20070812 Ver 1_US Geological Survey Draft comments_20070901

DRAFT
Review of Selected Documents
Related to Flooding at City of Salisbury Facilities
on the Yadkin River upstream from High Rock Dam, North Carolina
Jerad Bales
U.S. Geological Survey
September 2007
prepared in cooperation with
North Carolina Division of Water Resources

DRAFT
Introduction
The City of Salisbury has operated awater-supply intake at the confluence of the Yadkin
and South Yadkin Rivers since 1917 (fig. 1). The City maintains that the operation of
High Rock Dam, completed in 1927 and currently operated by Alcoa Power Generating
Inc. (APGI), has created a sediment delta in the headwaters of High Rock Reservoir, and
that this delta has caused (1) increased flooding of the water-supply pump station, (2)
sediment deposition around the pump intake, and (3) increased flooding at a wastewater
treatment plant pump station located near the mouth of Grants Creek (fig. 1).
Figure 1. Yadkin River, South Yadkin River, Grants Creek, and upstream end of High
Rock Lake, North Carolina
APGI is currently (2007) seeking to relicense four hydroelectric stations, including High
Rock Dam, on the Yadkin -Pee Dee River. The license for the hydroelectric stations and
the reservoirs that support them is granted by the Federal Energy Regulatory Commission
(FERC). The North Carolina Department of Environment and Natural Resources
(NCDENR) is participating in the relicensing process. NCDENR has signed a
Relicensing Settlement Agreement with APGI, and must subsequently respond to an
application by APGI for a section 401 water quality certification for the stations.
The City of Salisbury has requested that APGI address the issue of flooding and
sedimentation at the City's facilities on the Yadkin during the FERC relicensing process
2
~ 1 ~ a Mi9es

DRAFT
settlement agreement. APGI maintains that their operation is too far downstream to have
effects on the City's pump stations, that the Yadkin has a naturally high sediment load,
and that the pump station is in the floodplain. Because of APGI's position, Salisbury
retained consultants to develop hydraulic and sediment transport models for the reach in
question. The final report was submitted to FERC on February 26, 2007, as part of
Salisbury's relicensing scoping comments. APGI also has funded studies of their own,
including a review of the City's documents.
Because of the complexity of the issues, NCDENR asked the U.S. Geological Survey
(USGS) to review both APGI's and the City of Salisbury's documents related to flooding
at the City's facilities. The independent review is needed in order to assist NCDENR in
their review of APGI's application fora 401water-quality certification and subsequent
renewal of the FERC license for High Rock Dam.
Purpose and Scope
The purpose of this report is to document a review of the hydraulic and sediment
transport models developed by the City of Salisbury, APGI, and FERC, as well as related
data and information. The objective of the review was to determine if the documents
submitted by Salisbury clearly demonstrate that the presence of High Rock Dam has led
to an increase in water levels at Salisbury facilities, or conversely if APGI documents
demonstrate that High Rock Dam has not had an effect on water levels at Salisbury
facilities. Documents which were included in the review are listed in the following
section. The review included a site visit to the Yadkin and South Yadkin Rivers and
downstream to High Rock Lake. No new data were collected as a part of the review, and
the models developed by involved parties were not tested during the review. Some
historical discharge measurement notes and previously published reports were checked as
part of the review.
This report is structured as follows. First, the documents which were reviewed are listed.
Next, hydraulic modeling and related documents are reviewed. Sediment transport
modeling conducted by the City's consultant is then reviewed, along with comments
provided by APGI's consultant. A brief review of other documents provided by the City
and APGI is followed by a summary and recommendations.
Materials Reviewed
The documents reviewed during this study were as follows:
Salisbury Documents:
• Technical Report: High Rock Dam and High Rock Lake Sedimentation
Flooding Effects as Estimated using HEC-RAS Modeling, January 2006;
prepared by Hazen and Sawyer for Salisbury-Rowan Utilities. (Hereafter
referred to as SAL-1)
• Letter from V. Randall Tinsley to Secretary Magalie Salas, Federal Regulatory
Commission, February 23, 2006 (SAL-2)

