Project Overview

Full Title
Apparent Flow Losses in the Cape Fear River

Location
 Chatham, Wake, Lee, and Harnett Counties

Cooperating Agencies
Cape Fear Flow Study Committee of the Triangle J Council of Governments, NC Division of Water Resources

Project Chief
J. Curtis Weaver
jcweaver@usgs.gov

Period of Project
September 2008 - December 2010

Team Members
Geoff Cartano

USGS IN YOUR STATE

USGS Water Science Centers are located in each state.

Apparent Flow Losses in the Cape Fear River

Project Summary

Cape Fear River looking upstream from Buckhorn Dam

Background

The U.S. Army Corps of Engineers is required to release water from B. Everett Jordan Lake such that a minimum (target) flow of 600 ft³/s (± 50 ft³/s) is measured at the USGS streamgaging station located on the Cape Fear River at U.S. Highway 401 at Lillington (USGS Sta. 02102500).

The drainage area in this 24-mile reach increases from 1,689 mi² at Jordan Lake dam to 3,464 mi² at the Lillington gaging station (resulting in an intervening drainage area of 1,775 mi²). Much of the intervening drainage area in this reach includes the Deep River basin, and the downstream-most gaging station on the Deep River at Moncure (USGS Sta. 02102000) has a drainage area of 1,434 mi².

During some low-flow conditions, the sum of the estimated releases from Jordan Lake and the measured streamflow on the Deep River at Moncure is greater than the streamflow recorded downstream at the gaging station at Lillington. Consequently, either the target at Lillington is not met, or excessive water is released from Jordan Lake in an attempt to meet the target.

The difference between the inflow to this area (sum of releases from Jordan Lake dam and flow in Deep River at Moncure) and outflow from the reach (flow in the Cape Fear River at Lillington) could be because either (1) flow is measured inaccurately at one or more of the 3 measurement locations or (2) flow is being "lost" from the reach from surface-water diversions, possible alterations due to storage behind dam structures, ground-water withdrawals, evaporation, recharge (loss) to the ground water system, or a combination of these factors.

Figure 1. Map showing proposed study area of Cape Fear River between B. Everett Jordan Lake and Lillington.

The proposed study area is located in the Cape Fear River Basin from the dam at B. Everett Jordan Lake, an impoundment of the Haw River in Chatham County, downstream to U.S. Highway 401 near Lillington in Harnett County (figure 1). Jordan Lake has been in operation since February 1982, and the dam (USGS Sta. 02098197) is located 4.1 miles upstream from the confluence of the Haw and Deep Rivers where the Cape Fear River takes its name (figure 1). The study area also includes the downstream-most 4.5 miles of the Deep River (between the USGS gaging station located at Moncure and mouth of the Deep River). The confluence of the Haw River and Deep River is 19.7 miles upstream from the U.S. Highway 401 bridge at Lillington.

The Corps of Engineers estimates flows at Jordan Lake dam based on a rating between gate opening and flow. Streamflow records on the Deep River at Moncure began in July 1930; records began on the Cape Fear River at Lillington in December 1923.

The combined drainage area of the Deep River at Moncure and the basin upstream from Jordan Lake dam is 3,123 mi². Consequently, an additional area of 341 mi² drains to the Haw and Cape Fear Rivers between these two locations and the gaging station at Lillington. The drainage areas of these streams in the study reach have not all been determined.

The major streams draining to the study reach are (from upstream to downstream):

• Shaddox Creek - 20 miles upstream from U.S. Highway 401 bridge;
• Wombles Creek - 18.7 miles upstream from U.S. Highway 401 bridge;
• Lick Creek - 17.3 miles upstream from U.S. Highway 401 bridge;
• Gulf Creek - 16.5 miles upstream from U.S. Highway 401 bridge;
• Bush Creek - 14.9 miles upstream from U.S. Highway 401 bridge;
• Fall Creek - 13.2 miles upstream from U.S. Highway 401 bridge;
• Buckhorn Creek - 12.8 miles upstream from U.S. Highway 401 bridge;
• Parkers Creek - 11.8 miles upstream from U.S. Highway 401 bridge;
• Daniels Creek - 10.4 miles upstream from U.S. Highway 401 bridge;
• Cedar Creek - 9.2 miles upstream from U.S. Highway 401 bridge;
• Camels Creek - 8.1 miles upstream from U.S. Highway 401 bridge;
• Avents Creek - 7.4 miles upstream from U.S. Highway 401 bridge;
• Fish Creek - 4.7 miles upstream from U.S. Highway 401 bridge; and
• Hector Creek - 4.0 miles upstream from U.S. Highway 401 bridge.

