Cape_Fear_Model_of_Future_Use_Scenarios_Sept2008Revision

Cape Fear River Basin Surface Water
Assessment

Modeling of Future Water Use Scenarios

NC Division of Water Resources

October 2008

Drafted: September 29, 2008
2

Table of Contents
Page
Preface to the October 2008 Version 3
I. Introduction 5

Scope of the Model 7
Scenarios Modeled 8

II. Model Assumptions 10

Inputs 10
Outputs 11
Withdrawals and Discharges 12

III. Effects of Future Water Use on Jordan Lake 15
Jordan Lake Operation 15

Elevation Profile 16
Water Supply Pool Profile 17
Water Quality Pool Profile 18
Duration Curves 28
Jordan Lake Elevation Duration Curve 28
Water Supply Pool Duration Curve 28
Water Quality Pool Duration Curve 29
Impacts on Frequency of Jordan Lake Drought Stage Occurrence 33
Impacts on Boating at Jordan Lake 34

IV. Water Supply Demands vs. Delivery 35

V. Water Supply Intake Impacts 38

VI. In-stream Flow Evaluation 53

Analysis of In-stream Flows 54
Stream Flow Duration Curves 59
Flows at Lillington 59
Flows at Locks and Dams 59

VII. Other Model Results 66

VIII. Comments 67
Drafted: September 29, 2008
3
Preface to the October 2008 Version

Since the draft version of this document was released in March 2008, a number of changes
were made to the modeling scenarios presented in this analysis. The intention of these changes
is to improve the accuracy of how Jordan Lake operations are modeled and to better ensure that
water supply and irrigation demands are met by the model when adequate water is available at
the points of withdrawal. All changes have been incorporated into the modeling results
presented in this final version. A brief description of the changes will be provided here along
with an explanation of the impacts of the changes on modeling results.

Correction to Jordan Lake Water Quality Pool Accounting
For the model runs presented in the March 2008 draft, the model included inflows into Jordan
Lake estimations of the water quality releases from the water quality pool. Actually, the water
quality releases should only depend on the Lillington target flow and the flow from Deep River
at Moncure, and not on inflows to the reservoir. This revision has been made for all runs
presented in this final draft. Because the operation of the reservoir during drought depends on
how much storage remains in the water quality pool, this revision had some noticeable effects
on the storage remaining in the water quality pool. The corrected modeling results show in all
scenarios that Jordan Lake could go into drought operations, i.e. the water quality is drawn
down to drought levels, more often than under the previous scenarios. Detailed results are
presented in Section III: Effects of Future Water Use on Jordan Lake.

Changes to Jordan Lake Release Assumptions
Changes were made in the logic that guides how the model calculates the Jordan Lake releases
on days when the reservoir is not full. One of these changes relates to the model’s use of
perfect foresight. Perfect foresight means that the model knows precisely the inflows,
withdrawals, and discharges for every day. For example, with perfect foresight as the model
was previously set up, the model will exactly meet the Lillington target because it knows the
inflows to the Deep River and all local withdrawals and discharges between the reservoir and
Lillington. Reservoir operators of course do not have perfect foresight, and therefore the
Lillington target is not always exactly met. To better reflect how the reservoir is actually
operated, the model was revised to remove perfect foresight, and now releases the difference
between the Lillington flow target and the flow in the Deep River at Moncure on the previous
day (which is known to the reservoir operator). The model no longer exactly meets the
Lillington target as before, but it comes very close on most days. In order to even more closely
meet the Lillington target, the release was adjusted for withdrawals and discharges between the
dam and Lillington.

Another change made to the model’s logic relates to how the model handles demands
downstream of Lillington. Before, the model could have potentially released extra water from
Jordan Lake in order to meet demands downstream of Lillington. The logic was changed to
prevent this from happening.

Removing perfect foresight from the Jordan release logic caused a slight reduction in water
quality releases (See Section III). Regarding the change related to handling demands
downstream of Lillington, it was that found that even though the possibility existed, no
additional water was actually released to meet those demands under any of the previous
scenarios, because the river flows were already sufficient to meet the estimated demands
downstream. Therefore, the effects of this change had no impact on the results.
Drafted: September 29, 2008
4

Revision of the Model’s Weighting System

The OASIS model uses a weighting system to determine how to allocate water among
competing water uses. Simply stated, the model assigns points for allocating water to each of
the types of water use such as water supply demands, irrigation demands, minimum flow
requirements, and reservoir storage. To determine how to allocate the water, the model tries to
maximize the total number of weight points.

In the process of revising the analysis, it was discovered that at times, the model was not
meeting water supply and irrigation demands, even though there was adequate water in the
stream or reservoir at the point of demand. This was counter to the intent of the analysis. It
happened because the weight(s) for one or more competing water uses was set higher than the
demand weight.

To address this, all water supply demands and irrigation weight were examined and revised as
necessary to ensure that the demands are met whenever the stream flow or reservoir level
(whichever applies) is adequate to meet the demand.

This now means that if the model does not meet a water supply or irrigation demand, it is an
indication that the model has predicted a stream flow at or near zero at the point of the demand
or a lake level depleted to be at or very near its minimum elevation for that day.

Setting water supply and irrigation demand weights higher than weights for other water uses
does not mean that meeting these demands is more important than meeting demands for the
other uses. Rather, it was done to specifically evaluate how often these demands might not be
met under historical stream flow conditions and current policies. The model has the ability to
apply in-stream flow requirements if desired and as they become known.

Maximum Jordan Lake Water Supply Allocation
A scenario was added intended to evaluate Jordan Lake levels and downstream flow for the
situation in which the water supply pool is fully allocated. The safe yield of the water supply
pool is estimated to be 100 million gallons per day (mgd). Therefore, in order to develop this
scenario, starting with the 2050 Demands scenario, withdrawals from Jordan Lake were
increased so that the total water supply withdrawal from the reservoir is 100 mgd. This
scenario is included in output related to Jordan Lake levels and stream flows downstream of
the lake.

Long-Term Climate Change
A scenario has been added to this final draft which attempts to estimate what might be the
expected impacts of a long term increase in ambient temperatures in the Southeastern United
States. Accounting for the effects of long term global warming in hydrologic modeling is a
relatively new concept and it is not yet known how best to do so. However, one relatively easy
way to get an idea of the possible effects on the severity of droughts is to evaluate the impacts
of a scenario in which the natural inflows to the system have been reduced. A model scenario
was developed which depicts the system under projected water demands for the year 2050 in
which all of the natural inflows to the system have been reduced to 80% of their estimated
values based on historical flow measurements. Results from this scenario are compared to the
results for the natural flows scenarios.
Drafted: September 29, 2008
5
I. Introduction to the OASIS Model on the Surface Water Assessment

The Cape Fear River Basin Hydrologic Model (OASIS) is a computer based mathematical
model that simulates surface water flows in the Cape Fear River. It has the capability to take
into account a great deal of hydrologic information and water use data. It can be used to
evaluate the impacts of future projected future demands and operational scenarios. The version
of the model used for this analysis is based on the seventy-six year record of river flows from
1930 to 2005. The flows in this period of record include a wide range of flow conditions, like
several high-flow periods and several low-flow periods, including the exceptional drought
conditions of 2001-2002.

The 2003 demands scenario is used as the base case against which scenarios of projected future
demands and return flows are compared. Using the model to compare future demand
conditions with the base case conditions may help identify the possible impacts on reservoir
levels and stream flows at points of interest around the basin due to proposed increases in
water supply demands. This is the most comprehensive analysis that has been done so far
using the model.

The Division of Water Resources announced the beginning of this update to the Cape Fear
River Water Supply Plan in October 2007 and requested that water systems provide the
Division with any revised projections of future water supply demands. Except for the twenty
water users that submitted additional data, the modeled current and projected water supply
demands were derived from the 2002 local water supply plans submitted by the water systems.
Also at the October 2007 meeting, attendees heard presentations describing several new and
expanded withdrawals that are proposed or under development that will influence future
conditions of the Cape Fear River.

Recent Updates to Model Inputs

In the October 2007 meeting, Progress Energy representatives described the proposed addition
of more generation capacity at the site of the Harris Nuclear Plant in southwestern Wake
County. The analysis discussed in this report models the current demands and anticipated
demands for the existing facilities only. In future rounds of modeling we will include the
increased withdrawals needed to support the increased generation capacity.

The Lower Cape Fear Water and Sewer Authority (LCFWSA) presented information on two
projects. Currently, LCFWSA has a raw water intake just behind Lock and Dam #1 in Bladen
County on the main stem of the river. LCFWSA provides raw water to the City of Wilmington
and Brunswick County as well as some industrial facilities. In the future, LCFWSA also
anticipates supplying water to Pender County. LCFWSA expects to install an additional
pipeline and intake screens to carry water from the river to pumps located on shore. This will
increase capacity to withdraw water from 45 million gallons per day (mgd) to 96 mgd.

The City of Wilmington is beginning an expansion of the Sweeney Water Treatment Plant.
Raw water for this plant comes from Wilmington’s intake on the Cape Fear River behind Lock
and Dam #1 in the vicinity of the LCFWSA intake and the authority also supplies raw water to
the treatment plant. There is a node for each of these intakes in the model. In this modeling
exercise, all of Wilmington’s anticipated future demands that will be supplied by the Sweeney
Plant is assigned to Wilmington’s withdrawal node. The LCFWSA withdrawal node is
Drafted: September 29, 2008
6
assigned the anticipated future demands for their other customers. Since these two intakes
withdraw water from the same location in the river, the proportioning of water between the two
intakes has no effect on modeling results.

LCFWSA is also working with water users in the vicinity of Tar Heel in Bladen County on a
proposed new surface water intake on the Cape Fear River. This project is proposed to reduce
detrimental impacts to ground water resources in this area by shifting water users to surface
water. A node has been added to the model for this new withdrawal from the Cape Fear River.
As withdrawal needs get clarified, they can easily be added to the model in future model runs.

