relative Effectiveness of Two Ichthyoplankton Sampling
' Techniques. Canadian Journal of Fisheries and Aquatic Sciences 39(6): 811-814.
Cada and Loar (1992) collected triplicate pump and towed-net samples during both day and
night near the water's surface to detect differences in the effectiveness of a pumped sampler
and towed nets for collecting ichthyoplankton. The effectiveness of 1.1 m3/min (290 gallons per
minute [gpm]) pump°was compared to a 2 m long (6.6 feet), 0.5-m (1.6-feet) diameter, 243-pm
mesh conical plankton net in the headwaters of Watts Bar Reservoir in eastern Tennessee. The
net was towed for 5 minutes at velocities between 120 and 189 cm/s (4 to 6 feet per second),
which resulted in sample volumes of 85 to 100 m3. The pump system used a 7.6-cm (3-inch)
' diameter intake hose that had a cylindrical trash screen with 6.4-cm2 (1-inch) openings. This
hose inlet was placed 0.25 m (10 inches) below the surface and slowly moved through the
water.
Both gear collected clupeid larvae (Threadfin Shad and Gizzard Shad) between 4 and 17 mm
(0.16 to 0.67 inches) although larvae greater than 10 mm (0.4 inches) long were collected in
relatively low numbers. Clupeid larvae 4-10 mm (0.16 to 0.4 inches) long were collected in
sufficient numbers to make meaningful statistical comparisons based on date, station, diel
period (day vs. night), and fish length. Due to the difficulty in differentiating small Dorosoma spp.
both species were combined for the analysis. The differences in the mean densities of most
length groups between nighttime pump and tow samples were not statistically significant.
' °By comparison,the pumps proposed for use for entrainment sampling at Duke Energy facilities have a target capacity of 240 gpm
with a range from 5 to 380 gpm depending upon head.
Duke Energy 149

Entrainment
Marshall Steam Station
��on Study Plan
' However, the pump collected significantly more 4-mm larvae and the towed net collected more
5- to 10-mm larvae during the day. Cada and Loar (1992) concluded that based on sampler
' intake velocities and the potential for avoidance, it would generally be expected that pumps are
more effective than fixed nets, that towed nets are more effective than low volume pumps, and
that high volume pumps are equally or more effective than towed nets in collecting
ichthyoplankton. The lack of differences between densities of ichthyoplankton at night suggests
that gear avoidance is associated with a visual escape response and not a tactile response to
changes in hydraulic conditions. The authors caution that any gear comparisons should account
for sizes of sampled organisms and how that could impact performance and that care should be
used when using pumped samplers, because pumps may not adequately sample all entrainable
' organisms.
These results should be used with caution and may not be representative of what would be
observed with a fixed entrainment sampler at a cooling water intake. First, Cada and Loar
(1992) moved the pump inlet though the water which could alter the flow fields relative to what
would be observed around a static system. Second, the distance from the boat that these two
systems were deployed is unknown. Theoretically, boat-induced hydraulics, noise, and/or
shadow could have induced differential behavioral responses between the two gear types if
deployment distances were different. Third, the inlet to the pump had a trash screen with 1-inch
' openings. The authors did not describe the size of the trash screen. This screen and its wake
could have acted as visual avoidance stimuli.
Elder, J.A., J.W. Icanberry, D.J. Smith, D.G. Henriet, and C.E. Steitz. 1979. Assessment of
a Large Capacity Fish Pump for Sampling lchthyoplankton for Power-plant Entrainment
Studies. California Cooperative Oceanic Fisheries Investigation 20: 143-145
' A centrifugal, single-port bucket-style pump that delivered 3.0 m3/min (793 gpm) at a 3-m (10-
feet) head and in excess of 4.3 m3/min (1,136 gpm) at lower heads was evaluated for sampling
' entrainment. These pumps were originally designed for hatchery and aquaculture use and were
reported by the manufacturer to lift 30-cm (12-inch) trout 3 m (10 feet) above water surface with
99.5 percent survival. The pump discharged into a 505-pm mesh net with a cod-end bucket
' suspended in a 2 m3 (528 gallon) box. The inlet to the pump was a 15.2-cm (6-inch) diameter
pipe. The exit pipe was larger (25.4-cm [10-inch] diameter) to reduce velocity entering the net.