Consideration of possible flow losses is complicated by the presence of several factors that may affect the overall flow patterns in this reach:

The first factor is low-head, run-of-river dams located in the study area. The Lockville Dam is located on the Deep River about 2.7 miles upstream from the mouth of the Deep River (figure 1). Recent information obtained about this dam indicates that no power production currently occurs at this structure, and that run-of-river conditions are in effect across the dam. No other information or data is known to exist that would provide insight into flow patterns occurring as a result of this dam.

Figure 2. Buckhorn Dam on Cape Fear River between Jordan Lake and Lillington.

The Buckhorn dam (USGS Sta. 02102178, figure 2) is located in the Cape Fear River at 13.8 miles upstream from U.S. Highway 401 bridge, and 5.9 miles downstream from the confluence of the Haw and Deep Rivers (figure 1). Approximately 1,150 feet in length, the dam was completed and filled in 1908, then used until December 1962 by a regional power producer to regulate flows for the production of power (Walters and others, 2005). An estimated surface area of the river upstream from the dam up to Jordan Lake dam and the Lockville dam is about 460 acres. The volume of water in storage behind the dam is reported as being 69.7 million ft3 (Walters and others, 2005).

A second factor is the presence of water withdrawals from the river in the study reach. The largest withdrawal is associated with the operation of a coal-fired power plant owned by the regional power producer and located adjacent to the Cape Fear River just downstream from the confluence of the Deep and Haw Rivers. Water is withdrawn from the river, then discharged to a canal that generally runs parallel to the mainstem, merging back into the river immediately upstream of the Buckhorn dam. In a recent report on the low-flow characteristics in the Cape Fear River basin (Weaver, 2001), the average daily withdrawal and return discharge in 1998 for the plant was about 207 and 204 Mgal/day, respectively. Long-term consumptive use of water by the plant is not known at this time. However, a stream gaging station was installed on the canal during March 2008 and should help provide some additional information about the flow diversions through this facility.

A second water-supply withdrawal is located on the Cape Fear River just upstream from the gaging station at Lillington and the associated wastewater discharge is located downstream from the gaging station There are no other known major public surface-water supply withdrawals in the study reach, but there may be other small withdrawals.

Ground-water withdrawals from a quarry operation near the Lockville dam may potentially intercept ground-water that would be otherwise discharging to the river.

A third possible factor may be the underlying geologic setting and Fall Line transition from Piedmont physiographic province to Coastal Plain physiographic province. For most of the intervening drainage area between Jordan Lake dam and Lillington (with exception of most of Deep River basin), the underlying geologic rocks are Triassic Basin rocks, which generally have low permeabilities. Where streamflow data has been collected in areas within the Triassic Basin, the potentials for sustained base flows are extremely low. The presence of the Fall Line transition between the physiographic provinces raises the question about possible flow losses attributed to natural causes in these hydrogeologic formations as opposed to man-made causes.

Objectives

The objectives of the proposed study are to:

1. Determine the conditions under which flow apparently is lost from the study reach (actual loss or measurement error)

2. Determine the reasons for the loss (whether real or apparent?). Real losses would be based on consumptive use somewhere within the intervening drainage area or presence of a losing reach due to natural factors (underlying geology). Apparent losses would be reflective of flow dynamics that occur primarily as a result of diversion through canal, storage or potential flow regulations caused by the dams, and/or measurement error at observation points.