During 2004-2005, Harnett County Public Utilities expanded its raw water intake in the Cape
Fear River at Lillington. The new intake structure will improve reliability of the supply and
increase the capacity to withdraw water to 48 mgd.

Drafted: September 29, 2008
7
Scope of the Model

The geographic scope of the model includes the Deep River Basin, the Haw River Basin and
the Cape Fear River Basin above Lock and Dam #1 in Bladen County. The following
schematic map of the basin shows the geographic coverage of the model and the relative
location of the various model nodes.

Figure 1: Cape Fear Hydrologic Model Schematic

Figure 2 shows an example of the complexity of the model. Each of the polygons in the
schematic represents a node where the model performs a calculation to sum the effects of
inflows and outflows of water. The result of each calculation is used as an input for the next
downstream node. The section titled “Modeled Assumptions” describes the different types of
nodes.
Drafted: September 29, 2008
8
Figure 2: Cape Fear Hydrologic Model Detailed Schematic

Scenarios Modeled

For this round of modeling four different scenarios were modeled: a simulation of conditions
without any withdrawals, discharges or storage impoundments; a characterization of current
conditions and two scenarios of future withdrawals.

Scenario 1: Unregulated Flow Scenario

This scenario models stream flows which are the estimated natural flows in the basin,
unaffected by impoundments, water withdrawals, or wastewater discharges. To model this
scenario, all demands and discharges were set to zero. All reservoirs were assumed to have
zero usable storage, meaning they are modeled to remain full and release exactly the amount of
water that flows into them.

Drafted: September 29, 2008
9
Scenario 2: 2003 Demands Base Case

This scenario is intended to reflect the current water conditions. Modeled water demands were
estimated using local water supply plan data and additional information received from water
systems and other registered water users. The results of the other scenarios are compared to
this base case to identify possible changes in impacts due to projected changes in withdrawals
and return flows.

Scenario 3: 2030 Demands

Water demands are those projected for the year 2030 using local water supply plan data with
any updated projections received from water systems. Jordan Lake water supply withdrawals
may, in some cases, be greater than current water supply allocations. Withdrawals are assumed
to follow future water use projections provided by the allocation holders.

Scenario 4: 2050 Demands

The 2050 demand scenario is similar to Scenario 3 except that the water demands are those
needed to meet water demands projected for the year 2050 in the local water supply plans.

Scenario 5: 2050 Demands with Jordan Lake Water Supply Demands set to 100 MGD

This scenario is the same as the 2050 Demands scenario except that the water supply demand
from Jordan Lake are set to 100 MGD which is the estimated safe yield of the water supply
pool. Under the 2050 Demands scenario, a total of 73.5 mgd is withdrawn from Jordan Lake
for water supply. This scenario was developed by adding a water supply demand node to
Jordan Lake and setting the annual withdrawal from the node to 26.5 mgd, bringing the total
water supply withdrawal to 100 mgd. Note that the additional 26.5 mgd of water withdrawn is
assumed to be a 100% consumptive use, none of the additional withdrawal being returned to
the basin. This is a conservative assumption chosen to assess the maximum impacts to the
Jordan Lake level of the additional withdrawal.

Scenario 6: 2050 Demands with 80% of Historic Natural Inflows

This scenario is the same as the 2050 Demands scenario except that the natural inflows to the
system have all been multiplied by 0.8. The purpose of this scenario is to make an attempt to
assess the potential impacts of a long period of increased ambient temperatures. The idea is
that if ambient temperatures are consistently higher, this will cause an increase in evaporation
and possibly cause lower net inflows to the system.

Drafted: September 29, 2008
10
II. Model Assumptions

The Cape Fear Hydrologic Model uses a software program called OASIS (Operational
Analysis and Simulation of Integrated Systems) developed by Hydrologics Inc. OASIS is a
mass balance model that uses sequential calculations to simulate the routing of water through
the watershed. OASIS balances water coming in with water going out at all nodes, subject to
goals and constraints established for each node. The model also assigns weights to each type
of water use which allows the model to make allocation decisions between competing uses. At
the reservoir nodes, water is stored and released subject to user-defined operating rules. The
model operates on a daily time step making one set of calculations for each day and uses daily
average values for each calculation.

Inputs

Inputs to the model calculations include the following:

1. Estimated Daily Natural Inflows: The model uses a set of daily natural inflows which
estimate the water entering the system due to runoff. These inflow data were developed using
seventy-six years of flow records and are adjusted for upstream withdrawals, discharges, and
reservoir operations. These inflows are modeled as entering the systems at discrete points
scattered throughout the watershed. In the schematic, they are shown as purple arrows.

2. Daily Withdrawals: Water is removed from the system at discrete
points, represented in the model as withdrawal nodes. These nodes show up
as blue boxes on the schematic.
These withdrawals can be for water supply systems, industrial water users, or agricultural
water users. Public water supply withdrawals are based on local water supply plan data which
in some cases were updated to reflect improvements in projections of future demands. Self-
supplied industrial water withdrawals were derived from data submitted under the Division’s
water withdrawal registration program. It is assumed to remain the same in 2030 and 2050 as it
is in the base case unless additional information was available to justify changes in projections.
Agricultural demands are the same as those used in previous versions of the model.
Agricultural uses for livestock and irrigation were estimated with the help of county
agricultural extension agents and an agricultural extension irrigation specialist. Water use
estimates were developed for crops, taking into consideration variations in planting times in the
upper, middle and lower regions of the basin. Livestock water needs are based on animal head
counts in each county and the water use factors used by the USGS in the 1995 Estimated Water
Use in North Carolina. Percentages of irrigated crops and livestock in the basin were developed
for each county in consultation with county agricultural extension agents. There are individual
nodes for agricultural water use in each county in the basin.

3. Daily Wastewater Discharges: Return flows from wastewater discharges are modeled
similar to natural inflows, as water inputs at discrete nodes. They are represented in the
schematic as brown arrows.
Drafted: September 29, 2008
11
Inflows from wastewater discharges come from industrial and municipal wastewater treatment
plants and water reclamation facilities. The inputs used in the base case were used to calculate
the percentage of a facility’s water withdrawal that is directly returned to the surface waters of
the basin as wastewater return flow. . This percentage was then applied to estimated future
withdrawals to estimate future wastewater return flows. For example, if a town withdraws 10
mgd on average and returns 6 mgd of treated wastewater, then 60% of the withdrawal is
returned directly to the surface waters of the basin. In the 2030 and 2050 scenarios, the
assumed wastewater discharge is again 60% of the withdrawal.

4. Reservoir Operating Guidelines and Data: The model balances inflows and
outflows at each node. Inflows equal outflows on all days for all nodes except
reservoir nodes, represented by red triangles in the schematic. In the case of a
reservoir, the change in daily storage is considered in the balance equation. Each
reservoir in the model has a set of operating guidelines. Only two reservoirs in the system
have minimum release requirements, Jordan Lake and Randleman Lake. Jordan Lake has a
fairly complex set of operating rules, which are explained in Section III. Randleman Lake is
operated to maintain a minimum release that varies according to reservoir level. The minimum
release is assumed in the model as follows:

Outputs

The OASIS model can provide a variety of model run outputs in a variety of configurations.
The primary outputs used for this analysis include the following:

1. Daily Flows: The model outputs daily flows into a node, out of a node, or
through an arc. An arc connects two nodes, and is represented in the schematic as a
black arrow between two nodes.
2. Daily Reservoir Levels
3. Daily Reservoir Releases
4. Daily Accounting of Jordan Lake Conservation Storage: The model keeps track of how
much water is remaining in the water supply storage pool and the water quality storage pool.
This information is used to determine the release from the reservoir during droughts.
5. Drought Stage at Jordan Lake: According to the percentage of storage remaining in
water quality storage, the model outputs the daily drought stage.

Drafted: September 29, 2008
12
Withdrawals and Discharges

Table 1 summarizes the estimated withdrawals and return flows for the base case and the 2030
and 2050 demand scenarios for the major water users modeled for this analysis. All volumes
are shown in million gallons per day (MGD).

Table 1: Demands and Discharges Assumed in the Modeling (All units are in MGD)
System Node Description
Node
# Node Type 2003 2030 2050
Angier NC0082597 (AngiersWW) 654
WWTP
Discharge 0.43 0.81 0.90
Archdale Randleman Lake 904 Withdrawal 1.20 1.20
Asheboro NC0026123 (AsheboroWW) 282
WWTP
Discharge 5.62 7.57 9.94
Broadway NC0059242 (BroadwayWW) 940
WWTP
Discharge 0.11 0.13
Burlington Lake Mackintosh 341 Withdrawal 10.86 11.73 15.51
NC0083828 (BurlingtonMackintoshWW) 352 WTP Discharge 0.39 0.41 0.54
Stoney Creek Reservoir 71 Withdrawal 7.30 5.97 7.91
NC0023868 (BurlingtonEastWW) 106
WWTP
Discharge 0.07 9.87 12.93
NC0023876 (BurlingtonSouthWW) 362
WWTP
Discharge 6.40 9.48 12.48
Carthage Nicks Creek 701 Withdrawal 0.26 0.59 0.70
Carolina Trace WS NC0038831 (CarolinaTraceUtilWW) 674
WWTP
Discharge 0.25 0.27 0.27
Cary Jordan Lake 471 Withdrawal 14.02 32.09 34.88
Apex NC0081591 (CaryApxWW) 472 WTP Discharge 0.69 0.00 0.00