This pump sampler was compared to the sampling efficiency of a 1.0-m [3.3-feet] 505-pm nylon
mesh net towed across the mouth of the intakes near where the pump was operating. During 20
paired samples, there were no statistical differences in larval fish densities captured by the net
and the pump. There were statistically higher densities of opossum shrimp (Neomysis spp.)
collected by the pump than by the nets. Damage to pumped specimens was limited. Specimens
collected by pump remained highly identifiable and were in the same physical condition as net-
captured specimens. Given that the pumps evaluated by Elder et al. (1979) had a much higher
capacity than those proposed for use at the Duke Energy facilities, it is difficult to make a direct
comparison, but the results indicate that a pumped sampler could be as efficient as nets for
' sampling entrainment.
Duke Energy 150

Entrainment
Marshall Steam S tioonzadon Study Plan
' Gale, W. R, and H. W. Mohr, Jr. 1978. Larval Fish Drift in a Large River with a Comparison
of Sampling Methods. Transactions of the American Fisheries Society 107:46-55.
While this study was primarily designed to test the drift of larvae in the Susquehanna River
(1974-1975), use of a pumped sampler and nets simultaneously allows for a comparison of the
two gear types. Samples were collected from April 10 to October 10, 1974 with boat mounted
nets. The nets were made of nylon and had 0.4 x 0.8-mm mesh and 24 x 54-cm (9.4 x 21.3-
inch) rectangular mouths. Simultaneous samples were collected at near each shore and the
main channel (about 80 m [262 feet] from the west shore) at set intervals (0800-1000 h and
2300-0100 h). Five-minute fixed net samples were collected from a stationary boat and push-net
samples were collected propelling the boat slowly downstream for abut 300 m (984 feet). From
March through August 1974 a high capacity trash pump, raised and lowered by a hand winch,
was used to collect samples. Replicate pump samples were collected about 50 cm from the
surface and 10-20 cm (4-8 inches) from the bottom at 3-hour intervals for a 24-hour period near
the middle of the river. Flow rate was about 2,500 liters per minute (660 gpm). Based on surface
samples only,70 the pumped and fixed net samples collected about the same number of larvae
per 10 m3 and no statistical differences were detected.
Harris, R.P. L. Fortier, and R.K. Young. 1986. A Large-Volume Pump System for Studies
of the Vertical Distribution of Fish Larvae Under Open Sea Conditions. Journal of the
Marine Biological Association of the United Kingdom 66(4): 845-854.
' Harris et al. (1986) evaluated a pumped sampler for application in an open ocean setting. The
pump was 2.8 m3/min (740 gpm) submersible, centrifugal pump designed for pumping waste
water. Comparative efficiency trials by day and night showed that the pump was generally as
efficient, or in some cases more efficient, in capturing larvae than towed 200-pm WP2 nets",
although there was some evidence of visual avoidance by particular larval size classes during
daylight. The authors indicated that fine-scale temporal and spatial resolution is necessary to
study the distribution of larval fish and that large-volume pumps, sampling at rates in excess of
1 m3/min (264 gpm) can be used as an altemative to conventional nets. The design and
application of the pump sampler used by Harris et al. (1986) is substantially different (e.g., 15-
cm [6-inch] intake line; boat-mounted and sampled while in motion; and marine open-water
ecosystem) than what is being proposed for use at the Duke Energy facilities. For this reason,
the results have limited direct applicability to the Duke Energy fleet.
'D Sampler efficiency was only presented for surface sampler in Gale and Mohr 1978.
" The WP2 Net is a vertical plankton net with messenger operated closing mechanism based on the design of the UNESCO
Working Party 2.
' Duke Energy) 51

Entrai
ent
Marshall Steam Station�oon Study Plan EN
King, L. R., B. A. Smith, R. L. Kellogg and E. S. Perry. 1981. Comparison of
Ichthyoplankton Collected with a Pump and Stationary Plankton Nets in a Power Plant
' Discharge Canal. Fifth National Workshop on Entrainment and Impingement. San
Francisco, CA, May 1980. Loren D. Jensen, Ed.
King et al. (1981) compared the performance of pumped samples and nets in the discharge
canal at Indian Point Generating Station on the Hudson River. Simultaneous ichthyoplankton
sampling was completed using a 15-cm (6-inch) pump/larval table system and stationary 0.5-m
(1.6-feet) diameter conical plankton nets. A total of 79 paired samples were collected on 6 days
in June and July 1978. The average density of total ichthyoplankton collected was 3.0/m3 for
' pump samples and 3.31m3 for net samples. No significant differences (P > 0.05) were detected
between density estimates for total ichthyoplankton determined from pump and net samples for
eggs, yolk-sac larvae, post yolk-sac larvae, and juveniles. Thirteen out of 14 taxa compared
showed no significant difference between pump and net collections. The pump and net
collection systems were equally effective for estimating densities of most ichthyoplankton.