DRAFT
• Numerical Sedimentation Investigation, Yadkin River, North Carolina,
February 20, 2007; prepared by R.R. Copeland, Mobile Boundary Hydraulics
for City of Salisbury (SAL-3).
• High Rock Dam and Sediment Delta Flooding and Sedimentation Effects
(1927 - 2058) on City of Salisbury Critical Infrastructure, February 2007;
prepared by Martin Doyle for City of Salisbury (SAL-4).
• Letter from R.R. Copeland and W.A. Thomas to Matt Berhardt, Salisbury-
Rowan Utilities, May 8, 2007, in response to Williams Affidavit (SAL-5).
• Letter from Martin Doyle to Matt Bernhardt, Salisbury-Rowan Utilities, May
9, 2007, in response to Shiers Affidavit (SAL-6).
Yadkin / Alcoa Power Generating, Inc. Documents:
• Review of January 1998 Flood of Yadkin river, February 1998; prepared by
Stone & Webster for Yadkin, Inc. (hereafter referred to as APGI-1).
• Yadkin Hydroelectric Project-FERC No. P-2197-073, Alcoa Power
Generating Inc., Responses to Federal Energy Regulatory Commission's
September 14, 2006 and November 22, 2006 Additional Information Requests;
April 6, 1998; prepared by Julian Polk, Yadkin, Inc. (APGI-2).
• Letter from Mr. Gene Ellis, Licensing and Property Manager, APGI, to
Secretary Magalie Salas, FERC, December 13, 2006; response to the
Additional Information Request for Yadkin Project (APGI-3).
• Affidavit of David Williams, Ph.D., P.E., March 26, 2007 (APGI-4).
• Affidavit of Paul F. Shiers, P.E., March 27, 2007 (APGI-5).
• Consolidated answer of Alcoa Power Generating Inc. to petitions to intervene
and comments in response to Scoping Document 1; prepared by LeBoeuf,
Lamb, Greene, and MacRae LLP, Counsel to Alcoa Power Generating Inc.
undated (APGI-6).
• Sediment Fate and Transport Report, Final Report, November 2005; prepared
by Normandeau and Associates, Inc. and PB Power (APGI-7).
FERC Documents:
• Letter from Jerrold Gotzmer, Director, FERC to Mr. Julian Polk, Yadkin, Inc.,
dated March 11, 1998 (hereafter referred to as FERC-1).
• Letters from Jerrold Gotzmer, Director, FERC to Mr. Ron Qualkenbush, dated
May 6, 1998 and May 21, 1998 (FERC-2).
• Hydraulic Modeling Report for Yadkin Project North Carolina, Alcoa Power
Generating, Inc. (APGI), June 2003; prepared by Federal Energy Regulatory
Commission (FERC)., and transmittal letter from FERC to Mr. David Treme,
Mr. Ron Qualkenbush, and Mr. Milton Crowther (FERC-3).
4

DRAFT
One-Dimensional Hydraulic Modeling
One-dimensional hydraulic models were developed by Stone & Webster for APGI
(APGI-1), FERC (FERC-2), and Hazen and Sawyer for Salisbury -Rowan Utilities
(SAL-1). Important aspects of these models are summarized in table 1.
Stone & Webster: In January 1998, a campground approximately 20 miles (mi)
upstream from High Rock Dam was flooded during high flows. The Stone & Webster
model was created to evaluate "the effect that the operation of High Rock Dam may have
had on this flooding" (APGI-1). The model extended from High Rock Dam to the
confluence of the Yadkin and South Yadkin Rivers, 19.4 miles (mi) upstream from High
Rock Dam (or RM 19.4). The model uses an estimated inflow hydrograph at the
upstream boundary, and a range of downstream water-surface- elevations at the dam.
Cross sections were scaled from topographic maps and were assumed to trapezoidal in
shape. The report does not specify how the bottom and/or top width of the cross sections
were determined, nor how the side slopes were determined. Apparently more detailed
cross-sectional data were available (APGI-7) from 1997 surveys, but these data were not
used. The channel bed slope was based on the assumption of a linear slope between the
dam and RM 18.1; the slope was assumed to be zero between RM 18.1 and RM 19.4. It
does not appear that the model was calibrated.
The report concluded that for existing conditions, the water level at RM 19.4 was
independent of the lake level at High Rock Dam. The report also concluded that water
levels at RM 19.4 were affected by a channel constriction at the railroad bridge near the
NC 150 bridge and by a narrow bend in the river at about RM 18.4.
The conclusion regarding the relation of water levels at High Rock Dam and water levels }
at RM 19.4 for current conditions is reasonably supported by the modeling for current I I
conditions. The conclusions could be strengthened by conducting a sensitivity analysis
of the effects of the assumed channel geometry and bed slope on simulated water levels
at RM 19.4. There is no evidence given in the report for the conclusion that the channel
constriction and narrow bend in the river controlled water levels at RM 19.4 during high
flows. A sensitivity analysis of the effects of the expansion and contraction coefficients
and simulations without the bridges in place are needed to give credibility to the
conclusion regarding the effects of bridge constrictions. It is not clear how results from
one-dimensional model were used to make conclusions regarding the effects of the bend.
Doyle (SAL-4) gives a brief discussion of HEC-RAS modeling that was conducted to
evaluate the effects of the bridges on water levels at the Salisbury facilities. This analysis
indicates that the upstream effects of the bridges are minimal, even under high flows.
There is essentially no detail given in that report regarding any details of the model, so it
is impossible to assess these conclusions.