Approach

To fulfill the objectives of the study, the proposed work will include six main tasks as follows:

1. Re-establish a streamgaging station to obtain continuous records of discharge below the Jordan Lake dam. Because flow conditions below the dam are affected by backwater from the Buckhorn dam and the Deep River confluence, hydroacoustics instrumentation will be needed to collected discharge records. Discharge records collected at this site will help verify the outflows from Jordan Lake dam.

2. Establish a stage gage at the Buckhorn dam to continuously monitor the stage under various conditions in the Cape Fear River. Indirect measurements of discharge may be possible using the dam as a weir to compute flows based on the stage, depending on the sensitivity of the dam as a weir.

3. Install six shallow ground water piezometer clusters and continuously monitor temperature and ground-water level for at least 12 months. The piezometer clusters will be located at three transects in the study reach. The transects will be located adjacent to the Cape Fear River upstream and downstream from the Buckhorn dam, and downstream of the confluence of the Haw and Deep rivers. Within each transect, two piezometer clusters will be located in sediments beneath the river bed, along a cross section from the bank toward the center of the river channel. Each cluster will contain one shallow piezometer (<1 meter deep) and two deep piezometers (>1 and 2 meters deep) for vertical gradient determination. Data from the wells and the river stage at the Buckhorn dam will be used to identify periods during which river may be losing water to the ground water system, as indicated by the hydraulic and thermal gradients in the shallow ground water system.

4. Complete data reviews and analyses as follows:

   • Analyze streamflow records to identify periods during which the outflow from the study reach is less than the inflow to the reach. Data collected during the period 1982-2007 will be analyzed (data collected following the completion of Jordan Lake dam). The conditions (primarily season and magnitude of flows) under which outflow is less than inflow within the study area will characterized.

   • Review streamgaging station stage-discharge ratings at Cape Fear River at Lillington and Deep River at Moncure. Particular emphasis will be given to those flows for which the records analysis indicated that outflow was less than inflow. The effects of small changes in the ratings on the computed inflow and outflow will be characterized.

   • Work with North Carolina Division of Water Resources to obtain and compile historical records of withdrawals and discharges to the study reach. The effects of these withdrawals and discharges on changes in flow magnitude within the study reach will be characterized.

5. Conduct five sets of synoptic streamflow measurements in the study reach. Measurements will be made at the upstream and downstream ends of the study reach, at the mouths of selected ungaged tributaries, and within the study reach at locations upstream and downstream of the tributaries on which measurements are made. Each set of measurements will include about 10 individual measurements. The measurements will be made over a two-day period when flows and releases from Jordan Lake dam are steady. Measurements will be made for flow conditions that were previously identified as potentially having outflows less than inflows.

   Discharge measurements are generally considered to be accurate within plus or minus 5 percent of the actual flow. Consequently, if the difference between the outflow and inflow is less than 5 percent, losses may be difficult to quantify from the synoptic measurements. However, the uncertainty associated with the measurements, and the effects of the uncertainty on conclusions regarding losses (and stage-discharge ratings) will be quantified.

   Prior to the measurements, all known facilities withdrawing or discharging water to the study reach will be contacted and requested to maintain detailed records of withdrawals and discharges during the measurement period.

6. If data collection at any of the 3 piezometer transects reveal fluctuations in the temperature indicating a potentially losing reach, the spatial variation in temperature will be measured using fiber optic technology along that section of the Cape Fear. A distributed temperature system will be anchored onto (or embedded in) the river bed along a relatively long stretch of reach (possibly from a 1,000 feet to 0.5 mile?) during a time of year when the difference between the water temperature in the Cape Fear river and the ground water is at a maximum. Unlike monitoring temperature in piezometer clusters, the distributed temperature system will detect spatial variability in fluid exchanges, which may be missed in point measurements. If data collection at the 3 piezometer transects do not indicate any temperature fluctuations, then this task will not be carried out, resulting in an adjustment to the overall project costs.

   Complete interpretive analyses of the ground-water data and other newly-collected data (discharge records from acoustical streamgage, synoptic measurements), will be discussed in a USGS report summarizing all results of the investigation.