Petering, R W; Van Den Avyle, MJ. 1988. Relative Efficiency of a Pump for Sampling
Larval Gizzard and Threadfin Shad. Transactions of the American Fisheries Society 117:
78-83.
' Efficiency of a pump sampler and a plankton net were compared based on monthly nighttime
collections on Lake Oconee, Georgia12 from April through August 1982. The two samplers were
' used on the same night or on two consecutive nights. The gasoline-powered pump had a 1.18
m3/sec (312 gpm) capacity. Pumped samples were discharged into a 297-pm mesh net
suspend over the side of a boat. A cylindrical metal sieve with 8 mm (5/16 inch) holes was
' attached to the inlet of the pump to exclude large debris. Three, 10-minute samples were
collected by slowly raising and lowering the intake hose in the upper 2-m (6.6 feet) of the water
column. Sample volumes were assumed to be 11.8 m3 (3,117 gpm) based on manufacturer's
pumping specifications. Triplicate samples (75 m3) were collected with a towed net that was
0.25-m2 (2.7 feet) with 0.5-m (5.4 foot) square opening that was towed at 1.0 m/s (3.3 feet per
second)for 5 minutes. Net samples contained seven fish taxa whereas only two were collected
by pumps; however, Gizzard Shad and Threadfin Shad accounted for more than 97 percent of
the specimens caught in both samplers. The authors conclude that the pumped sampler
collected fewer taxa and that the estimates of shad density were lower and less precise than net
samples.
There are serious concerns with this study design that call into question the authors' findings.
First, when calculating organism density, the authors used pump flow rates as defined by the
manufacturer's specifications rather than metering the flow. The authors evaluated the pump in
the laboratory and determined that the intake rates were typically within 10 percent of advertised
values, but no methods or results for the laboratory evaluation are provided to allow a critique.
This lack of metering casts doubt on the accuracy of the entrainment estimates. Second, similar
'2 A 7,709-hedare pumped storage reservoir in central Georgia. Not to be confused with Duke Energy's Oconee Nuclear Station on
Lake Keowee in South Carolina.
' Duke Energy 152

Marrsshallll Stea S�donnzation Study Plan
to Cada and Loar(1982), the inlet to the pumped sampler was screened and moved through the
water. The screening could alter the localized hydraulics (e.g., increase in velocity entering the
' screening), which may be perceived and avoided by later larvae and early juveniles with
sufficient swimming capacity. Moving the inlet through the water could also induce hydraulic
changes perceptible to fish. It is also unknown if, or to what extent, entrainment rates with the
pumped sampler could have been affected by its proximity to a moving boat. Finally, samples
were not necessarily collected on the same night with both gear types. Daily variability in
ichthyoplankton density was likely an unaccounted for confounding factor.
Leithiser, R.M., K.F. Ehrlich, and A.B. Thum. 1979. Comparison of a High Volume Pump
and Conventional Plankton Nets for Collecting Fish Larvae Entrained in Power Plant
Cooling Systems. Journal of the Fisheries Research Board of Canada, 1979, 36(1): 81-84
Leithiser et al. (1979) compared a high volume pump to conventional ichthyoplankton nets for
entrainment monitoring. The pumping system had a capacity of about 2.5 m3/sec(660 gpm). No
information was given about the size of the piping on the suction end of the pump. The nets
used by Leithiser et al. (1979)were 0.5-m (363-pm) and 1.0-m (335-pm mesh) conical plankton
nets.
The average density of large larvae (> 5 mm Total Length [fL])was significantly greater with the
pumped sampler than what was observed with the 1-m plankton nets. Compared to the pump,
both sizes of plankton nets (0.5 and 1.0 m diameter) in each test greatly under sampled larvae
' over 5.0 mm TL. The data suggest that the pump and plankton nets sampled the small larvae
equally well, but that the larger larvae were better able to avoid the plankton net than the pump
inlet. Leithiser et al. (1979) collected samples in the power plant intake canal at different
approach velocities. While it was expected that larval avoidance would decrease with increasing
channel velocity, this was not observed with either the pumped or net samples, but the number
of channel velocities tested were limited.
The authors concluded that the high volume pump was a more effective larval fish sampler than
the conventional plankton nets. It is important to note that on the California coast where these
studies were undertaken, there are greater densities of small larvae (< 5 mm TL) than would be
expected in southeastern Piedmont reservoirs, which would make the differences between
these two gear types more dramatic at the Duke Energy facilities. It is unclear how these results
can be transferred directly to what is being proposed at the Duke Energy facilities because of
the higher flow rate used in this study(660 gpm vs. 240 gpm).