Western Wake Regional WRF
(CaryRegWW) 930
WWTP
Discharge 18.40 20.60
Chatham Co North Jordan Lake 473 Withdrawal 1.03 9.63 15.88
NC0035866 (NorthChathamWW) 452
WWTP
Discharge 0.01 0.05 0.08
Dunn Cape Fear River 663 Withdrawal 3.49 11.77 17.59
NC0078955 (DunnWW) 682 WTP Discharge 0.51 0.76
NC0043176 (DunnWWTP) 692
WWTP
Discharge 3.04 9.99 15.35
Durham Jordan Lake 476 Withdrawal 10.00 10.00
NC0047957 (DurhamReclamationWW) 462
WWTP
Discharge 10.73 11.29 12.90
NC0026051 (DurhamCtyTriangleWW) 454
WWTP
Discharge 4.49 4.02 4.59
Elizabethtown NC006671 (ElizabethtownWW) 960 Discharge 1.04 1.25
Erwin Swift Textiles Reservoir 661 Withdrawal 0.65 0.89 1.06
NC0064521 (ErwinSouthWW) 686
WWTP
Discharge 0.95 0.98 1.17
NC0001406 (BurlingtonIndustriesWW) 684
WWTP
Discharge 8.74 0.00 0.00
Fayetteville Cape Fear River 733 Withdrawal 20.00 69.18 83.11
NC0076783 (FayettevillePOHofferWW) 744 WTP Discharge 1.23 4.71 5.73
Little Cross Creek 761 Withdrawal 0.00 0.00 0.00

NC0023957
(FayettevilleCrossCreekWW) 742
WWTP
Discharge 12.39 31.87 43.24

NC0050105
(FayettevilleRockfishCreekWW) 774
WWTP
Discharge 13.04 24.00 24.00
Fort Bragg Little Upper River Dam 721 Withdrawal 6.27 0.00 0.00
Drafted: September 29, 2008
13
System Node Description
Node
# Node Type 2003 2030 2050
Franklinville NC0007820 (FranklinvilleWW) 910
WWTP
Discharge 0.04 0.05
Fuquay-Varina NC0028118 (FuquayVarinayWW) 552
WWTP
Discharge 1.01 0.39 1.18
Goldston Gulf SD Deep River 605 Withdrawal 0.13 1.94
Graham Graham-Mebane Lake 321 Withdrawal 3.23 6.05 8.11
Mebane NC0045292 (GrahamMebaneWW) 102 WTP Discharge 0.323 0.60 0.81
NC0021211 (GrahamCtyWW) 108
WWTP
Discharge 1.85 2.46 3.14
NC0021474 (MebaneWW) 104
WWTP
Discharge 1.84 2.63
Greensboro Lake Townsend 141 Withdrawal 19.65 17.53 23.19

NC 0081671
(GreensboroLakeTownsendWW) 142 WTP Discharge 12.76 1.56 2.07
Lake Brandt 121 Withdrawal 11.44 8.77 11.59
NC 0081426 (GreensboroMitchellWW) 174 WTP Discharge 0.26 0.15 0.20
Randleman Lake 901 Withdrawal 20.83 27.54

NC0047384
(GreensboroTZOsborneWW) 182
WWTP
Discharge 23.08 26.60 40.34

NC0024325
(GreensboroNBuffaloCrkWW) 176
WWTP
Discharge 1.97 16.00 16.00

UNC Greensboro (formerly NC0082082)
(UNCGreensboroWW) 172 Discharge 0.03 0.00 0.00
Harnett Co Cape Fear River 551 Withdrawal 7.04 27.47 40.03
NC0021636 (LillingtonWW) 664
WWTP
Discharge 0.43 0.95
NC0030091 (BuiesCreekWW) 656
WWTP
Discharge 0.50 0.50
NC0031470 (HarnettCoWW) 950
WWTP
Discharge 0.40 0.40
High Point City and Oak Hollow Lakes 221 Withdrawal 13.12 10.58 12.30
NC0081256 (HighPointWW) 236 WTP Discharge 0.86 0.66 0.77
Randleman Lake 902 Withdrawal 4.80 5.44
NC0024210 (HighPointEastWW) 232
WWTP
Discharge 15.08 19.06 22.82
Holly Springs Jordan Lake Release 924 Withdrawal 0.00 0.00
Cape Fear River 923 Withdrawal 0.00 0.00
NC0063096 (HollySpringsWW) 522
WWTP
Discharge 0.92 4.01 4.83
Jamestown Randleman Lake 903 Withdrawal 0.67 0.71
Lee County Cumnock
Golden Poultry Deep River 601 Withdrawal 0.65 2.50 2.50
NC0083852 (LeeCtyWW) 616 WTP Discharge 0.16 0.40 0.40
Lower Cape Fear WSA Cape Fear River 825 Withdrawal 17.58 21.31 25.70
Moore Co (Vass) Thagards Lake 711 Withdrawal 0.00 0.00
Morrisville Jordan Lake 477 Withdrawal 1.5 3.96 3.96
Orange-Alamance Jordan Lake 921 Withdrawal 0.00 0.00
Orange Co
Orange WASA Cane Creek Reservoir 391 Withdrawal 5.43 4.68 6.90
University Lake 431 Withdrawal 2.84 3.19 4.70
NC0082210 (OWASA_WTP_WW) 442 WTP Discharge 0.36 0.40 0.59
Jordan Lake 922 Withdrawal 5.00 5.00
NC0025241 (OWASA_MasonFarmWW) 444
WWTP
Discharge 8.30 9.94 12.82
Drafted: September 29, 2008
14
System Node Description
Node
# Node Type 2003 2030 2050
Pittsboro Haw River 401 Withdrawal 0.65 8.14 8.14
NC0020354 (PittsboroWW) 412
WWTP
Discharge 0.45 4.23 4.23
Progress Energy Shearon Harris 521 Withdrawal 31.41 31.41 31.41
524 Discharge 19.5 19.5 19.5
Progress Energy Cape Fear Plant 487 Withdrawal 194.15 194.15 194.15
512 Discharge 194 194 194
Raeford NC0026514 (RaefordWW) 772
WWTP
Discharge 1.88 3.61 4.10
Ramseur Sandy Creek 301 Withdrawal 0.58 0.92 1.09
NC0026565 (RamseurWW) 572
WWTP
Discharge 0.27 0.34 0.40
Randleman Randleman City Reservoir 261 Withdrawal 0.90 0.16 0.55
Randleman Lake 905 Withdrawal 1.01 1.01
NC0025445 (RandlemanWW) 252
WWTP
Discharge 1.09 1.08 1.45
Randolph Co Randleman Lake 906 Withdrawal 6.00 6.00
Reidsville Lake Reidsville 31 Withdrawal 5.75 5.41 5.76
NC0046345 (Reidsville_WTP_WW) 24 WTP Discharge 0.48 0.46 0.49
NC0024881 (ReidsvilleWW) 42
WWTP
Discharge 3.05 3.46 3.63
Robbins Brooks 591 Withdrawal 0.26 0.24 0.27
NC0062855 (RobbinsWW) 582
WWTP
Discharge 0.25 0.16 0.19
Sanford Cape Fear River 491 Withdrawal 6.53 21.06 39.73
NC0002861 (SanfordWW) 502 WTP Discharge 0.65 2.09 3.93
NC0024147 (Sanford_WWTP) 612
WWTP
Discharge 4.38 12.36 23.67
Siler City Rocky River 631 Withdrawal 2.97 5.83 6.00
NC0058548 (SilerCityWW) 632
WWTP
Discharge 2.96 4.56 5.40
Spring Lake NC0030970 (SpringLakeWW) 722
WWTP
Discharge 0.90 1.42 1.85
Star NC0058548 (StarWW) 592
WWTP
Discharge 0.16 0.11 0.12
Wake Co - RTP South Jordan Lake 474 Withdrawal 0.39 2.65 3.82
Wilmington Cape Fear River 823 Withdrawal 14.80 30.70 39.80

Drafted: September 29, 2008
15
III. Effects of Future Water Use on Jordan Lake
Jordan Lake Operation
Jordan Lake is operated by the US Army Corps of Engineers. It was designed to provide for
water supply, recreation, flood control, fish and wildlife management, and low-flow
augmentation. As is typical for multi-purpose reservoirs, the Lake's storage volume of the
impoundment is divided vertically into pools that are delineated by elevation, above sea level.
These levels are shown in Figure 3. The normally empty flood control storage provides about
538,000 acre feet of controlled flood storage above the conservation pool. The conservation
pool provides storage for water supply and low flow augmentation. Below the conservation
pool the sediment pool, provides space for the accumulation of sediment.

The top of the conservation pool corresponds with the normal lake level of 216 feet above
mean sea level (MSL). At this elevation, Jordan Lake covers 13,900 acres. As the following
Figure shows, usable water in the lake at its normal elevation amounts to a total volume of
approximately 140,400 acre-feet and is referred to as the conservation storage. Approximately
45,800 acre-feet in conservation storage, or about 15 billion gallons, is designated to provide
water supply, and is called the water supply pool. This amount of storage is estimated to be
able to furnish approximately 100 million gallons per day (MGD) during the severest drought.

Figure 3: Jordan Lake Storage Volume

In addition to water supply, the Lake's conservation storage provides 94,600 acre-feet of water
for downstream flow augmentation to benefit water quality and economic development. This
Drafted: September 29, 2008
16
storage is generally referred to as the flow augmentation or water quality pool. The water
quality pool is used to maintain a target minimum flow of about 388 MGD (600 cfs) at
Lillington during non-drought periods, and less during droughts. Inflows to and withdrawals
from each of these storage pools are accounted-for independently. Therefore, withdrawals
from the water supply storage pool do not reduce the amount of water remaining in the flow
augmentation pool.

Jordan Lake has more complex operating rules than the other reservoirs in the basin. During
droughts, the Army Corps of Engineers operates the lake according to a schedule that indicates
how releases should be varied as the water quality pool draws down. The Corps is in the
process of recommending that these operating rules be approved for permanent operations.

In general, releases from the lake depend on the amount of water remaining in the water quality
pool. There is a downstream target flow at Lillington that affects how much water must be
released from the dam. The operating schedule shown in Table 2 summarizes the proposed
drought management protocol and is included in the model used for this analysis.

Table 2: Jordan Lake Operating Rules During Drought

Minimum Lillington
Drought % Remaining in Release Target
Stage WQ Pool (cfs) (cfs)
0 80-100 40 600
1 60-80 40 450-600
2 40-60 40 300-450
3 20-40 200 None
4 0-20 100 None

Presentation of Modeling Results

The following sections present the results of modeling the various scenarios used in this
analysis. The results are shown in several different presentation formats to aid understanding.

