NC0024406_WQ Impacts of Pumping Report_20030501

WATER QUALITY -IMPACTS OF
PUMPING DAN RIVER WATER TO
AUGMENT BELEWS LAKE LEVEL,
WINTER 2002-2003
May 2003
Keith A. Finley

Table of Contents Page
ExecutiveSummary............................................................................................................3
Introduction......................................................................................................................... 5
Methodsand Materials........................................................................................................6
Resultsand Discussion......................................................................................................6
Pumping Operation and Lake Elevation..................................................................... 6
Spatial Analysis of Water Quality Profiles.................................................................7
Historical Analysis of Water Quality Profiles........................................................................ 8
Water Clarity and Suspended Solids...................................................................................... 9
Comparison of Water Quality to Historical Trends...................................................10
Conclusions.......................................................................................................................11
LiteratureCited................................................................................................................. I I
Tables................................................................................................................................12
Figures...............................................................................................................................14
2

Executive Summary
Due to persistent drought conditions in the Southeast during mid-1998 through 2002, by
early 2002, groundwater levels, stream flows and reservoir levels were substantially
reduced. Duke Power engineers had projected that if the surface elevation of Belews
Lake dropped below 717 feet above mean sea level (msl), Belews Creek Steam Station
(BCSS) would become unavailable for generation due to condenser cooling water (CCW)
pump cavitation. Subsequently, in early 2002 structures were added to the CCW inlets to
suppress potential vortices, permitting the plant to operate down to a surface elevation of
715 ft msl. Yet, the drought remained in effect and future operability was by no means
ensured for this important base -loaded facility.
In the summer of 2002, Duke Power consulted with regulatory agencies and stakeholders
seeking permission to install a temporary pumping operation providing Dan River water
to Belews Lake. Permission was subsequently granted for a winter 2002-2003 seasonal
withdrawal and pumping of Dan River water to provide aid for Belews Lake water
management. The agreement included a request by the North Carolina Department of
Environment and Natural Resources (NCDENR) to monitor Belews Lake water quality
during the pumping operation to ascertain whether water quality impacts had occurred.
Pumping from the Dan River commenced December 23, 2002 and was terminated on
February 28, 2003. Belews Lake was restored to near full pond by early March, 2003.
Pumping from the Dan River had supplied an estimated 45 percent of the restored Belews
Lake water volume, with the remaining 55 percent supplied by increased winter seasonal
precipitation and inflows.
The December 2002 though March 2003 monitoring effort showed only localized effects
due to the Dan River pumping operation. In the immediate vicinity of the pump inflow
(near the spillway), slightly lower- water temperatures, coupled with higher dissolved
oxygen saturation, and slightly lower specific conductance were noted in vertical profiles
collected during pumping. Lake -wide, lower than normal water temperatures during the
winter 2003 season were primarily attributable to reduced generation at BCSS during
much of the period. Winter 2002-2003 climatology provided near normal temperatures,
but approximately 21 percent higher than normal precipitation.
3

Indicators of water clarity (i.e., Secchi depth, suspended solids and turbidity) also showed
only temporary, highly localized effects attributable to the pumping operation. Most
notable increases observed in particulate loading appeared to have resulted from
upstream watershed inflows, particularly during the February -March 2003 timeframe.
These watershed -related effects were most notable at the slightly uplake BCSS CCW
discharge zone.
Examination of sequential seasonal observations for solids, major nutrients, and minerals
tended to show that the winter 2002-2003 pumping operations had less overall impact on
the lake -wide water quality than did the preceding mid -1998-2002 regional drought. The
winter 2002-2003 Dan River pumping, combined with the coincident watershed inputs,
actually served to partially counteract drought -related increases that had been observed in
the concentrations of certain constituents, particularly calcium (and alkalinity), chloride,
and sodium.
In summary, water quality impacts from the Dan River pumping operation were: 1)
limited primarily to the immediate vicinity of the pump outlets in the lake; 2) of transient
duration (i.e., days as opposed to weeks); and 3) insignificant when compared to effects
of the 1998-2002 drought or seasonal storm runoff -related events.
0