I
Duke Energy 153

' arehnment Characterization n Study PlanSteam
Leonard, T.J. and G.E. Vaughn. 1985. A Comparison of Four Gear Types to Measure
Entrainment of Larval Fish. Proceedings of the Annual Conference of the Southeast
Association of Fish and Wildlife Agencies 39: 288-297.
A study was undertaken at the McGuire Nuclear Station (MNS) on Lake Norman, North
Carolina. The purpose of this program was to determine the relative efficiency, reliability, and
cost of four different systems for measuring larval fish entrainment into the cooling water
system: 1) a tap valve in the condenser, 2) a pump-net system, 3) a fine-mesh screen; and 4) a
stationary net. Night samples were collected on five consecutive nights 6-10 June 1982, which
was anticipated to coincide with the peak abundance of Threadfin Shad and Gizzard Shad.
The stationary net was a single 0.5-m conical net (2.5-m [8-feet] long, 800-pm mesh) was
suspended from a barge located 7 m [23 feet] upstream of the intake. A flow meter was placed
in the mouth of the net to monitor the volume of water sampled. The net was fished by quickly
lowering the net to the bottom of the intake structure and raising it in 1-m intervals over the 9.5-
m high effective opening to the intake. Rapidly lowering and raising of the net caused the net to
collapse, preventing the flow meter from recording flow during the decent and retrieval. On the
first night the net was raised at 1-m (3-feet) increments at 2-minute intervals, but this resulted in
less than desirable volumes sampled and on subsequent nights the net was raised at 1-m (3-
feet) increments at 3-minute intervals for the remainder of the study, resulting in an average
sampling duration of 27 minutes.
' The fine-mesh screen was a 9-cm (3.5-inch) deep rectangular wooden frame that had a 50.8 x
57.5-cm (20 x 22.6-inch) mouth opening covered on one side with 800-pm mesh. This panel
was attached to a traveling screen in the intake bay being sampled. The traveling screen was
rotated manually to position the fine-mesh panel into the flow. The screen was rotated so that
the fine-mesh panel was raised in approximate 1-m (3-feet) increments at 4 minute intervals,
resulting in an average duration of 38 minutes. Once at the water surface, the screen was
rotated without stopping to retrieve the sample.
The pumped sample was collected using a 10.1-cm (4-inch) diameter centrifugal pump
(2,270 L/min [600 gpm] pumping capacity at 0.6 m head). Water was withdrawn through a 10.1-
cm (4-inch) diameter flex hose and discharged via a 10.1-cm (4-inch) PVC pipe into a plankton
net. An ultrasonic flow meter was used to measure flow through the discharge pipe. The hose
was lowered as close to the bar racks as possible and raised at 1-m (3-feet) increments at 3-
minute intervals starting at just above the bottom of the intake opening. Water from the pump
was discharged just below the water surface into a 2.4-m (8-feet) long, 794-pm mesh
suspended from the side of the barge.
Water was sampled through a 7.6-cm (3-inch) gate valve on the condenser and discharged into
an 800-pm mesh plankton net suspended in a 208-L (55-gallon) drum. Two consecutive
samples averaging 192 minutes each were collected on each night. Though longer duration
than the other sampling techniques, the volumes of water sampled were similar.
All four gear types were compared to Tucker trawls taken from the intake embayment and used
to estimate density and length frequency distributions of larvae potentially susceptible to
Duke Energy 154

Entrainment Charact
Marshall Steam S potion Study Plan Fn
' entrainment. The Tucker trawl had 710-pm mesh with a 1 m2 (10.8 feetf) effective opening.
Samples were collected perpendicular to the intake (starting as near as possible to the intake).
' Samples were collected at two depths—surface to the top of the intake structure opening (about
4.6 m [15 feet]) and from the intake opening to the bottom of the intake structure (4.6 m [15 feet]
to 14 m [50 feet]). Each stratum was towed obliquely. Trawl durations were 2 to 5.5 minutes and
collected every two hours starting at about 30 minutes after sunset and extending until sampling
with all the other gear types was completed. Two or three sets of tows were made on each
night.