Elevation Profile

Elevation profiles show how reservoir levels vary over a specified period of time. They are
useful for examining the shorter term fluctuations in reservoir elevation. The Jordan Lake
elevation profile in Figure 4 shows the expected daily lake elevation for each of the three
demand scenarios over the entire 76-year period of record.

It is of particular interest to notice the behavior of the lake elevation during drought periods,
when it is drawn down to the lowest levels. As expected, the elevation profile for the three
demand scenario in Figure 4 shows that during the major droughts on record, Jordan Lake is
expected to be drawn down increasingly further as water supply demands increase from the
base case to the 2050 demand levels. The profile shows a minimum lake elevation of about
207.5 feet being reached under the estimated 2050 demands scenario during the 2002 drought.
Another deep drawdown occurs in the 1952-53 drought, drawing the water level down to just
below 208 feet.

Drafted: September 29, 2008
17
Figure 5 shows a comparison between the elevation profiles for Jordan Lake for the 2050
demands scenarios assuming the actual projected Jordan Lake water supply demands in one
scenario and water supply demands of 100 mgd, the full estimated safe yield of the water
supply pool, in the other scenario. The added water supply demands tend to further draw down
the lake from about 0.5 to 2 feet for most droughts. The lowest expected elevation is just
above 206 feet for the droughts of the 1950s and 2002.

Figure 6 shows a comparison between the 2050 demands scenario with natural inflows and the
2050 scenario with 80% of natural inflows. Reducing the natural inflows tends to further draw
down the elevation by about 0.5 to 1 foot for most droughts. However, for a 20% reduction of
the 2002 drought conditions, the water level would be expected to decline an additional 2.5
feet to an elevation of about 205 feet.

Water Supply Pool Profile

Table 3 shows that Jordan Lake currently has combined estimated water supply withdrawals of
16.94 mgd. This increases to 63.33 mgd in 2030 and 73.54 mgd in 2050.

Table 3

The water supply pool profiles for the three demands scenarios in Figure 7 shows, as expected,
that the pool is drawn down increasingly as water supply demand increases from the base case
up to 2050 demand levels. It shows that the minimum predicted storage remaining in the water
supply pool is about 50% reached during the 1952-53 drought under the 2050 demands
scenario. It also shows that the water supply pool would again drop below 55% remaining
during a repeat of the 2002 drought conditions with the projected 2050 demands.

Figure 8 shows that the water supply pool profile is lowered significantly under the 100 mgd
water supply scenario, lowering the pool as low as 20% full in three different droughts over the
period of record. Because the pool was not fully depleted under this scenario would indicate
that the safe yield of the water supply pool may be slightly higher than 100 mgd.

Drafted: September 29, 2008
18
Figure 9 shows that reducing system inflows to 80% of the historical natural inflows under
2050 demands is expected to impact the water supply pool by drawing it down an additional 2
to 10 percent for most droughts, down to a low of about 45 % full.

Water Quality Pool Profile

The water quality (WQ) or flow augmentation pool profile in Figure 10 shows some interesting
and apparently non-intuitive results. However, on closer look, the results are easily explained.
An interesting observation is that the periods when the water quality pool is drawn down the
most do not always occur under the 2050 demands scenario, but rather sometimes under the
base case demand scenarios. The lowest expected level predicted by the model is just above
20% remaining under the base case demands and the inflows during the 1952-53 drought
conditions. There were no expected occurrences of Stage 4 drought (WQ pool below 20%)
under any of the demands scenarios. This is an indication that the Corps of Engineers’ drought
response measures are effective at maintaining the water quality pool storage even during the
most severe drought conditions.

It is important to understand why the model sometimes predicts the water quality pool to drawn
down further under current base case demands than under the higher 2050 demands. The
Town of Cary is currently in the process of developing the Western Wake wastewater
treatment plant that is expected to discharge treated wastewater to the Cape Fear River below
the Jordan Lake and upstream of the Lillington gage. Discharges from this plant will flow by
the Lillington gage and therefore reduce the amount of water that must be released from the
Jordan dam to meet the target flow at Lillington specified under the drought operating rules.
Since water released from the dam to meet the Lillington target comes out of the water quality
pool, the increased future discharges from the Western Wake plant in effect relieve stress on
the water quality pool to meet the Lillington in-stream flow target. Therefore, as withdrawals
from the water supply pool increase with increasing future water supply demands, releases
from the water quality pool that are required to meet the in-stream flow target at Lillington
during drought tend to decrease.

Figure 11 is a comparison of the water quality pool profiles for the normal 2050 demands and
scenario and the scenario in which the water supply demands from Jordan Lake are increased
to 100 mgd. It shows that as expected, increasing water supply demands to 100 mgd have little
or no noticeable impact on the water quality pool.

Figure 12 is a comparison of the water quality pool profiles for the normal projected 2050
demands and scenario and the scenario in which the inflow to the system are reduced to 80% of
historical inflows. It shows an impact of an additional 3 to 10 percent of draw down on the
water quality pool for most droughts. This would cause an increase in occurrences of drought
response measures which will be described later in this document.
Drafted: September 29, 2008
19
Figure 4: Jordan Lake Elevation Profile

Drafted: September 29, 2008
20
Figure 5: Jordan Lake Elevation Profile – 100 MGD Water Supply Demand Scenario

Drafted: September 29, 2008
21
Figure 6: Jordan Lake Elevation Profile – 80 % Inflows Scenario

Drafted: September 29, 2008
22
Figure 7: Jordan Lake Water Supply Pool Profile – Demands Scenarios

Drafted: September 29, 2008
23

Figure 8: Jordan Lake Water Supply Pool Profile – 100 MGD Scenario

Drafted: September 29, 2008
24

Figure 9: Jordan Lake Water Supply Pool Profile – 80%
Inflows

Drafted: September 29, 2008
25

Figure 10: Jordan Lake Water Quality Pool Profile – Demands Scenarios

Drafted: September 29, 2008
26

Figure 11: Jordan Lake Water Quality Pool Profile – 100 MGD Water Supply Demand

Drafted: September 29, 2008
27

Figure 12: Jordan Lake Water Quality Pool Profile – 80% Inflow

F
Drafted: September 29, 2008
28

Duration Curves

Another way to present predicted reservoir storage impacts shown by the model is by the use
of duration curves. Duration curves are cumulative frequency curves showing the percentage
of time over the period of record that a particular reservoir level is less than or equal to a
specified amount. The curves show in a probabilistic way how reservoir levels vary over the
entire period of record based on the conditions specified in the model. They are useful for
comparing scenarios over a long time scale. The area under the curve is a long term
representation of the amount of water in storage. Therefore, the lower the curve, the less water
being stored over the entire 76 year record.

Jordan Lake Elevation Duration Curve

Figure 13 shows elevation duration curves for all five scenarios. The curves show that, as
expected, less water is stored in the lake as water supply demands increase. The differences
between the scenarios are only seen to the left of 50% on the x-axis. Therefore, the duration
curve only shows the driest 50% of the days on record. The duration curves again show the
minimum elevation reached as about 207.5 feet for the 2050 demands scenario. The vertical
difference between the curves for the 2030 and 2050 scenarios indicate the predicted impact to
reservoir water levels due to the conditions modeled in these scenarios.

The incremental impact of increasing water supply demands to 100 mgd under 2050 conditions
is shown to pull the 2050 demands curve down from 0.5 to 1 foot. The impacts of reducing
inflows to 80% of historical inflows even further lowers the 2050 demands curve another 0.25
to 0.5 foot.

Water Supply Pool Duration Curve

Since the conditions of the water supply pool are not directly related to the level of water in the
reservoir, water supply pool conditions are shown as the percent of storage remaining.

The water supply duration curves in Figure 14 show a clear impact of the increased water
supply demands predicted for Jordan Lake in the 2030 and 2050 demands scenarios. The 2030
and 2050 cases are both significantly below the base case 2003 demands curve. The difference
between the 2030 and 2050 curves is not as drastic. The curves show that the lowest predicted
draw down of the water supply pool for the three primary scenarios modeled is to about 50% of
storage remaining under the projected 2050 demands scenario.

Reducing inflows to 80% of historical has a noticeable impact on the 2050 curve, but an even
greater impact is that of increasing water supply demands to 100 mgd.

Drafted: September 29, 2008
29
Water Quality Pool Duration Curve

The water quality pool duration curves in Figure 15 show how future water supply withdrawals
impact the water quality pool. The 2003 base case curve is the highest of the curves, meaning
that over the entire period of record, more water is held in water quality storage for this
scenario than the other two. During wetter periods, between 5% and 35% on the x-axis, the
base case scenario is noticeably higher than the 2030 and 2050 scenarios. However, all
scenarios are very similar during the driest 2-3% of the time. This is another indication that the
Jordan Lake drought response measures for the deeper stages of drought, Stage 3 and 4, tend to
prevent further lowering of the water quality pool once the pool reaches 40% full.

As expected, increasing water supply demands from Jordan Lake to 100 mgd shows little
noticeable impact on the duration curve. However, reducing inflows to 80% of historical
shows a very noticeable lowering of the 2050 curve.
Drafted: September 29, 2008
30
Figure 13: Jordan Lake Elevation Duration Curve

Drafted: September 29, 2008
31

Figure 14: Jordan Lake Water Supply Pool Duration Curve

Drafted: September 29, 2008
32
Figure 15: Jordan Lake Water Quality Pool Duration Curve

Drafted: September 29, 2008
33
Impacts on Frequency of Jordan Lake Drought Stage Occurrence

Figure 16 depicts how frequently each stage of drought is expected to be reached for the five
scenarios. The model counts the number of days during the 76-year period that the reservoir
operates at each of the drought stages and this chart summarizes the totals.