Introduction
Drought conditions during mid-1998 through 2002 led to substantial decreases in
groundwater levels, stream flows, and reservoir levels in the Piedmont region. After
winter 2001 and spring 2002 precipitation failed to appreciably raise the Belews Lake
surface elevation, it became highly likely that with the absence of very substantial and
seemingly improbable amounts of rainfall, the lake would fall to critically low levels
during the normally drier, summer and fall of the year. Previous data had demonstrated
that in a dry year, lake levels could easily drop by 3 to 4 feet between the end of the
relatively wet spring season and the end of the calendar year. This eventuality did indeed
materialize in 2002, as the prolonged drought continued and losses of surface water to
groundwater and evaporation accelerated throughout the summer and continued beyond.
Duke Power engineers had projected that lake levels falling below 717 feet above mean
sea level (msl) would lead to Belews Creek Steam Station (BOSS) condensing cooling
water (CCW) pump cavitation and subsequent inoperability of the generating station. In
anticipation of this problem, Duke Power installed modifications at the CCW pump
intake bays early in 2002. Structures were added to the CCW inlets to suppress potential
vortices and to hypothetically allow the plant to operate down to a surface elevation of
71.5 feet without significant pump cavitation occurring. As the drought continued,
however, resulting in unprecedented low lake levels, it became clear that additional
measures would be needed to ensure station operability in 2003, should substantially
above-average rainfall not occur during that year.
In the summer of 2002, Duke Power Fossil -Hydro Environmental Health and Safety staff
consulted with regulatory agencies, including the North Carolina Department- of .
Environment and Natural Resources (NCDENR), North Carolina Wildlife Resources
Commission, US Fish and Wildlife Services, US Army Corps of Engineers, and other
stakeholders seeking permission to install a temporary pumping operation providing Dan
River water to Belews Lake. Permission was subsequently granted for a winter 2002-
2003 seasonal withdrawal of Dan River water to provide a source for Belews Lake water
management. This report provides a summary of Belews Lake water quality monitoring
conducted prior to, during and following the termination of the Dan River pump
operation.
5

Flow Multipliers for the Temporary Dan River Pumping Project
Introduction
Withdrawals will be variable based upon river flow. Pumping will not lower the river flow
rate downstream of the pumping site to less than 100 cfs, which is the sum of the 7Q10
flow rate and near -field municipal water withdrawals. The maximum pumping rate will be
98 cfs. Therefore, if flow is less than 100 cfs at the pumping site, there will be no
pumping, and if flow is greater, than 198 cfs, maximum pumping will occur. If flow is
between 100 and 198 cfs, the pumping rate will vary. A method of determining flow at
the pumping site is necessary to determine the correct number of pumps to operate.
There are currently two USGS flow gages on the Dan River in the vicinity of the
proposed pumping project. USGS Gage 02068500 near Francisco is located upstream
from the pumping site, and USGS Gage 02071000 near Wentworth is located
downstream of the pumping site. Retired USGS Gage 02069000 was located slightly
upstream of the pumping site. The upstream gage near Francisco will be used to
determine river flow rate at the pumping site. Data from all three gages was analyzed to
determine the appropriate multiplier between the Francisco gage and the pumping site.
A summary of the data used in the assessment is shown in Table 1.
Flow Rate Prosection
Flow rates along a river can 'be projected by a function of the ratio of the drainage areas.
x
Qz — Q1
*(DrainageAreal
DrainageArea2
The exponent typically varies between % and 1, depending upon river characteristics
and flow rates. Working downstream, the lower the exponent, the lower the predicted
flow will be. The exponent may vary with baseflow rates and increase when flows
increase. Additionally, during storm. events, the exponent usually increases because the
intensity and duration of the storm hydrograph increases as flow moves downstream and
drainage area increases. However, for consistency and convenience, a single method is
desired to project flows at Pine Hall based on real-time flows at Francisco. Therefore,
the projection was developed in a way that would yield conservative results. The
analysis focused on low flow periods, which have lower exponents. Because only one
exponent will be used, the projected storm hydrograph at Pine Hall will be the same
intensity and duration as Francisco, which will be less than the actual hydrograph
duration and intensity.
More than six months of recent data, in 15 -minute time increments, for the Francisco
and Wentworth gages was analyzed to determine the travel time between the two
gages. Analysis of peak flows indicates that the travel time between the Francisco gage
and the Wentworth gage is about 14 hours, and interpolation indicates the travel time
between Francisco and the pumping site is about 7 hours. The flows at Francisco were
lagged 14 hours for the exponent analysis so that the hydrograph peaks would match.
Duke Energy/TLB