Results were variable by gear type (Table C-2). Few shad were collected by pumped sampler
on 6 and 7 June due to a ripped seam in the net. Trawl samples from the upper stratum were
greater in number than the lower stratum. While not identified by the authors, it should be noted
that it is difficult to measure the flow sampled with the fine-mesh screen approach. Differential
' open area between the fine-mesh overlay and the surrounding coarse-mesh panels would likely
have resulted in flow diverting to the coarse-mesh panels or gaps between panels, side seals,
and the screen boot preventing accurate measurement of volumes sampled. In most cases, the
1 samples collected at the condenser tap were higher than what was observed in the lower trawls.
There were no significant differences in ichthyoplankton density by date (P = 0.83), but there
were differences by gear type. Mean densities of the pumped samples were not significantly
different from the tap samples (P= 0.60). Stationary nets and fine-mesh screen collections were
significantly different from one another and from the pumped and tap collected samples (P <
0.05). Length frequencies between the pump and tap samples were not significantly different
from one another. Mechanical damage from collection was minimal for all four techniques with
the highest observed damage associated with pump passage (3 percent unidentifiable). Tap
sampling was considered for use at McGuire for the 2016 entrainment program, but was
eliminated because of concems with access to secure locations within the power plant. This
study indicates that the selected method for entrainment monitoring (pumped sampling) is
statistically no different than measuring ichthyoplankton density at the condenser tap and better
than a streamed net.
1
1
1
1
Duke Energy I 55

I Entrainment Characterization Study Plan ���
Marshall Steam Station
ITable C-2. Total Number (N) and Mean Densities (MD) (mean number of shad/ 1,000 m3) of
All Shad Collected with Comparison Gear and Shad <28 mm Total Length Collected with
I a Tucker Trawl on Lake Norman, North Carolina, 6-10 June 1982, with Average Volume of
Water Filtered per Sample (m3) (Leonard and Vaughn 1985)
Net Screen Pump Tap Upper Trawl Lower Trawl
I
Date © MD N MD N MD © MD N MD N MD
Jun 6 0 0.0 2 -° 1° -C 5 56.4 164 195.3 27 27.9
I Jun 7 0 0.0 8 34.0 6°� idlt
182.0 590 I0 ,19 70.7
Jun 8 1 4.7 15 53.0 .38 : 103.0 659 '31536.1 751: 43.1
Jun 9 1 7.0 11 43.0 1130 '1'"196.9 4115 346.1 511 1.66111 8U 76.4
Jun 10 0 5.0 10 32.1,1 36 ji 82.4 4 82.0 279 406.5" 1 85.7
ilk, w1
.
Tota 4 611 16. mu PI
Ave.
IIIMII
.317,;
Sample ..,
I
Volume
(m3)
i a-Unable to calculate volume
b-Number not considered valid due to malfunctioning equipment
c-Density not calculated on invalid data
1 Taggart, C.T. and W.C. Leggett. 1984. Efficiency of Large-Volume Plankton Pumps, and
Evaluation of a Design Suitable for Deployment from Small Boats. Canadian Journal of
IFisheries and Aquatic Sciences 41(10) 1428-1435.
Taggart and Leggett (1984) identified and evaluated five major studies that compared the
I efficiency of large-volume pumps (defined as withdrawing > 0.5 m3/min [132 gpm]) and nets
(Table C-3). In addition the authors evaluated a boat mounted large-volume pumping system.
I Taggart and Leggett (1984) found, in general, that among the reviewed studies, densities of
organisms sampled with the pumps was equal to or greater than the densities in the towed net
samples, but there were differences based on length classes and time of day. The authors
1 pointed out that the gear configurations relative to towed net diameters, pump intake diameter,
the presence of intake screens, intake velocities and mesh sizes were highly variable between
the studies, which made it difficult to compare. The authors were particularly critical of the lack
Iof accurate flow measurement used in many of these studies. Despite these challenges, no
systematic biases were detected.
In addition to the literature review, Taggart and Leggett (1984) tested a large-volume pump
system with standard plankton nets. The boat mounted system used a 22.2-cm (8.7-inch)
I impeller that could pump up to 1.7 m3/min (450 gpm) depending upon head. Divers confirmed
the inlet was oriented into the direction of travel. Simultaneously to pump sampling, a 0.5-m
(1.6-feet) diameter 2-m (6.6 feet) long 80- and 153-pm mesh standard plankton nets were also
I
fished. Three sets of comparisons were made. In 1981 an 80-pm net was towed immediately
IDuke Energy i 56

Entrainent
Marshall Steam Station
Study Plan Ln
below the surface 2 m (6.6 feet) behind the pump intake. In 1982 and 1983 a 153-pm net was
towed immediately below the surface 13-15 m (43-50 feet) astern of the pump sampler. The
pump intake was maintained at a depth of 0.25 m (0.8 feet)for all comparisons.