Because the drought stage designation is related to the percent of storage remaining in the
water quality pool, these results are an alternative way to depict the previously presented
duration curves and profiles of the water quality pool storage. For all scenarios, the reservoir
operates at stage 0 more than 70% of the time. The 2030 and 2050 demands scenarios show an
increase in the number of days at drought stage 1 from 8.2% for the base case to 9.8%. There
is also an increase in the percent of days at Stage 2 from 7.9% for the base to 9.2% and 8.5%
respectively for the 2030 and 2050 scenarios. The percent of days at Stage 3 is 2.1% and 2.4%
for the 2003 and 2030 scenarios, but increases to 4.4% for the 2050 scenario. Finally, none of
the five scenarios have Jordan Lake reaching Stage 4 drought level on any days.

Increasing water supply demands to 100 mgd has a small impact on drought stage occurrences
as compared to the 2050 scenario. However, reducing inflows to 80% of historical has a great
impact on the occurrence of Stage 2 and 3 drought occurrence. Stage 2 occurrences are
increased from 8.5% to 12.0% of the days, and Stage 3 occurrences are increased from 4.4% to
7.1% of the days.

Figure 16: Jordan Lake Drought Stage Occurrence

Jordan Lake Drought Stage
81.8%78.5%77.2%76.1%70.9%
8.2%9.8%9.8%10.3%10.0%
7.9%9.2%8.5%9.1%12.0%
7.1%4.5%2.1% 4.4%2.4%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2003 Demands 2030 Demands 2050 Demands 2050 100 MGD
WS Demand
2050 80%
Inflows
Scenario% of Days at Drought StageStage 4
Stage 3
Stage 2
Stage 1
Stage 0

Drafted: September 29, 2008
34
Impacts on Boating at Jordan Lake

Figure 17 shows the duration curves for Jordan Lake for the five scenarios and the elevation
levels of each of the boat ramps. From this curve, it can be estimated how often to expect that
the various boat ramps might be impacted.

Figure 17: Jordan Lake Boating Impacts

Drafted: September 29, 2008
35
IV. Water Supply Demands vs. Delivery

Note: The results presented in this section are only for the three water demands scenarios; the
base case, 2030, and 2050 demands scenarios.

There are 42 modeled water supply demand nodes. In the scenarios, the nodes were individually
examined to determine if the projected available quantity for surface water would be sufficient to
meet projected demand at each of those nodes. The withdrawal amounts assumed by the model at
each water supply node are summarized in the previous section titled “Withdrawals and
Discharges”.

For demand nodes on run-of-river sections of streams, the model has a set of weights and goals
that determine whether water supply demands can be met. The model uses these weights to
prioritize water uses. Simply stated, the weights assign points to each type of use such as a water
supply demand, irrigation demand, minimum release from a reservoir, or reservoir storage. The
model allocates water by choosing the allocation which maximizes the total weight points. At
this time, minimum in-stream flow needs have not been identified and therefore have not been
assigned a weight. Using the model to analyze in-stream flows at additional nodes may help
identify in-stream flow concerns downstream of water supply withdrawals. In future model runs,
in-stream flow targets may be set as needed which may further constrain water supply
withdrawals.

For demand nodes from reservoirs, the water supply demand is met if the reservoir has sufficient
water remaining in storage. If the model predicts that a demand from a reservoir is not met, this
is an indication that the reservoir has been depleted.

Water demand deliveries were compared to water supply demand for each of the three demands
scenarios. The model predicts that for 31 of the 42 water supply nodes, the full demand is
met for all days for all three water demand scenarios. However, there are 10 nodes for which
the full demand is not met under all three water demand scenarios.

Table 4 summarizes the instances in which the model predicts that the full water supply demand
would not be met in one or more scenarios. They are listed in alphabetical order and divided into
systems which have only a single water supply and systems which have multiple water supplies.

As explained in the preface, the water supply demand weights have been closely examined since
the draft version of this report was released in March 2008. Many water supply demands
weights were adjusted to ensure that the demands would be met if adequate water is available at
the withdrawal point. For this reason, a number of systems which previously had a small
predicted water supply deficit, no longer have one. Also, in the draft version, there was a
significant predicted deficit for the Fayetteville Glenville Lake withdrawal. For modeling
purposes, this demand was moved to Fayetteville’s run-of-river intake on the Cape Fear River.
Subsequently, no deficit is now predicted for Fayetteville’s water supply, as the flow at the run-
of-river intake is expected to be adequate to meet its demands.
The largest water supply deficit predicted by the model is for Orange Water and Sewer Authority
(OWASA) University Lake, which shows a deficit in 24 of the 76 years in the 2050 scenario.
However, this deficit is likely due not to an actual water supply shortage, but rather how the
Drafted: September 29, 2008
36
model deals with the OWASA two-reservoir water system. The model assigns separate water
demands to each water supply. In actuality, if University Lake is depleted, the Cane Creek
Reservoir is used to meet the demands. However, the model is currently not set up to take this
into account. This is also the case with Greensboro, High Point, and Fort Bragg, which have
smaller deficits at one of their sources. It should be noted that the Cane Creek Reservoir water
supply withdrawals showed no deficits in the 2003 and 2030 scenarios, but did have one 2-day
deficit under the 2050 scenario.
The model predicts a deficit for Fort Bragg under the base case scenario in 7 of the 76 years.
However, in future scenarios, the Fort Bragg demand is to be met mostly by Fayetteville, so no
deficit was predicted in either 2030 or 2050. Fort Bragg’s future demands were considered by
the model in Fayetteville’s demands for future scenarios.

Among the water systems with only one water supply, deficits were predicted for Cone Mills
Richland, Dupont, Ramseur, Randleman, and Robbins. However, only for Ramseur was a deficit
predicted in more than two of the 76 years.

The model predicts significant water supply deficits for Ramseur both under the base case
scenario and future scenarios. The Ramseur reservoir is small as is Ramseur’s projected water
supply demand, not expected to exceed 1.1 mgd before the year 2050. However, further
attention may be necessary as the model shows a clear potential water supply shortage.
Drafted: September 29, 2008 37 Table 4: Water Supply Demand Deficits Predicted by the OASIS Model Model Scenario 2003 # of Days Longest Years 2030 # of Days Longest Years 2050 # of Days Longest Years Demand Per Year Deficit Demand Demand Per Year Deficit Demand Demand Per Year Deficit Demand Water Systems (mgd) Demand Not Met (Days) Not Met (mgd) Demand Not Met (Days) Not Met (mgd) Demand Not Met (Days) Not Met Out of 76 Out of 76 Out of 76 Systems With a Single Water Sources Cone Mills Richland Lake 0.7 0.1 4 2 Dupont 12.3 0.0 1 1 12.3 0.0 1 1 Ramseur 0.58 1.8 34 12 0.9 2.8 35 16 1.1 3.8 43 20 Randleman 0.2 0.8 16 2 0.6 0.3 15 1 Robbins CB Brooks 0.26 0.8 33 2 0.24 0.8 33 2 0.27 0.8 33 2 Systems With Multiple Water Sources Ft. Bragg 6.3 0.6 12 7 Greensboro Townsend Lake 23.2 5.5 36 3 High Point - F Ward 13.12 2.1 34 5 12.3 1.2 16 4 OWASA Cane Creek 7.0 0.0 2 1 OWASA University Lake 2.8 1.9 48 7 3.2 1.9 7 7 4.7 11.5 92 24 # of Days Per Year Demand Not Met = Number of days out of the full 27,394 days of record that the model shows the full demand maybe is not met, divided by 76 (years of record). Longest Deficit (Days) = The greatest number of consecutive days over the entire 76 year record that the full water supply demand maybe is not met. Years Demand Not Met = The number of years out of a total of 76 that the full water supply demand maybe is not met. Systems in Red are those for which a deficit is predicted in any scenario seven or more years out of the 76 year record.
Drafted: September 29, 2008
38
V. Water Supply Intake Impacts

There are fourteen reservoirs that are considered in the model. Following are duration curves for
the water surface elevation for each of the reservoirs. Each plot includes a curve for all five
scenarios. If the information is available, the levels of the water withdrawal intakes are indicated
on the plots. These plots are useful to estimate the percentage of time that reservoir levels are
predicted to drop low enough to impact the water supply intakes. For most reservoirs, very few
impacts to water supply intakes are predicted. However, the Ramsuer Reservoir is expected to
fall below the level of the intake more often than for any of the other reservoirs. Also, in actual
practice, water system operators would take actions by imposing water conservation measures to
reduce the likelihood of the reservoir levels dropping below the level of the intake. In cases
where a water system has more than one water source, one reservoir could be allowed to be
exhausted while holding the other sources in reserve.
Drafted: September 29, 2008
39
Figure 18: Jordan Lake Elevation Duration Curve

Drafted: September 29, 2008
40

Figure 19: Lake Mackintosh Elevation Duration Curve
Drafted: September 29, 2008
41
Figure 20: Graham Mebane Reservoir Elevation Duration Curve

Drafted: September 29, 2008
42
Figure 21: Cane Creek Reservoir Elevation Duration Curve

Drafted: September 29, 2008
43
Figure22: Glenville Reservoir Elevation Duration Curve

Note: The model assumes no water supply withdrawals from Glenville Reservoir.

Drafted: September 29, 2008
44
Figure 23: Townsend Reservoir Elevation Duration Curve

Drafted: September 29, 2008
45
Figure24: Harris Lake Elevation Duration Curve
Note: Water supply demands are assumed the same from Harris Lake for all scenarios.
Drafted: September 29, 2008
46
Figure25: High Point Reservoir Elevation Duration Curve

Drafted: September 29, 2008
47
Figure 26: Old Stoney Creek Reservoir Elevation Duration Curve

Drafted: September 29, 2008
48

Figure 27: Ramseur Reservoir Elevation Duration Curve

Drafted: September 29, 2008
49
Figure 28: Randleman Reservoir Elevation Duration Curve

Drafted: September 29, 2008
50
Figure 29: Lake Reidsville Elevation Duration Curve

Drafted: September 29, 2008
51
Figure 30: University Lake Elevation Duration Curve

Drafted: September 29, 2008
52
Figure 31: Brandt Reservoir Elevation Duration Curve

Drafted: September 29, 2008
53

VI. In-stream Flow Evaluation

Predicted stream flows at certain points of interest were evaluated. The purpose of examining in-
stream flows is to evaluate potential impacts on aquatic ecosystems, including fish and other
aquatic organisms, that could be caused by changes in flows resulting from reservoir operations
or water supply withdrawals. The following table shows the nodes of interest that were
identified through discussions with the NC Wildlife Resources Commission. NCWRC also
expressed interest in assessing in-stream flows on Rockfish Creek, Upper Little River and Rocky
River, but because of the way the system is modeled, this was not possible.

