Surface-water quality and flow Modeling Interest Group
USGS, Water Resources Division
10615 SE Cherry Blossom Drive
Portland, OR 97216
Internet: sarounds@usgs.gov
Phone: (503) 251-3280
FAX: (503) 251-3470
with help from
Paul Conrads, Wladmir Guimaraes, and Toby Feaster
USGS - Columbia, South Carolina
So, what's an Oregonian doing writing about South Carolina projects? Well, I was recently in the SC district helping out with a district review, and I was impressed by the number of modeling projects that were in progress. These notes are only meant to sketch out the basics of these projects; some of the text was taken from project proposals and project reviews. For more information, please consult the project chiefs.
Charleston Harbor is located near the middle of the South Carolina coast and is formed by the confluence of the Ashley, Cooper, and Wando Rivers (map, 113k JPEG). The Ashley and Wando Rivers have little freshwater inflow; water quality is greatly affected by the biochemical oxygen demand (BOD) that emanates from the adjoining tidal marsh flats. The Cooper River receives freshwater input via a flow diversion from the Santee River Basin, through Lake Moultrie. Vertical and longitudinal density and salinity gradients are present in all three rivers.
The primary water quality problems in Charleston Harbor and its three main tributaries are:
In particular, Paul Conrads, working with water-quality modelers from the South Carolina Department of Health and Environmental Control (SCDHEC), has developed flow and water-quality models of the Ashley, Cooper, and Wando Rivers using the USGS models BRANCH and BLTM. Of course, a two- or three-dimensional model of the entire harbor, linked to simpler models of the rivers is the ideal, and project personnel were involved in an attempt to develop such a model, but the main focus of this project was the development of the one-dimensional river models.
Originally, the plan was to use the EPA model WASP4 for the water-quality simulations. Because of some initial problems with excessive numerical dispersion, however, parallel development of both WASP4 and BLTM models was pursued. In the end, the performance of BLTM was better than that of WASP4, especially for the Wando River, and further use of WASP4 was abandoned. It is really pretty amazing how well the BRANCH/BLTM combination was able to simulate salinity and dissolved oxygen in these rivers, despite the multidimensional nature of the hydrodynamics.
Paul is currently testing the 1D models of the Ashley, Cooper, and Wando Rivers with several hypothetical scenarios to explore their response to changing hydrodynamic and water-quality inputs. Several reports are in the works; one is already out:
Located in the north central part of South Carolina, the Catawba River (map, 16k GIF) flows south from Lake Wylie near the Charlotte metropolitan area downstream to Lake Wateree. The regional utilities in North and South Carolina have discussed siting a large wastewater treatment facility in the area that would discharge between 25 and 100 million gallons per day of treated effluent into this watershed. Before that can occur, the South Carolina Department of Health and Environmental Control (SCDHEC) would require the development of a dynamic water-quality model of the 26-mile reach of the Catawba River downstream of Wylie Dam.
The objectives of this study are to:
Data collection to date has included a wide range of discharge and water-quality measurements as well as dye travel-time and reaeration studies. In particular, the reaeration studies should prove to be quite interesting. While USGS personnel were using the standard propane method, a crew from EPA was using, in the same reach, a new method that employs krypton gas as the tracer for air/water exchange. A comparison of these two methods will be very useful.
Planned products include a USGS Water Resources Investigations Report that describes the data analyses, documentation of the model calibration/validation and sensitivity testing, and results of the model simulations. The final calibrated model will be released to the cooperator (Lancaster County Water and Sewer District) and to SCDHEC for their use.
The Wateree River watershed has an area of 358 square miles and is located in Kershaw, Sumter, and Richland Counties, SC (map, 35k GIF). Within this watershed, the Wateree River flows from Lake Wateree Dam to its confluence with the Congaree River, a reach of 76.8 miles.
A number of industries and municipalities discharge treated effluent into the Wateree River. These point source discharges are regulated by the South Carolina Department of Health and Environmental Control (SCDHEC). Currently, the quantity and quality of such discharges are based on insight derived from a one-dimensional, steady-state flow and water-quality model. Unfortunately, the model does not cover the entire reach of interest and the hydrodynamics of the Wateree River are far from steady-state. Duke Power Company operates the Wateree Dam for hydropower generation, and flows are controlled at least in part to meet peaking power demand. As a result, releases from the dam vary from 300 to 12,000 cubic feet per second, sometimes over the course of a single day.