The authors concluded that the nets and pumped sampler were equally effective in capturing
capelin (Mallotus villosus) larvae (5-mm length), herring (Clupea harengus) larvae (9-mm
length), large copepods (>750 pm), small jellyfish, and hyperiid amphipods, despite only
sampling 8 percent of the water volume sampled by the nets. In addition, the pump was more
' efficient at collecting crab zoea and megalops larvae and efficiency of collecting euphausiids
and chaetognaths increased as their natural densities increased. Nets were superior in the
capture of fish eggs (primarily cunner, Tautogolabrus adspersus), possibly due to the vertical
distribution of eggs in the water column. The average length of capelin larvae captured by
pumping was consistently 0.2 mm longer than that of larvae taken in nets, but the length—
, frequency distribution of larvae sampled was similar to that of larvae entering the pelagic
environment.
The very fine meshes and small organism sizes likely contributed to the high collection
efficiency of the pumped sampler used by Taggart and Leggett(1984)as other studies reviewed
here indicate it is the larger more motile life stages that tend to be collected more efficiently by
' nets than by pumped samplers. Like other studies that have towed the pumped sampler inlet
through the water, it is unclear whether these results are directly applicable to the application
being implemented by Duke Energy. Despite these uncertainties, this study seems to support
use of pumped samples for estimating ichthyoplankton and zooplankton densities.
1
1
1
1
Duke Energy 157

IIIIII IIIII 1111111 IIII. 1111. 11111 Ille 111111 1.111 111.1 111111 1.1. dill 1011 111111 a a 1.1111
Entrainment Characterization Study Plan F01
Marshall Steam Station
Table C-3. Summary of Major Studies Designed to Comparatively Evaluate the Sampling Efficiency of Various Large-Volume
Pumps and Tow Nets (Taggart and Leggett 1984)
Pump
Tow or current
Suction speed Volume sampled
Type and Tow net and (m/s) (m3)
mesh size Flow Diameter Velocity mesh size
Reference (lam) (m3/min) (m) (m/s) (pm) Pump Net Wasp Net Gear comparison protocol
Aron 1958 Centrifugal, 1.514 0.076 5.55 0.5-m-dia.std., 1.7' 1.7' 15 200 50 paired hauls"near
544 silk 476 nitex surface"(marine)
Posner and Rohde Madan propeller, 8.6 0.20 4.60 0.5-m-dia.std., Local current 86' 44' 111 paired stationary
1977 500 nitex 500 nitex 0.4 samples at 4.5,8,and
9 m(riverine)
Gale and Mohr Open impeller 2.5 0.10 5.30 0.24 x 0.54-m rect., Local current 7 stationary sets of 4 pump
1978 centrifugal, 400 x 800 nitex "moderate-strong" - - and 8 net replicates
400 x 800 nitex (0.24)° (0.92)° at surface and bottom
(riverime)
Leilhiser et al. Fish treader, 2.1 0.15 1.92 (a)I-m-dia.cyl.-cone, Local current (a)62 121 (a)10 stationary pairs at
1979 335 obex 335 nitex (a)0.26 0.5-1.5 m depth
(riverine)
(b)as in(a)and (b)0.30 (b)64 414 (b)as in(a)above
0.5-mdia.std., 64 105
363 nitex
Cada and Loar Open impeller 1.10 0.076 4.04 0.5-m-dia. "Slowly" 1.2-1.9 17 85-100 3 sets of 3 replicates not
1982 243 niter (0.021)° (l.I)° Hensen net, paired in time at
243 mirex 0-0.S m depth(riverine)
'Estimated from data provided in paper referenced.
°Measured at inhke,which differs in size from suction hose.