L.L

~' ~.
~ ~
bn v1
~ ^ty
a>n `~
W ~
x
a~
a~
w
a~
~ ~ a
W
~~ ~
~ ~
a~ ~
.~
w
~ °
~ ~
.., x
,~ U
O
3 ~
~ x
~,
-~ `° ,a
~ .c
N~ 3
O ~
~ ~ C
N C
~~7 ~
'~ N
a~ x
o ~
0
0 0
..~, • ~
U ° ~
U y
~ ~
a~
c~
F~
.-.
~ o~
~ rn ~ ~ o
~ o ~ ~ ~ o ~ M ~, o ~ ~
~ ,~ bn a~ ~ o ~ ~ ~ ~ ~ ~ rn ~ rn
~ ~ ~ a ~ o ~ ~ ~ ~~ ~~ ~,~
° ~ o ° ~
.,.., ~ cd O ~ i N
N a~ >' ~ vOi ~ ~ ~ O U ~+ ~ ~" '~ .O
(~ SUS". ~ 'J"' vii U °~ ¢+ O ~i O ~ ~ N ~
... ~ ~
u, ~

O ~,
~ ~ a~
~
~
U >
x
-^7 O
~ ~ O L."
,-~ •
~ ~•
O
O ~ y ~ ~
0 0 ~~ ~ `~' ~ ° ~ ~' ~ 3

~
~
~"
3
~ -~
`~ ~
o i
'
~' .~•~ as N
° > ~ ~ ~ ~ °~ ~ o °~'
°' °
'
~ a' ern ~ o.~ o > ~
a ~.~
o• ~ ~ ~ ~ o
° ~ o
~
~ .°
~• .
a~ ~ ~
_
• ~ •~ ~ ~ ~
O ~
•r o ~ b o
~ o ~
~
N ~ , `1. M ~
,
°
C ~,~ ~
O
0 ~
~r
0
° ~ U cC
t ~ ,~
~
° o `'-' ° W ~ '~ ° ~ ~ ' an
N v -~ '~

'N ~, ~
Q ~ ~ ~ ~ ~ ~ y ~ ~ ~
~

~~'~ ~~ ~ ~ ~ ~~ ° ~ o en°
_
x~ '~ cC O p y N ~ 0 0 N~ 'd
~ O cd
~ O ~"
U
O ~
~ v~ U ~ ~ ~
U ~
' ~ , ~ x ~+ ~ U
~
N ~ c
~ 'd
~ bA ~"
U ,~ fYi b `~
~ W N ~
0
~ ~ Z ~ ~ >
O N ' ~' O
O
^" O N ~
~ ~ .~
0 w ~ U ~
~ a N ~ ~ O ~
o ° ~ ~ .
~ ~ ~~.~ ~ '3 ° a 3
•
c~ °o ~ o ~ a, o ~ c~ ~ ~ o

N `~ +rr
W o ~
Qi U cd ~ ~
~ ~ •
~ ~ .~ bA

o
c~ i
•
~ ~ ,~ o

a U U o° ,~ v o
°~
v~ cw U ~ c
n

L.L
o ~ -~
V ~ ~ ~ ~
~ ~ ~ ~„ N ~ ~ i..i ~ i-i Vi ~ (Uj
~ G~ ~ O ~ 3 ~ II O ~ O ~ O ~ ~
~ 3 ~ ~
~ ~ ~ ~ ~ ~' ~ ~~ ~~ ob o
-~ ~ ~ ~ ~ .~ ~ ~ ~ b ~ ~ ~ v ~'
U ~ ~ U O U ti ~ cn ~ ~ cd
Gtr ~, "' ~ ~ ~" ~ ~ ~ ~ ~ ~ ~ y ~ ~
,~ N Q Vj L: ~ cd '~ ~ ~ ~ ~ ~O O ~' O
~ ~ ~ ~ N ~ ~ ~' ~ p ~ ~ ~ ~ ~ :~ y Q
~ ~ ~ ~ O ~ ~ ~ ~ O O ~ O ~ O
3 x~~ z a~ ~ ao z~ z~ Z a.~, Z
~° ~ ~ ~ ~
`nom a. ~ -~ ~ 3 0~ ~ ~ ~ ~' M
~ b o x ~ ~ ~ ~ ~ o ~ ~ ~ ~ W
w 'O V O ~ N ~ ~ ~ .~ ~ ~ J, ~ ~ ~ bUA ..
t. ~ ^" ~ ~ ~ ~ i ~ ~ N
0 o U-d U .° ~°o ~° ~ ~ ~ ~-d ~-~ o o ~
,~, ~, i`ri 4-a U Zi V] ¢, U v~ U O O z ~ ,7.. ~ N U ..
~ ~ d; U M ~
~ ~, rn ~ o z ~ a°,
d
~ ~ ~ ~
~ ~~ Q
~ ~ ~ o ~ ~ o ~ ~ o O ~' ~ ~ Z
O U U ~y
'C ,4? v> > ~ 'C O M ~ N
U U ~ cd •• O ~ Gy O cd ~ O O ~--~ ~ ~ ~ O ~
"'~
~ ~ O
t~ .~ '~=+
~, O
~Ui ~ ~ ~ p w N ~ O .O ~ ~ ^O "C
~ U ~ U ~ ~ L." ~ ~ ~ ~
[~ ~ U ~ ~ ~ W v ° U'~ ~