The cooperator in this project, the Kershaw County Water and Sewer Authority (KCWSA), is interested in monitoring and modeling the effects of the releases from Wateree Dam on the dissolved oxygen concentrations in the Wateree River. Therefore, the objectives of this project are to:
One particularly interesting result from the data collected thus far is that the largest dissolved oxygen sag observed in this reach occurs upstream of the permitted industrial and municipal dischargers (example profile, 17k GIF). The sag occurs downstream of a riffle reach that is located just downstream of Wateree Dam. By the time the water reaches the first dischargers, the sag typically has reached its minimum and begins to recover. So, although the current dischargers don't appear to be having a substantial detrimental effect on the dissolved oxygen concentration, the capacity of the Wateree River to assimilate further oxygen-demanding loads may be nonexistent under certain conditions. The model being developed in this study will help to quantify that capacity.
Several theories have been investigated regarding the potential cause of the dissolved oxygen sag in the upper part of the study reach, including:
Planned products from this investigation are a USGS Water Resources Investigations Report describing the data collection and analyses as well as documentation of the model calibration and validation. The final calibrated model will be delivered to the KCWSA and the SCDHEC for their use. The USGS will provide some amount of technical support and training to the KCWSA and the SCDHEC for model usage.
Results are quantified on graphs showing the number of days required to move a particle of water (retention time or days-to-flush) from Durham Canal to the Charleston Commissioners of Public Works (CPW) intake, reflecting changes in the amount of withdrawal by Charleston CPW; the location of the CPW intake at its present location on Foster Creek or at a new location 0.9 mile north of Foster Creek, on the Back River; flow through hypothetical flap-type tide gates at (6-foot concrete pipes) Bushy Park Dam; and whether a thermoelectric power plant is withdrawing water from the reservoir.
The withdrawals by the thermoelectric power plant are large enough to improve the quality of water in the reservoir from Durham Canal to its intake to a degree that the water can be economically treated by CPW. Target flushing rates of 3.1 and 5.2 days were established for hypothetical CPW intakes located 0.9 mile north of Foster Creek on the Back River and at the current Foster Creek intake, respectively. This flushing rate uses the same flushing rate achieved by the power plant. Combined maximum CPW withdrawals from the Edisto River and Foster Creek are 50, 118, and 150 million gallons per day (Mgal/d) for current, short-term, and ultimate demand projections, respectively. With 50 Mgal/d withdrawals at Foster Creek, the days-to-flush is about twice the estimated target for eight 6-foot concrete pipes. If the withdrawal rate is increased to 118 or 150 Mgal/d for the same number of pipes, the target rate is exceeded by one day. If the CPW intake is moved to the site on the Back River, the target days-to-flush can be reached by withdrawals of 50 to 150 Mgal/d with six to eight 6-foot concrete pipes. Significant improvement in flushing characteristics could be achieved if the intake was located on the Back River, 0.9 mile north of Foster Creek. A sensitivity analysis showed that flushing rates were insensitive to model roughness estimates, cross-section datums, or boundary-condition stage datums.
Low-flow surface-water samples were generally free of toxic compounds with the exception of laboratory artifacts and naturally occurring trace metals. Storm-runoff samples generally contained very low concentrations (near detection limits) of a small number of volatile and semivolatile organics and naturally occurring trace metals. Concentrations of toxic compounds in excess of current (1995) South Carolina Department of Health and Environmental Control and U.S. Environmental Protection Agency regulations were not detected in surface-water samples collected from Foster Creek. Chemical analyses of streambed sediments indicated minimal anthropogenic effects on sediment quality.
The particle-tracking option of the U.S. Geological Survey one-dimensional unsteady-flow model BRANCH indicated that as the simulated volume of rainfall runoff increased in the Foster Creek Basin, simulated particles in Foster Creek were transported greater distances. Simulating flow through the Bushy Park Dam (also known as Back River Dam) had little effect on particle movement in Foster Creek. Simulating typical withdrawal rates at a water-supply intake resulted in a slight attraction of particles toward the intake during conditions of relatively low runoff. These withdrawals had a greater influence on particles downstream of the intake than on those upstream of the intake. Simulations confirmed earlier findings which suggested that the creek would not flush during baseflow conditions, with the exception of the lower 1-mile reach, where flushing results from tidal movements. According to the simulations, Foster Creek will fully flush if a 2-year, 7-day storm occurs. Flushing appears to be affected more by the total volume of storm runoff than by typical municipal withdrawals or tidal effects.
Back to the SMIG Features Page
Home | Mailing List | Features | Conferences | Classes | Reading | Model Archives | Feedback