Table 5: Nodes of Interest for In-stream Flow

River Location / Section Node
Deep River Middle portion 280
Haw River Middle portion 360
Haw River Lower portion 410
Cape Fear River Lillington 550
Deep River Lower portion 640
Little River Lower portion 720
Cape Fear River Lock and Dam #3 780
Cape Fear River Lock and Dam #2 790
Cape Fear River Lock and Dam #1 820

Drafted: September 29, 2008
54
Figure 32 shows the geographic locations of the points of interest.
Figure 32: In-stream Flow Analysis Nodes of Interest

Analysis of In-stream Flows

An adaptation of the Tennant Method1 for evaluating in-stream flows was used for evaluating the
modeled in-stream flows. Under this method, daily stream flows are compared to the historical
average annual flow at the point of interest. The historical average annual flow was determined
using the model under the Unregulated scenario.

Depending on the percentage of annual flow, the Tennant Method provides guidelines for
evaluating the adequacy of the flow for the given time of the year. Table 6 summarizes these
guidelines.

1 In-stream Flows for Riverine Resource Stewardship: 2004 Revised Edition, Multiple Authors .

Drafted: September 29, 2008
55
Table 6: Modified Tennant Method Guidelines for Evaluating In-stream Flows

The Tennant method is used as a preliminary screening device to see
how projected increases in water use will affect stream flows at selected
locations. When new or increased water withdrawals are planned, the
permitting process will require site-specific in-stream flow studies to
determine required in-stream flow levels.
The following plots show an example of a summary of stream flow levels using
the Tennant Method for one of the points of interests identified in Table 5.
Daily stream flows at all points of interest were estimated using the model for
the entire 76-year record. Then, the percentage of days over the 76-year period
within each of the various stream flow ranges was calculated.
The complete summary of results at all points of interest will be made available
on the Division of Water Resources website under Cape Fear River Basin
Planning.

Description of
Flow Levels March to May June to November December to February
Level 1 < 10% of QAA* Severe Degradation Severe Degradation Severe Degradation
Level 2 10 - 20% of QAA Poor or Minimum Fair or Degrading Fair or Degrading
Level 3 20 - 30% of QAA Fair or Degrading Good Good
Level 4 30 - 40% of QAA Good Excellent Excellent
Level 5 40 - 50% of QAA Excellent Outstanding Outstanding
Level 6 50 - 60% of QAA Outstanding Outstanding Outstanding
Level 7 60 - 100% of QAA Optimum Optimum Optimum
Level 8 100 - 200% of QAA Optimum to Flushing Optimum to Flushing Optimum to Flushing
Level 9 >200 of QAA Flushing or Maximum Flow Flushing or Maximum Flow Flushing or Maximum Flow

*QAA is the Average
Annual Flow
Drafted: September 29, 2008
56
Figure 33: Stream Condition for the Middle Deep River – Dec-Feb
Stream Condition Middle Deep River (Node 280)
December- February
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
Unregulated 2003 Demands 2030 Demands 2050 Demands% of Days at Flow Level< 10% of QAA*
10 - 20% of QAA
20 - 30% of QAA
30 - 40% of QAA
40 - 50% of QAA
50 - 60% of QAA
60 - 100% of QAA
100 - 200% of QAA
>200 of QAA

*QAA (average annual flow) at Node 280 = 227 mgd

Table 7
Level Dec-Feb
Unregulated

2003
Demands
2030
Demands
2050
Demands

1 < 10% of QAA* 14.1% 9.6% 1.7% 2.6%
2 10 - 20% of QAA 16.7% 23.3% 12.5% 11.5%
3 20 - 30% of QAA 12.5% 11.7% 14.4% 15.4%
4 30 - 40% of QAA 10.8% 10.1% 10.0% 10.7%
5 40 - 50% of QAA 7.7% 6.7% 8.1% 8.3%
6 50 - 60% of QAA 6.0% 6.1% 7.1% 7.1%
7 60 - 100% of QAA 13.2% 14.0% 17.4% 16.9%
8
100 - 200% of
QAA 10.6% 10.1% 15.5% 14.8%
9 >200 of QAA 8.3% 8.4% 13.2% 12.7%

Drafted: September 29, 2008
57
Figure 34: Stream Condition for the Middle Deep River – Mar-May
Stream Condition Middle Deep River (Node 280)
March-May
Spawning Impacts
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
Unregulated 2003 Demands 2030 Demands 2050 Demands% of Days at Flow Level< 10% of QAA*
10 - 20% of QAA
20 - 30% of QAA
30 - 40% of QAA
40 - 50% of QAA
50 - 60% of QAA
60 - 100% of QAA
100 - 200% of QAA
>200 of QAA

*QAA (average annual flow) at Node 280 = 227 mgd

Table 8
Mar-May
Unregulated

2003
Demands
2030
Demands
2050
Demands

< 10% of QAA* 11.5% 9.1% 0.6% 0.6%
10 - 20% of QAA 15.1% 20.3% 10.8% 10.8%
20 - 30% of QAA 13.3% 13.2% 18.8% 19.3%
30 - 40% of QAA 10.0% 9.3% 11.7% 12.5%
40 - 50% of QAA 8.0% 7.2% 7.9% 8.3%
50 - 60% of QAA 6.4% 5.9% 6.1% 5.7%
60 - 100% of QAA 15.0% 14.6% 17.1% 16.5%
100 - 200% of
QAA 11.2% 10.9% 14.5% 13.9%
>200 of QAA 9.4% 9.5% 12.6% 12.4%

Drafted: September 29, 2008
58
Figure 35: Stream Condition for the Middle Deep River – June-Nov
Stream Condition Middle Deep River (Node 280)
June - November
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%
40.0%
45.0%
Unregulated 2003 Demands 2030 Demands 2050 Demands% of Days at Flow Level< 10% of QAA*
10 - 20% of QAA
20 - 30% of QAA
30 - 40% of QAA
40 - 50% of QAA
50 - 60% of QAA
60 - 100% of QAA
100 - 200% of QAA
>200 of QAA

*QAA (average annual flow) at Node 280 = 227 mgd

Table 9
June-Nov
Unregulated

2003
Demands
2030
Demands
2050
Demands

< 10% of QAA* 8.9% 7.6% 10.7% 11.3%
10 - 20% of QAA 14.6% 20.7% 41.0% 41.1%
20 - 30% of QAA 12.3% 11.6% 17.1% 16.5%
30 - 40% of QAA 10.6% 8.1% 7.6% 7.7%
40 - 50% of QAA 7.3% 6.8% 5.1% 5.0%
50 - 60% of QAA 6.0% 5.8% 3.6% 4.2%
60 - 100% of QAA 16.3% 15.7% 7.5% 7.2%
100 - 200% of
QAA 13.3% 13.0% 3.9% 3.7%
>200 of QAA 10.6% 10.6% 3.5% 3.3%

Drafted: September 29, 2008
59

Stream Flow Duration Curves

The following flow duration curves for the flows at Lillington and Lock and Dam #1 are
included because of their particular interest. These results and flow results for other nodes and
other model outputs are available at:

http://www.ncwater.org/Data_and_Modeling/CF/

Flows at Lillington

The duration curves for flows at Lillington have an unusual shape during low flow times. This is
explained by how the model handles Jordan Lake operations. During low flow periods, the
model operates Jordan Lake by strictly following the drought operations plan. The model
releases water from the lake to meet a downstream target flow at Lillington. For example, during
Drought Stage 1, the model meets a downstream flow of 600 cfs at Lillington. For this reason,
the curves tend to flatten at the 600 cfs flow level. The length of the flat segment represents the
amount of time the lake is operating under Drought Stage 1. The same is true for the other
drought stages but at different flow levels.

As expected, all of the demand scenarios are above the unregulated curve during low flow
periods. This is true because during low flow periods, the lake is releasing water to maintain-
stream flows above what would occur if the lake were absent. During higher flow periods, the
unregulated curve is higher than the demand scenarios. This is because during high flows, the
lake often stores much of the inflow, releasing less downstream than would occur were the lake
absent.

Flows at the Locks and Dams

The duration curves at the lock and dams are smooth. This is because the locks and dams are far
enough downstream of the lake that the lake releases are no longer the dominant factor in
determining the stream flows. However, the effects of Jordan Lake are still apparent. Flows
under the three demand scenarios are higher than the unregulated flows during low flow periods
and lower during high flow periods.

Drafted: September 29, 2008
60
Figure 36: Lillington Flow Duration Curve

Drafted: September 29, 2008
61
Figure 37: Lillington Flow Duration Curve - Close-up

Drafted: September 29, 2008
62
Figure 38: Lock and Dam #3 Flow Duration Curve

Drafted: September 29, 2008
63

Figure 39: Lock and Dam #3 Flow Duration Curve _ Close-up

Drafted: September 29, 2008
64

Figure 40: Lock and Dam #1 Flow Duration Curve

Drafted: September 29, 2008
65
Figure 41: Lock and Dam #1 Flow Duration Curve - Close-up

Drafted: September 29, 2008
66

VII. Other Model Results

All of the model results in this reports along with output for any other model nodes of interest are
available at the following Division of Water Resources website:

http://www.ncwater.org/Data_and_Modeling/CF/

Drafted: September 29, 2008
67
VII. Comments

After the meeting in May, we received some written comments from Professor Richard
Whisnant, Sydney Miller, Paul E. Peterson, Mick Noland, Mick Greeson that have helped to
improve with the final version of this document.