Duke Energy 158

Entrain
ent
Marshall SteaCharacterization
r Sboon Study Plan
' Results
' The literature review, as described in detail above, indicates that there have been several
studies that have compared the effectiveness of pumps and net sampling. In general, the results
have been equivocal with no clear pattern of one gear type out performing the other. The most
extensive power plant entrainment gear comparison studies were undertaken at the Indian Point
Generating Station on the Hudson River (EA 1978; 1979; 1981; King et al. 1981; NAI 1982;
1987—as cited in EPRI 2014)13. The results of these studies indicated no consistent differences
in estimated densities between the two gear types (EPRI 2014). There is some indication that at
relative low velocities, high-volume pumped samples collect higher densities of ichthyoplankton
than nets (Leithiser et al. 1979). In a southeastern reservoir(Tennessee), Cada and Loar(1982)
found that for most length classes there were no significant differences in density between
pump and net samples. The data suggest that avoidance is more common during daylight hours
in clear water when organisms can observe and avoid the gear. At night and/or in turbid water
avoidance is minimized (Cada and Loar 1982; Harris et al. 1986). Petering and Van Den Avyle
(1988) collected significantly lower densities of fish larvae in a Georgia reservoir using a
pumped sampler as compared to nets, but there are concerns with the methods and gear, as
described in detail above, that could account for some of these differences. Leonard and
Vaughn (1985) sampled Threadfin Shad (Dorosoma petenense) and Gizzard Shad (Dorosoma
cepedianum) using pumps, a streamed net, a fine-mesh panel on a traveling screen, and a tap
at the condenser at the McGuire Nuclear Station and reported the highest density at the
condenser tap where the water was well mixed. Leonard and Vaughn (1985) reported the
highest rate of damaged larvae from the pumped samples, but damage was s 3 percent. Gear
avoidance occurs with both pumps and nets and increases with increasing fish length, which is
likely a result of increased swimming ability, maturing fish sensory systems, and avoidance
behavior.
Caution should be used when applying the results of specific previous gear efficiency studies to
the Duke Energy entrainment program because of the variability in gear types and sampling
techniques used in these studies (e.g., net type and shape; mesh sizes and material; methods
' of gear deployment; deployment location; flow metering; inlet orifice size, shape, and
orientation; inlet velocity; impeller size, shape and material; mechanism for suction [vacuum,
diaphragm, centrifugal]; and pumping capacity). Instead, the weight-of-evidence across
' scientifically sound studies should be used which supports pumped samples as being generally
equal in terms of collection efficiency to towed and streamed nets.
' Conclusions
Nets and pumps are the two primary methods by which ichthyoplankton densities are estimated
at power plant intakes. Pumped samplers have been used successfully in a wide variety of
" During the peer review kick off meeting, a biological peer review mentioned an Indian Point Generating study evaluating the
effectiveness of fine-mesh wedgewire screens to reduce entrainment.As part of that study,samples were collected by nets and
pumps. That study was not reviewed here for several reasons: a) the study was not designed to compare gear types; b)the
sampler used for pump sampling at Indian Point was a unique design that is dissimilar from what is being proposed at the Duce
Energy facilities;and c)Indian Point is part of on-going 316(b)-related litigation and those study reports are not readily available to
the public.
Duke Energy 159

Entrainment Characterization Study Plan
Marshall Steam Station
' deployment conditions at power plants in the U.S. routinely since the 1970s. This method has
been well vetted by regulatory agencies including in the northeastern U.S., which along with
' California, has historically applied the most stringent 316(b) requirements in the nation. In
addition, pumped samplers continue to be accepted for 316(b) monitoring as evidenced by the
list of states where these samplers are deployed or approved to be deployed to support 316(b)
evaluations under the final Rule. Importantly, the final Rule became effective October 14, 2014
and many facilities have not yet chosen the method by which they intend to sample entrainment
' and the list of states approving pumped samplers is likely to increase.
Studies that have evaluated nets and pumps have shown no clear pattern of one technology
outperforming the other. The results from published and unpublished industry studies are
' equivocal with some examples suggesting that pumped samples outperform netted samples
and vise-versa.
Perhaps the greatest impediment to determining the collection efficiency of gear from existing
data is the lack of standardized techniques for netting (e.g., net type, net shape, mesh sizes,
mesh materials, methods of deployment, deployment location, and flow metering) or pumped
samples (e.g., orifice size, orifice orientation, inlet velocity, impeller size, impeller shape,
impeller material, mechanism for suction [vacuum, diaphragm, centrifugal], deployment location,
and pumping capacity). These factors, along with site-specific hydraulics and life history
characteristics of the early life stages of fish sampled, are likely to impact the performance of
both collection systems. Further, short of a duplicative sampling with both sampling systems in
tandem, it is unlikely that a study program could be developed that would provide the necessary
data to develop a universal correction factor that could be used to quantitative adjust
entrainment estimates.
We acknowledge that no system for sampling ichthyoplankton densities is perfect given the
patchy temporal and spatial distribution of fish eggs and larvae and inherent biases of all gear
' types. That said, the available data indicate that pumped samplers can be as efficient as nets.
Because entrainment monitoring by pumped sampling is commonly used, widely accepted, and
has practical advantages over using nets at the Duke Energy facilities, our recommendation is
that no gear efficiency testing is necessary or warranted.