LL
1..1~

a~ ~
w ~
~ ~
~ b
w `~
~ ~
x
~~
'd .O
O ~
s..
N W
Q. y ~
b~A ~
• -. ~
~~
~ U
O
•s
~ x
~~
~°
~;
~"
O C/~
~ ~ i.r
~ ~
~~ 3
~x
~~ ~
°~ ~
~ o
.~ .o
Q., .~
U o
U
~~
~~
~~
~~
H
~
0 ~~

~ d
~,
~ a~ ~
b1
~ •~

' a, >,
b ~ 3 ~.~ ~, ~~

~
~ ~ ~ o a~ ~ ~ a~
~ ,~ U ~ ...~ Cd ~ ~ ~ ~ „~

~~. O
•.. bq
O O r-'
cC 4-i
O ~
~ ~ O ~
N

~
3 ~ ~ U
~

a
°
~
~~ ~,
~'
~ o ... Q.
~~
°
. -
o
z

~ ~~ ~
O ~ ~ "O ~. ~ O ,~
>

~ ~ ,~ O O ~ 4-, . ~ ~, 'C
U ~
O ~+ p ..~
~ y
~ V~ „~,,
O
~
~ O ~" ~ cd cn O
rr
~'`•' f' ! ~ +' ~ ~ N cn
~ 'x O
4
'CS O ~ ~ '~ O ~
-•i ..~
cH
o ~' ~ ~
~" ~ ~ U
. ~
°' °o z
O ~
~,

O N T O ~~'
.-.. • O V1
~' O •~ ~
~

°
~ ° ~
~~
o
~ z - ~

w ~
~ .~

~
"
~ ~

~
U ~ ~
.
cd ~ '~ U
'. 4-i
p~
~ ~ ~' '
~
~ N ~ U
i
~ C/~ ~.
v~ .
~
~ i~. O ~, v
~ ~
~
~ Q„
O L".
~ O ~
cd v~ ~.. ~"+
O O 4
.i
O
~~ +-~
~
~ ~ ~ r-.~ ~~
O ~ O O ~ ~ O
•
... ~ v
i

'
an ~
~
~ o U
"z ~
~
° ~ ~ . ~ N
o ° ~
3
Q ~

~ 3 ~x •
• z w
~
~~
~, ~ ~
~
o
•~ 0
•~

~, ~ •~
~° a
00

DRAFT
The results of the Stone & Webster modeling are not directly relevant to the issue of the
effects of the presence of High Rock Dam or the delta on flooding at RM 19.4. This is
because no simulations for pre-dam or pre-delta conditions were performed for
comparison with 1998 conditions. Moreover, simulations for conditions with and
without the bridges were not conducted, making it impossible to assess the effects of the
bridges on water levels at the Salisbury facilities.
FERC: FERC obtained the HEC-RAS model used by Stone & Webster (APGI-1) to
conduct additional one-dimensional hydraulic modeling investigations (FERC-3). The
purpose of modeling was to "Determine the influence of the High Rock Reservoir
elevation on the water surface elevation at the confluence of the Yadkin and South
Yadkin Rivers," and in particular during a 2003 flood. The model was extended
upstream to about RM 21.3, and additional cross sections were added to the model to
provide greater resolution at locations where FERC concluded that Stone & Webster
cross sections were too widely spaced. The model was used to simulate water-surface
profiles for the March 2003 flood for a range of water levels at High Rock Dam. FERC
used a digital elevation model to develop a topographic map that perhaps provided more
detailed cross-sectional information than was in the Stone & Webster model, but the
detailed data from the 1997 survey were not used in the model. The approach used by
FERC was much the same as that used in the Stone & Webster study, other than the . ,
source of the topography data and the flood to which the model was applied. As a result,
the conclusions from this study were essentially the same as for the Stone & Webster
study-water level at RM 19.4 is essentially independent of water level at High Rock
Dam for flows between 40,000 cfs and 70,000 cfs under current conditions. As with the
Stone & Webster study, no simulations were done for pre-dam or for 1917 bathymetric
conditions for comparison to current conditions.
Hazen and Sawyer: Hazen and Sawyer (SAL-1) constructed a HEC-RAS model for
High Rock Lake and a portion of the Yadkin River to "evaluate the effects of High Rock
Dam and dam-induced sediment accumulation on the frequency and magnitude of
flooding ... at the Yadkin -South Yadkin confluence." This statement presumes the
conclusion that all of the sediment accumulation in the reservoir is "dam-induced." The
modeling effort was unique in two ways: (1) 1997 topography was used to construct
cross sections for current conditions model input and (2) pre-impoundment conditions
were simulated.
The 1997 topographic survey was conducted by APGI contractors when the lake level
was at 12 ft below full pool. The survey extended out from the full pool elevation about
0.25 mi. A topographic map with 2 ft contour intervals was prepared (APGI-7). Hazen
and Sawyer also simulated pre-impoundment conditions by using 1917 topographic data,
obtained from maps provided by APGI, and estimating channel cross-sectional shape
below the 1917 water surface. Simulations were made assuming the dam was not in
place and using 1917 topography.
Results indicated that the presence of the dam alone, with no sediment accumulation (i.e.,
1917 bathymetry), may have led to an increase in water levels at RM 19.4 for flows less
9