Jucilene,
Here is some feedback for you on the report. I thought the report was very well written,
particularly in its summary description of the model and the Jordan Reservoir storage system. I
appreciate that.

I have the following suggestions for future reports. First, I do not recall a discussion of
uncertainty in the report. I think the authors of future such reports should explicitly discuss the
uncertainty and confidence levels associated with the inputs, assumptions and outputs in the
model. Even better would be some explicit sensitivity analysis--what are the most sensitive
variables?

Second, I believe any future such reports by the Division should include a model run with
explicitly conservative low-flow assumptions. The question many people will want an answer to
is: if flows are again as low as they were in 2007-08, and demand is up to 2050 projections
including those left out of the March 2008 report, and it is on average 4 degrees F. hotter as the
IPCC says it likely will be, and there is a lot more flashiness in the basin both from urbanization
and changes in precipitation patterns, what are the concerns about water supply in the basin, for
drinking water, cooling water, irrigation and in-stream flows?

We have started a page on the water wiki (our website for collecting and discussing information
relevant to the water allocation study) that discusses these assumptions, the ones we are aware
of. I would very much appreciate any input that division modelers have on this page of the wiki.
http://sogweb.sog.unc.edu/Water/index.php/River_basin_models Having missed all the meetings
thus far on the Cape Fear model, I am sure there is a lot more I have to learn about it.

Finally, I am having trouble finding a summary of the agricultural water use data plugged into
the model. Your data page has a .pdf document describing the methodology for collecting the
data, but I assume there is a table somewhere comparable to the table for public water supplies
that was included in the March report document, showing what water use values were actually
plugged into the model. I think those values should be made explicit and public--probably they
are and I just can't seem to find them.

Thanks again to all in the division who are working on these models,
Richard Whisnant
UNC Chapel Hill
919.962.9320
Drafted: September 29, 2008
68
MEMORANDUM

Date: March 26, 2008

To: Phil Fragapane, NC Division of Water Resources

From: Sydney Miller, Water Resources Program Manager

Subject: Draft Cape Fear River Basin Water Supply Plan (March 20, 2008)

I have reviewed the Draft Cape Fear River Basin Water Supply Plan: Modeling of Future Water
Use Scenarios dated March 20, 2008. I offer the following comments and questions.

1. Pages 6 and 14 – Do the 2030 Demands and 2050 Demands scenarios use current Jordan
Lake water supply storage allocations? The 2030 and 2050 scenarios I had developed assume
increased allocations in some cases, consistent with the methodology for the previous
iteration of the Draft Cape Fear River Basin Water Supply Plan.

2. Page 7 – Natural inflows are represented by purple arrows and wastewater discharges are
represented by brown arrows in the OASIS model.

3. Pages 12-14 – Are the 2030 Demands and 2050 Demands scenarios based on the draft 2030
and 2050 scenarios that I provided on March 13, 2008? I incorporated some changes in the
“Jordan_WQ_WS_Accounts” OCL file that prevent water supply storage accounts from
going into negative quantities and also limit water supply deliveries to no more than the
amount available in storage plus inflow.

4. Page 14 – With the Jordan Lake water supply storage pool fully allocated and under a total
demand of 100 mgd in the 2050 Demands scenario, the water quality pool performs more
similarly to the 2030 Demands scenario.

5. Pages 18-51 – I think the plots of duration curves for each of the scenarios by parameter are
the best way to show the relative differences between scenarios.

6. Page 25 – In the 2050 Demands scenario with current Jordan Lake allocations and with the
corrected “Jordan_WQ_WS_Accounts” OCL file, there is a shortage in meeting
Morrisville’s demand for 9 days in the period of record.

7. Pages 41-45 – In displaying the results of the modified Tennant Method for evaluating in-
stream flows, consider showing the cumulative percentages for the flow levels as a single bar
for each scenario. For example, in the chart on page 43, rather than having 9 bars for each
scenario, there would be one bar for each of the four scenarios. The first bar (the Unregulated
scenario) would show level 1 at 0.3%, level 2 at 2.2%, level 3 at 7.5%, level 4 at 14%, etc.

Drafted: September 29, 2008
69

The only other comment I have at this time (other than those listed
below by Paul Peterson, consultant to PWC) is as follows: How will the
State verify that the Rocky River Reservoir and Randleman Reservoir are
being operated in accordance with the guidelines that are assumed in the
modeling runs? There needs to be some type of reporting requirement
that is timely enough for the Corps to use in determining release rates
during drought conditions.

Mick Noland
Chief Operating Officer
Water Resources Division
Public Works Commission of the
City of Fayetteville
Drafted: September 29, 2008
70

Mick,

I took a look at DWR's March 20, 2008 draft Cape Fear River Basin Water
Supply Plan (Modeling of Future Water Use Scenarios). I have a few initial
comments:

Page 9: For 2050, the assumed Western Wake Regional WRF discharge is 20.6
mgd. Is that consistent with current plans for this WRF? This is an
important assumption since page 14 highlights that the increased future
discharges from the Western Wake WRF will relieve stress on the Jordan Lake
water quality pool to meet the Lillington in-stream flow target.

Page 24: Fayetteville's demand at the Cape Fear demand node was reportedly
met for the full record, while it is stated that the deficit at Glenville
Lake would be met by increasing withdrawals from the Cape Fear River. Is
there some way to easily confirm that the deficit at Glenville Lake could
have been offset in all cases? If deficits are simulated for PWC and other
utilities, and could have been offset through other sources for those
systems, then aren't we underestimating total basin demand during critical
low flow periods by not simulating those offsets?

Page 25: The simulated Glenville Lake deficit of 97.1 days per year on
average (or 27% of the time) highlights the need to confirm that the deficit
at Glenville Lake could have been offset in all cases through increased
withdrawals from the Cape Fear River.

Other: Is there an outline for the complete Cape Fear River Basin Water
Supply Plan? In DWR's March 2002 draft plan, available supply was quantified
for each system and compared against the demand projections.
Is that sort of comparative analysis planned for this updated plan?

Paul E. Peterson
Malcolm Pirnie, Inc.
701 Town Center Drive
Suite 600
Newport News, VA 23606
757-873-4347 (phone)
757-593-0193 (mobile)
757-873-8723 (fax)
Drafted: September 29, 2008
71
April 30, 2008
Via Email
Mr. Phil Fragapane
DENR – Division of Water Resources
1611 Mail Service Center
Raleigh, NC 27699-1611
Subject: Cape Fear Water Model
Dear Mr. Fragapane:
The North Carolina Wildlife Resources Commission (NCWRC) provides the following
comments upon reviewing The North Carolina Division of Water Resource’s draft “Cape Fear
River Basin Water Supply Plan: Modeling of Future Water Use Scenarios” dated March 20,
2008.
Introduction
This section would benefit from the addition of several paragraphs explaining the purpose and
need for the modeling effort and how the results will be used. Explaining how this modeling
process is related to the statewide and basin water supply plans would help the reader understand
the context of the report. We believe the title of the report and the content of the report should
compatible. This document is not the water supply plan for the Cape Fear River basin, but rather
a summary of modeling results which are an important input to developing/revising the water
supply plan. Furthermore, the modeling report is useful not only for (human) water supply
decisions, but could be used for water management in a larger sense.
The first paragraph seems more appropriately located in Section II, Model Assumptions, which
should be retitled as Model Description and Assumptions. Paragraphs three through seven in
this section are further details of the model and should be moved to Section II.
Although this model is based on a fairly significant period of record, it is important to continue
to add data for the model to be as accurate as possible. Specifically, data from the 2007 drought
should be added to the Cape Fear River model as it has shed light on many areas with
insufficient water to meet demand.
Scope of the Model
The model does not appear to include the Black River or Northeast Cape Fear River sub-basins.
Are there other portions of the Cape Fear basin that are not included in the model? Please
explicitly state which portions are not included, explain why they are not part of the model, and
if they will be included in a future update.
Scenarios Modeled
Please describe how evaporation from the reservoirs was handled by the model for Scenario 1. It
should have been set to zero.
Drafted: September 29, 2008
72
It is unclear in the descriptions for Scenarios 2 - 4 exactly how the proposed plans described in
paragraphs 3-7 of the Introduction were handled in the model. Please clarify.
We suggest that a new scenario be modeled; one that fully allocates the water supply pool
(approximately 100 mgd) of Jordan Lake.
Model Assumptions – Inputs
Estimated Daily Natural Inflows – Please describe how the Estimated Daily Natural Inflow
dataset was developed. Was it based on a variety of stream flow gages in the watershed? If so,
which ones? If the gages were not on the modeled stream node, how were the gage records
adjusted for differences in watershed size. It appears from the figures (please number the
figures) on pages 4 and 5 that only two Daily Natural Inflow input nodes were used – one on the
Rocky River and one on the Deep River. Please describe why natural inflows were not used on
the many other streams in the basin.
Daily Withdrawals – During the March 20, 2008 meeting it was explained that the annual
average withdrawals for each node was adjusted on a monthly basis to reflect differences in
demand throughout the year. Please include a description of this in the report and provide an
example in a table.
The unnumbered table (please number the tables) on pages 9 – 11 lists the water withdrawals and
discharges for the base case scenario and for 2030 and 2050 projections. In some cases the table
indicates a drop in use from 2003 to 2030 and then an increase again between 2030 and 2050
(e.g., Fuquay-Varina). Please explain. It would also help if you could provide a summary table
that indicates how the number and percentage of nodes, by type of withdrawal (e.g., water
supply, industry, agriculture, etc.) showing a decrease, no change or an increase over time. For
those nodes that show no change or a decrease from the base case condition, please explain the
reason. Also, please describe how the individual water users take into account any changes in
use due to a drop in per capita use due to more efficient use or mandatory restrictions?
Please explain the assumptions and rationale for dealing with groundwater and wells. For
example, increases in the number of private and public wells could affect the baseflow of a
stream, and therefore, the Natural Inflow dataset. Also, wastewater discharges from well
systems could increase the stream flows in certain areas.
Daily Wastewater Discharges – Water reclamation and reuse is becoming more viable and
common. Often the water is applied on the land surface instead of being returned to the stream.
How does, or will, the model account for this change in water use, particularly during the
growing season, when much of it will be lost to evaporation?
The assumption of projecting the same percentage return of wastewater in the future as under the
base case scenario may not be valid in that it is basically assuming the same mixture of use by a
system’s customers. For example, if the current return percentage of 60% is based on a mix of
65% residential and 35% industrial, a future return percentage of 60% may not be correct if the
overall demand increases and the mix becomes 90% residential and 10% industrial.
Reservoir Operating Guidelines and Data – Please describe how evaporation from the reservoirs
was handled throughout the year.
Effects of Future Water Use on Jordan Lake
Drafted: September 29, 2008
73
We understand that the Western Wake Partners are investigating a discharge to Harris Lake
which would eliminate the direct discharge to Cape Fear River. Depending on available data and
the timing, this may or may not be feasible. If the discharge is moved to Harris Lake, this affects
several things: 1) the discharge to Cape Fear River would be reduced by about 38 mgd; 2) this
would affect the model predictions for Jordan Lake and downstream, particularly for 2030 and
2050; 3) this could potentially reduce the amount of water Progress Energy withdraws directly
from Cape Fear River; 4) currently Harris Lake does not have a minimum release, if the new
reactors are approved and new construction on the dam occurs, it’s likely this would include a
minimum release from the dam, but the anticipated minimum release of 20 cfs might be affected.
Please discuss these possibilities.
The report includes several figures and tables showing the percent of time that Jordan Lake is at
various elevations. However, it is equally necessary to understand the absolute number and
duration of occurrences that elevations drop below a given point. For example, the figure on
page 17 should indicate the number of times that the stage 1, 2 and 3 triggers are made, along
with the median and maximum duration of the events.
Water Supply Demands vs. Delivery
The second paragraph on page 24 include the sentence: “In general, for the Cape Fear model,
water supply demands have a higher weight than the in-stream flow needs, and therefore are met
first.” We are unsure how to interpret the meaning of this sentence. As used here, what does the
term “weight” mean? Did the model logic require that water supply demand be met before
providing an in-stream flow or was a percentage of in-stream flow met?
We recognize the need to meet public water supply demand, but since the model does not yet
include the worst drought on record, we believe the model should be extremely conservative in
allocating water to meet future water supply demands (e.g., a local government requesting an
increase from 40 mgd to 90 mgd). The model should err on the side of leaving more water in the
river rather than allocating the maximum amount available based on the current model. We
strongly request that we be included in the process of developing appropriate in-stream flow
targets and requirements so that the model can provide a better indication of the areas of concern.
Water Supply Intake Impacts
The report states: “In case where a water system has more than one water source, one reservoir
could be allowed to be exhausted while holding the other source in reserve.” Is this how some
water supply reservoirs are actually managed, or is this just how the model was designed? We
believe that there should be some provision, in the model and in real practice, to protect the
aquatic resources of the reservoirs from being eliminated due to a complete drawdown.
In-stream Flow Evaluation
Our interest in the in-stream flow evaluation is at two levels – a general concern for protecting
the fish and wildlife resources of the Cape Fear basin, and a specific concern for protecting rare
species. There are two federally endangered species in the Cape Fear Basin and a number of
state-listed species. The Shortnose Sturgeon is located downstream of lock and dam #1. The
Cape Fear Shiner is found in the Deep, Rocky, Haw, and Cape Fear rivers, although the
populations in the Haw and Cape Fear are very small. The complete list of priority species in the
Cape Fear basin is quite large (Table 1).
Drafted: September 29, 2008
74