1
' Duke Energy 160

ent
Marsha I Steam Station
Study Plan
References
Aron, W. 1958. The Use of Large Capacity Portable Pump for Plankton Sampling, with Notes on
Plankton Patchiness. Journal of Marine Research 16: 158-173.
' Bowles, R.R. and J.V. Merriner. 1978. Evaluation of ichthyoplankton sampling gear used in
power plant entrainment studies. In L. Jensen (ed.). Fifth National Workshop on
Entrainment and Impingement. Pp. 149-158.
111 Cada, G.F. and J.M. Loar. 1982. Relative Effectiveness of Two Ichthyoplankton Sampling
Techniques. Canadian Journal of Fisheries and Aquatic Sciences 39(6): 811-814.
' Ecological Analysts, Inc. (EA). 1978. Indian Point Generating Station Entrainment Survival and
Related Studies 1977 Annual Report. Consolidated Edison Company of New York, Inc.
EA. 1981. Indian Point Generating Station Entrainment and Near Field River Studies: 1979
Annual Report. Consolidated Edison Company of New York, Inc.; Power Authority of the
' State of New York.
Elder, J.A., J.W. Icanberry, D.J. Smith, D.G. Henriet, and C.E. Steitz. 1979. Assessment of a
Large Capacity Fish Pump for Sampling Ichthyoplankton for Power-plant Entrainment
Studies. California Cooperative Oceanic Fisheries Investigation 20: 143-145
Electric Power Research Institute (EPRI). 2014. Entrainment Abundance Monitoring Technical
Support Document: Updated for the New Clean Water Act §316(b) Rule. 3002001425.
EPRI, Palo Alto, CA. 156 pp.
Gale, W. R, and H. W. Mohr, Jr. 1978. Larval Fish Drift in a Large River with a Comparison of
Sampling Methods. Transactions of the American Fisheries Society 107:46-55.
' Gibbons, S.G. and J. H. Fraser. 1937. The Centrifugal Pump and Suction Hose as a Method of
Collecting Plankton Samples. Journal du Conseil / Conseil Permanent International pour
' ('Exploration de la Mer. 12: 155-170.
Hams, R.P. L. Fortier, and R.K. Young. 1986. A Large-Volume Pump System for Studies of the
Vertical Distribution of Fish Larvae Under Open Sea Conditions. Journal of the Marine
Biological Association of the United Kingdom 66(4): 845-854.
King, L.R., B.A. Smith, R.L. Kellogg, and E.S. Perry. 1981. Comparison of ichthyoplankton
' collected with a pump and stationary plankton net in a power plant discharge canal. In L.
Jensen (ed.), Issues associated with Impact Assessment, Fifth National Workshop on
Entrainment and Impingement. Pp. 267-276.
Leithiser, R.M., K.F. Ehrlich, and A.B. Thum. 1979. Comparison of a High Volume Pump and
Conventional Plankton Nets for Collecting Fish Larvae Entrained in Power Plant Cooling
Systems. Journal of the Fisheries Research Board of Canada, 1979, 36(1): 81-84.
Duke Energy I e1

Entrainent
i Marshall SteamStation ion Study Plan Ln
Leonard, T.J. and G.E. Vaughn. 1985. A Comparison of Four Gear Types to Measure
Entrainment of Larval Fish. Proceedings of the Annual Conference of the Southeast
Association of Fish and Wildlife Agencies 39: 288-297.
Normandeau Associates, Inc. (NAI). 1982. Gear comparability study for entrainment sampling of
juvenile fish at the Indian Point Station, 1981.
NAI. 1987. Indian Point Generating Station Entrainment Abundance Program 1985 Annual
Report. Prepared for Consolidated Edison Company of New York, Inc. and New York
Power Authority.
Petering, R W; Van Den Avyle, MJ. 1988. Relative Efficiency of a Pump for Sampling Larval
Gizzard and Threadfin Shad. Transactions of the American Fisheries Society 117: 78-83.
Taggart, C.T. and W.C. Leggett. 1984. Efficiency of Large-Volume Plankton Pumps, and
Evaluation of a Design Suitable for Deployment from Small Boats. Canadian Journal of
Fisheries and Aquatic Sciences 41(10) 1428-1435.
IWaite, S.W. and S.M. O'Grady. 1980. Description of a new submersible filter-pump apparatus
for sampling plankton. Hydrobiologia 74(2): 187-191.
Welch, P.S. 1948. Limnological Methods. McGraw-Hill Co., New York. 318 pp.
I
I
I
I
I
I
i
I
I
Duke Energy 162