DRAFT
than about 80,000 cfs. At a flow of about 20,000 cfs, the increase may be about 6 ft
relative to pre-impoundment conditions. The increases relative to pre-impoundment
conditions were greater when simulations were made using 1997 topography/bathymetry,
with an increase of about 10 ft at a flow of about 20,000 cfs. Water levels for estimated
2038 conditions were linearly extrapolated by assuming that the average annual increase
in water-surface elevation for each flow rate during 1927 - 1997 continued through 2038.
Various aspects of the Hazen and Sawyer model are poorly documented, including the
Manning's n, expansion and contraction coefficients, and treatment of bridges for historic
conditions (i.e., did the 1917 simulations include the 2005 bridges, 1917 bridges, or no
bridges). In addition, like the Stone & Webster and the FERC models, the Hazen and
Sawyer model does not appear to have been calibrated or compared to observed
conditions. Nevertheless, this model seems to clearly demonstrate that the presence of 4 ~;
the dam and the changes in the bathymetry likely have had an effect on water levels at the
Salisbury intakes relative to the pre-dam conditions. The exact magnitude of that change
is less certain primarily because of uncertainties in current and historic bathymetry,
because model performance has not been tested through comparison to measured
conditions, and because all aspects of the model were not documented.
Doyle (SAL-4) summarized the work of Hazen and Sawyer (SAL-1) in his report. The
hydraulic modeling aspects of Doyle's report were subsequently reviewed by APGI's
consultant, Parsons Brinkerhoff (PB) (APGI-5). The PB review identified four major
areas of concern, which are listed below, along with some observations regarding the
concerns. It should be noted, however, that none of these criticisms by PB relate directly
to the issue of the effects of High Rock Lake on possible increased water levels at
Salisbury's facilities, but rather to some peripheral issues related to the details of the
analysis by Salisbury's consultants. Doyle (SAL-6) subsequently responded to PB's
criticisms (APGI-5).
The resolution of the topographic data used in Hazen and Sawyer's modeling, in
particular the pre-dam conditions, does not juste Doyle's conclusions. In fact,
Doyle does report water-surface elevations to the nearest 0.1 ft for future conditions.
These numbers should be used with caution. Copeland (SAL-3) does, however,
address the issue of uncertainty topography in the sediment transport modeling,
although the sediment transport model does not address the effects of High Rock
Dam on water levels at Salisbury's intake. Doyle (SAL-6) also notes that the Stone
and Webster's modeling (APGI-1) did not included an analysis of the sensitivity of
model results to topographic elevations.
Doyle calculated the "design flow " of 121, 000 cubic feet per second (cfs) at RM 15.2
rather than at the water-supply intake at RM 19.4. Doyle apparently did use records
collected at RM 15.2 to represent conditions at RM 19.4. The increase in drainage
area between the two sites is about 90 mil, or less than 3 percent of the total drainage
area. The at-station 100-year recurrence interval flow at RM 15.2 is 166,000 cfs
based on records collected 1896 - 1927 (Pope and others, 2001), which is less than
Doyle's "design flow." These issues, however, do not directly relate to the question
of the effect of High Rock Lake on water levels at Salisbury's intake, but rather to the
magnitude of the effect at selected conditions (100-year flow; "design flow," etc.).
10