Table 1. Priority aquatic species in the Cape Fear River Basin as listed in the “North Carolina
Wildlife Action Plan (2005).
Group Common Name State Status (Federal Status)
Shortnose Sturgeon Endangered (Endangered)
Cape Fear Shiner Endangered (Endangered)
Atlantic Sturgeon Special Concern
Highfin Carpsucker Special Concern
Thinlip Chub Special Concern
Carolina Darter Special Concern
Bluefin Killifish Special Concern
Least Killifish Special Concern
Broadtail Madtom Special Concern
Sandhills Chub Special Concern
Roanoke Bass Significantly Rare
Pinewoods Shiner Significantly Rare
Carolina Redhorse Significantly Rare
Snail Bullhead
Everglades Pygmy Sunfish
Banded Pygmy Sunfish
Blackbanded Sunfish
Banded Sunfish
Lake Chubsucker
Banded Killifish
Lined Topminnow
Dollar Sunfish
Spotted Sunfish
Notchlip Redhorse
Shorthead Redhorse
V-lip Redhorse
Comely Shiner
Ironcolor Shiner
Taillight Shiner
Sea Lamprey
Fish
Sailfin Molly
Brook Floater Endangered
Barrel Floater (possibly extirpated) Endangered
Atlantic Pigtoe Endangered
Yellow Lampmussel Endangered
Savannah Lilliput Endangered
Carolina Creekshell Endangered
Triangle Floater Threatened
Roanoke Slabshell Threatened
Eastern Lampmussel Threatened
Eastern Pondmussel Threatened
Creeper (Squawfoot) Threatened
Notched Rainbow Special Concern
Pod Lance Special Concern
Cape Fear Spike Special Concern
Eastern Creekshell Significantly Rare
Box Spike
Carolina Slabshell
Mussels
Variable Spike
Greensboro Burrowing Crayfish Special Concern
Carolina Ladle Crayfish Significantly Rare
Sandhills Spiny Crayfish Significantly Rare
Croatan Crayfish Significantly Rare
Crayfish
Edisto Crayfish
Greenfield Rams-horn Endangered
Magnificent Rams-horn Endangered
Snails
Rotund Mysterysnail Significantly Rare

Drafted: September 29, 2008
75
Based on the list of rare species and other factors, the North Carolina Wildlife Action Plan
(2005) lists the following watersheds in the Cape Fear River basin as priority watersheds for
conservation (see Figure 1 below):
• Upper Haw River
• Middle Haw River tributaries
• Deep/Rocky/Haw/Cape Fear Rivers
• New Hope Watershed above B. Everett Jordan Reservoir
• Cape Fear sandhills tributaries
• Lower Cape Fear/Black/South Rivers
• Northeast Cape Fear River
• Town Creek
• Merrick’s Creek/Holly Shelter Game Lands
• Orton Pond/Military Ocean Terminal Sunny Point
Currently, the Cape Fear hydrologic model analyzes nodes only the mainstem Cape Fear and
three major tributaries. In order for this model to be more useful for understanding potential
impacts to in-stream flows, we would need to see the daily flows for the period of record for
each scenario at each junction node and gage node. Additional nodes may need to be added to
represent some of the priority watersheds. A good quality map clearly showing each node and
arc is also required. The maps in the document and on the web site are of poor quality.

Drafted: September 29, 2008
76
Figure 1. Maps from the Wildlife Action Plan (2005) showing watersheds containing state and
federal listed species (left) and priority watershed for conservation (right.

We also reviewed the separate document “Cape Fear River Basin Model In-stream Flow
Analysis”. For each node we suggest that the median flow be reported in addition to the mean
annual flow. The graph for node 280 (December – February) is missing. The graph and table for
node 640 (March – May) is repeated.
In general, the in-stream flow analysis indicates that most of the nodes modeled showed a
moderate to substantial increase in the amount of time at Level 2 (10-20% of QAA) under the
current and future scenarios compared to the Unregulated scenario for the June to November
time period. These impacts were seen on the Deep, Haw and Cape Fear Rivers, all of which are
included on our list of priority watersheds for protecting rare species. The Cape Fear River at
Lillington showed impacts during all months, with more than 50% of the time in Level 2,
compared to 21% in the Unregulated scenario.
The analysis seems to show that the impacts are less at the lowermost nodes, suggesting that
greater impacts may be occurring on the smaller rivers and streams in the upper portions of the
basin. If this is the case, our need to have additional data for the smaller streams using this
model or some other approach becomes critical.
Finally, as we stated in our February 12, 2008 letter, we believe that the analysis conducted as
part of this study should only be used as a basic screening tool, not to analyze scenarios at a site-
specific level, and we have concerns about using the Tennant method in North Carolina. We
repeat our recommendations that other methods of screening (e.g., IHA and ELOHA) should be
investigated as tools for understanding the impacts to in-stream flows.
General Comments
As population increases in the Cape Fear River, so will the demand for water. Minor droughts
may be exacerbated by larger populations and the perceived need to maintain water-dependent
landscaping. Local water conservation plans vary widely in terms of conservation measures and
triggers for implementing conservation measures (e.g., voluntary, mandatory, etc.). The
disparities in the initiation and magnitude of conservation measures among local water suppliers
led to confusion among the populace and a lack of confidence in the State’s ability to manage
water. We recommend the development of basinwide water conservation plans based on
primarily on hydrological triggers (e.g., reservoir level, water quality and/or water supply pool
remaining, etc.), and less on financial triggers. When the hydrologic trigger is reached, then all
local governments should be required to adopt similar water conservation measures. The model
should be updated to include a basinwide conservation plan (triggers and reductions) to
understand how the impacts to water supplies and in-stream flows would be reduced.
The 7Q10 table in the first draft of the report was useful and should be included in the final
report as a way to understand how water quality, and therefore habitat, are expected to be
affected by current and future scenarios.
Requests to construct “off-line” reservoirs and skim high flows have been increasing and it is
likely these will increase even more following this drought. While one or two of these projects
may not have a significant impact on in-stream flow, we expect that multiple off-line reservoirs
Drafted: September 29, 2008
77
could affect the river system. Updates to the model should take into account projected use of this
technique.
Thank you for the opportunity to provide input. If DWR is contemplating similar efforts for
other river basins, please let us know at an early stage so we may have sufficient time to provide
you with meaningful input. If you have any questions concerning these comments, please
contact me at 828-652-4360 ext. 223.
Sincerely,

Christopher Goudreau
Special Projects Coordinator

Drafted: September 29, 2008
78