The duration of the ascension limbs were generally similar for hydrographs at both
locations, but the recession limbs were longer downstream, as expected'.
The projection should be most accurate during periods when the available flow is less
than the level needed for maximum pumping. Therefore, the previously mentioned data
set was sorted to include only those flows. Including higher flows would raise the
median exponent for the data set because the exponent typically increases with flow.
The minimum flow rate required at Pine Hall for maximum pumping is 198 cfs. If the
projection exponent were 1, the corresponding flow at Wentworth would be 413 cfs.
Therefore, the maximum flow at Wentworth for the analysis was 413 cfs. The data set
included 13,301 flow ratios, and the corresponding median exponent was 0.51. This
exponent results in a flow multiplier of 2 between Francisco and Pine Hall and 0.69
between Pine Hall and Wentworth. Similar analysis was performed using the daily data
available for all three gages, and results were consistent. These multipliers will under
predict higher flows because the exponent typically increases with flow, and this
exponent was developed using only low flow data. However, this error will not affect
pumping rates because maximum pumping capacity will have already been realized.
Table 2 shows the protocol for pump operation when the available river flow is less than
the level required for maximum pumping. Pumping should not lower the downstream
river flow to less than 100 cfs, which is the sum of the 7Q10 level and the near -field
downstream municipal water withdrawals.
Duke Energy/TLB .
7,

Table 1
Francisco --Pine Hall*..
Wentworth
Gage # 02068500 02069000
02071000
Drainage. Area "(mi2) 129 501
1035
The Pine Hall gage is retired. It was located slightly upstream of the pumping site.
The pumping site is interchangeably referred to as Pine Hall.
Flow rates can be projected at Pine Hall
X
DrainageArea2
using this equation and flow rates at the Q __ Q
2 j
Francisco gage. X typically varies
DfainageAreal
between 0.5 and 1. X was determined
- -
using flow data from Francisco and Wentworth.
Minimum River Flow After Pumping (cfs)
100
Maximum Pumping Rate (cfs)
98
# Pumps
12
Pumping Rate Per Pump (cfs)
8
Minimum Flow at Pine Hall for Max Pumping (cfs)
198
Flow -Required at Francisco and Wentworthfor, Pumping_.
x-0.5 x=0.51
x=1.0
Francisco 101 99
51
Wentworth 283 287
413
X typically increases with flow; therefore, only flows less than 413 cfs at Wentworth
were used in determining the correct exponent for this portion of the Dan River.
The median value of X for those low flows was 0.51.
Multi tiers to Determine Flow at Pine Hall
x = 0.5 x = 0.51
x = 1.0
Francisco 1.97 2.00
3.88
Wentworth 0.70 0.69
0.48
Duke Energy/TLB -

Table 2
Putttp Operation Based on Flow at Francisco (x = 0.51)*
Flow at Francisco
Flow at Pine Hall
Pump Flow
Number
(cfs)
(cfs)
(cfs)
of Pumps
50.0
100.0
0.0
0
54.1
108.2
8.2
1
58.2
116.3
16.3
2
62.3
124.5
24.5
3
66.3
132.7
32.7
4
70.4
140.8
40.8
5
74.5
149.0
49.0
6
78.6
157.2
57.2
7
82.7
165.3
65.3
8
86.8
173.5
73.5
9
90.8
181.7
81.7
10
94.9
189.8
89.8
11
99.0
198.0
98.0
12
* Use of x=0.51 .at flows greater than those shown will produce conservative results
Also, this method assumes the shape of the storm hydrograph at Pine Hall is the
$'a,r ps at Francisco. This is conservative because the intensity and duration of
the hydrograph increases downstream.
Duke Energy/TLB
t
4,.