DRAFT
As an aside, the methods used by Doyle (SAL-4) and Copeland (SAL-3) might
underestimate the 100-year flood at the confluence of the Yadkin and South Yadkin
Rivers. Pope and others (2001) provide methods for estimating flood flows at
ungaged sites. Greater use of at-station flood-frequency information for Yadkin
River at Salisbury and South Yadkin River at Cooleemee might also be helpful.
Doyle (SAL-6) asserts that inflows from Grants Creek would not affect peak flows
measured at RM 15.2, because Grants Creek likely would peak before the Yadkin
River. This argument is reasonable, although there are no data to support the
assertion for the particular case of the 1916 flood.
PB believes that Doyle mischaracterizes APGI's statements regarding causes for
sediment accumulation in the reservoir. This criticism may be valid but is not
relevant to the issue of effects of High Rock Dam on water levels at Salisbury's
intake. APGI has argued that sediment accumulation is caused, in part, by natural
bends in the river and that channel constrictions from bridges are contributing to
flooding at Salisbury's facilities. As noted above (and by Doyle, SAL-6), however,
the evidence presented by APGI to substantiate these claims is weak to nonexistent, at
least in the documents available for review. PB also indicates that Doyle does not
present any evidence that the river was in geomorphic equilibrium prior to 1927.
Indeed, Doyle does not present any quantitative evidence for this assertion, although
Copeland (SAL-3) does appear to have examined the stage-discharge relation for
Yadkin River at Yadkin College and found the rating to be stable.
^ There is an issue related to Doyle's characterization of flooding at the Grants Creek
wastewater pump station, and design of the pump station to account for possible
flooding. This issue is unrelated to the effects of High Rock Lake on increased
flooding at Salisbury's facilities.
One-Dimensional Sediment Transport Modeling
A one-dimensional unsteady sediment transport model was constructed "to evaluate the
potential impact of continuing delta aggradation in the Yadkin River at the upstream end
of High Rock Lake" (SAL-3). The study seems to presume that all of the sediment
accumulation is directly attributable to the presence of High Rock Dam. The model was
applied to evaluate four alternatives for reducing sedimentation at City of Salisbury
facilities near RM 19.4. The HEC-6T model, which is a proprietary version of the U.S.
Army Corps of Engineers HEC-6 model, was used for the simulations.
Cross sections for the sediment transport model were obtained from the Hazen and
Sawyer HEC-RAS model, from LIDAR data, and from bathymetric survey data.
Additional cross sections were surveyed between RM 21.3 and 31.3, and these cross
sections were assumed to be stable over the simulation period. The pre-impoundment
river bottom elevations used in the HEC-6T model were, however, adjusted upward from
those used in the Hazen and Sawyer HEC-RAS model. Fairly extensive sediment inflow
records were available for the inflow sediment model boundary conditions. Sediment
inflow rates for future conditions were assumed to be the same as those for historical
11

DRAFT
conditions, but the sensitivity of model results to this assumption were evaluated.
Assumptions regarding inflow hydrographs were required, and these assumptions seem
reasonable, although a number of other approaches could have been used for estimating
inflow scenarios.
The model was calibrated such that measured and simulated bed profiles for 1917 and
1997 were in general agreement. The model also was calibrated to water levels measured
at RM 19.4 for selected events. The sensitivity of model results to changes in sediment
inflows, the sediment transport equation used in the model, and to variability in bottom
elevation was evaluated. The model does not simulate natural adjustments in channel
width; it has been noted elsewhere (APGI-7, SAL-3) that channel widths in the region of
concern have changed over the last 70 years.
The model was applied to simulate the growth of the delta in the headwaters of High
Rock Lake during 1928 - 97. Changes in water-surface elevations at Salisbury facilities
for 3 flow conditions as a function of changes in bed elevation are presented for the
period 1920 - 2057.
Copeland's report provides a good documentation of the data and assumptions used to
construct the model and provides a reasonable assessment of uncertainty in model results.
The data, analyses, and simulations presented in the report add to current understanding
of sediment transport and accumulation in the High Rock Lake and the Yadkin River.
The report does not, however, address the issue of the relative effects of High Rock Dam
on sediment accumulation, because the model was not applied to simulate conditions
without the dam in place. The study also does not address the relative effects of the
bridges on sediment accumulation in the upper reach of High Rock Lake. As a result,
there is much improved understanding of sediment transport and accumulation, but there
is no new information on the effects of High Rock Dam or bridges on sediment
accumulation relative to what might have occurred naturally.
Copeland's report was reviewed by Dr. David Williams (APGI-4). The review identified
three major areas of concern regarding Copeland's sediment transport modeling.
Copeland and Thomas (SAL-5) subsequently responded to Williams' Affidavit.
The base condition (or pre-impoundment condition) was not modeled, so the effect of
High Rock Dam on sediment accumulation is not known. This is a valid criticism of
Copeland's work, just as it is for the Stone & Webster (APGI-1) and FERC (FERC-3)
hydraulic modeling. The actual increase in sediment deposition in the reach of the
Yadkin River between High Rock Dam and RM 19.4 attributable to the presence of
the dam cannot be determined from Copeland's original results.
As noted above, objective of Copeland's work (SAL-3) was "to evaluate the potential
impact of continuing delta aggradation in the Yadkin River at the upstream end of
High Rock Lake." As a result of Williams' criticism, Copeland and Thomas (SAL-5)
used their sediment transport model to simulate sediment thalweg and water-surface
elevations in the study reach for the pre-impoundment condition. Widths of the
channel cross sections were set equal to original river bank widths, presumably
12

DRAFT
determined from 1917 maps, and no erosion of bed or bank material was simulated in
the model.
Comparison of the simulations for pre-impoundment and current (or base) conditions
suggest that water-surface elevations for the 10-year flood are about 8 ft. higher now
than would be expected if the dam were never constructed. Similar results are shown
for other flood flows. Results also demonstrate temporal increases in water-surface
elevation at RM 19.4 relative to the pre-impoundment condition as a result of the
growth of the delta. Copeland and Thomas do not state whether their pre-
impoundment model included bridges.
The sediment transport model appears to have numerical instabilities when results
are displayed for each computational time step. The computational time step used by
Copeland was one day, except during high-flow events (SAL-3). Figures 1, 3 and 4
of Williams' report (APGI-4) shows oscillations of about 6 ft in the bed elevation at
RM 19.4. Similar oscillations are not evident upstream of RM 19.4. The largest
oscillations (Williams' figures 3 and 4) begin at about day 22,000 from the beginning
of the simulations, which would seem to correspond to sometime in the mid-1980's
(although this is an estimate based on material presented in the reports). Oscillations
appear to have an approximately annual frequency.
Copeland notes that his model simulated sand extraction during 1965 - 1984 and
1988 -end of simulation at locations near RM 19.4; sand was extracted by the model
at the end of each water year. It seems likely that the oscillations in bed elevation
presented by Williams are the result of the annual sand extraction in the model,
particularly because the oscillations seem to be present at RM 19.4, where extraction
occurred, but not at other nearby cross sections.
Copeland and Thomas (SAL-5) confirm that the oscillations are indeed are result of
dredging operations on thalweg elevation. Copeland and Thomas present additional
results to demonstrate that the model is not numerically unstable, citing guidelines
from a Corps of Engineers training document containing guidelines for application of
HEC-6.
Williams also points out that model results seem unusual during the first two years of
the simulation (Williams' figure 2), in that there are large changes in bed elevation
during this period. Some numerical models require a `warm-up' period because
initial conditions cannot be known throughout the model domain. Therefore, this
`warm-up' period is used to transport the estimated initial conditions out of the model
domain before boundary conditions begin to affect model simulations throughout the
model domain. Simulation results are normally not considered useful during this
`warm-up' period.
Copeland did not address the issue of initial conditions, a `warm-up' period, or the
bed elevation changes during the first two years of simulation in his original report
(SAL-3). Copeland and Thomas, however, do confirm in their response to Williams'
Affidavit that these changes are a result of model `warm-up' (SAL-5).
13

DRAFT
Other Studies
Normendeau Associates, Inc. and PB Power prepared a report on sediment fate and
transport in the Yadkin -Pee Dee River basin (APGI-7). The study provides a very
comprehensive review of literature on sediment transport in the Yadkin River basin.
Although the material in the report is quite interesting and well-documented, most of the
information is not directly relevant to the question of the effects of High Rock Dam on
water levels at RM 19.4. The study does present a comparison of the 1917 bathymetry
with 1997 topography in the upper 12 ft of the reservoir, although a full bathymetric
survey of current conditions has not been completed. Sedimentation patterns also were
analyzed by comparing the 1917 and 1997 topographic maps. The report notes "the
deepest portion of the river has narrowed" from the I-85 bridge upstream to the Yadkin -
South Yadkin confluence.
Doyle (SAL-4) presents an overview of reservoir sedimentation processes. A time series
of aerial photographs of High Rock Lake also are presented, and Doyle discusses
sedimentation patterns and the growth of the delta in the upstream end of the reservoir.
The river bed profile also is reconstructed for pre-impoundment and 2000 conditions.
This discussion is a very useful documentation of sedimentation in High Rock Lake. The
discussion by itself, however, does not unequivocally demonstrate that High Rock Dam is
responsible for all or part of the sedimentation. Doyle's arguments, the aerial
.,
photographs, and a general understanding of river morphology would lead one to
conclude that sedimentation in the portion of the Yadkin River that is now the upper end ~~
of High Rock Lake has increased as a result of High Rock Dam. The amount of increase
attributable to the presence of High Rock Dam and to the bridge abutments in the reach,
respectively, has not been quantified through either Doyle's presentation or Copeland's
modeling.
Doyle argues that the Yadkin River was in morphological equilibrium prior to 1917 and
continues to be in equilibrium, meaning that the overall sediment deposition and erosion
in the river is in general balance with the sediment supply. Doyle offers no quantitative
evidence of this assertion. Cross sections measured at USGS gaging stations (Yadkin
River at Salisbury at about RM 15.2 and Yadkin River at Yadkin College) could be
examined to document changes in the river geometry at measurement sections for both
pre-dam and post-impoundment periods, although the channel width at these cross
sections may be constrained by bridge abutments. Given the known sediment legacy
issues in the basin (extremely high sediment loads in the past, somewhat lower loads
now), the assumption of morphological equilibrium seems to require greater support than
what was offered in SAL-4.
Summary and Recommendations
The primary conclusions from this review are as follows:
^ The hydraulic models of Stone & Webster and FERC did not assess the effects of
changes in bathymetry on changes in flood levels at the confluence of the Yadkin and
14

DRAFT
South Yadkin Rivers. In other words, pre-impoundment conditions were not
simulated, so the effect of High Rock Dam and post-impoundment sedimentation on
water levels at RM 19.4 can not be evaluated from these studies.
^ The hydraulic modeling performed by Hazen and Sawyer seems to indicate that both
the presence of the dam in the absence of any post-impoundment sedimentation and
changes in bathymetry between pre-dam and 1997 conditions have resulted in
increased water levels relative to pre-impoundment conditions at Salisbury facilities
on the Yadkin River for a fairly wide range of flows.
^ The degree to which the dam and the changes in bathymetry have affected flood ;" %
levels at the Salisbury facilities relative to pre-impoundment conditions is open to
dispute because of uncertainty in topographic/bathymetric data and the absence of
calibration and sensitivity testing of the hydraulic models.
^ None of the 3 hydraulic models appear to have been calibrated to or tested against
measurements, and no sensitivity testing was reported.
^ Copeland's sediment transport model was calibrated to estimated bed elevations and
to water levels at RM 19.4. The model is well documented (except for issues related
to initial conditions and possible numerical instabilities) and provides a good
understanding of the expected growth of the sediment delta in the upper end of High
Rock Lake.
^ In a response to a criticism from APGI, Copeland and Thomas simulated thalweg and Kew
water-surface elevations for the pre-impoundment condition and seems to have ~~ I ~ ~`'
demonstrated that the presence of the dam and the growth of the delta have resulted in
increases in water-surface elevations at RM 19.4 over a range of flows, and that these
increases are expected to increase through time if current conditions remain
unchanged.
Some recommendations for further analyses and studies to improve understanding of the
relation of High Rock Lake to sedimentation in the lake and changes in the flood regime
in the upper part of High Rock Lake include the following:
Several elevation datums are used in the various documents, including the Yadkin
datum, "USGS datum," NGVD 29 datum, and some datums which are unspecified.
This leads to a good bit of confusion in interpreting results from reports. All parties
should agree that elevations will be specified in the standard NAVD 88 datum.
A detailed bathymetric survey of High Rock Lake and the Yadkin River from High
Rock Dam to 20 mi upstream from the dam would provide the correct information
needed for hydraulic (and sediment transport) modeling of the reach for current
conditions. Detailed topographic data of the region is available from the North
Carolina Floodplain Mapping Program
http://www.ncfloodmaps.com/default swf.asp).
The current bathymetry and topography could be combined to generate realistic
channel cross-sections for hydraulic modeling of current conditions. Pre-dam
15

DRAFT
conditions were already simulated by Hazen and Sawyer, although some clarification
and documentation of their approach is needed.
^ Copeland's model could be used to simulate sediment accumulation as a result of the
presence of bridges. Sediment accumulation attributable to the dam and to bridges
might then be distinguished from what might have happened for a natural condition.
^ In the absence of realistic bathymetry for current conditions, a sensitivity analysis
could be performed to determine the effects of changes in channel cross sections and
slope on final results. A sensitivity analysis for pre-dam conditions, for which it is
not possible to know channel cross sectional geometry, could be conducted. Likewise
effects of bridges could be evaluated for current and pre-impoundment conditions.
^ Cross sections measured at USGS gaging stations (Yadkin River at Salisbury at about
RM 15.2, South Yadkin River at Cooleemee, and Yadkin River at Yadkin College)
could be examined to document changes in the river channel for both pre-dam and
post-impoundment periods.
^ Sediment coring and age-dating methods on both the bed and the floodplain could be
investigated as a means to evaluate sediment throughout the reaches of interest for
both the post-dam and pre-dam periods. It might be possible to estimate
accumulation rates for both pre-dam and post-impoundment conditions. It may also
be possible to estimate geomorphic equilibrium conditions prior to and following dam
operations, as well as the effect of the legacy sediment on current morphological
conditions.
Flood data from existing and discontinued gages could be examined to determine if
there has been a change in the flood flow regime. Improved estimates of flood
recurrence intervals could be obtained by using the methods of Pope and others
(2001) and historical at-gage flood frequencies.
References
Pope, B.F., Tasker, G.D, and Robbins, J.C., 2001, Estimating the magnitude and
frequency of floods in rural basins of North Carolina-Revised: U.S. Geological
Survey Water-Resources Investigations Report 01-4027, 44 p.
16