Developing a Protocol for Water Use During a Drought

A four-year drought in the Catawba-Wateree Basin in North and South Carolina highlighted the importance of planning for water use during low water periods. Duke Energy used computer modeling to develop a drought response plan that will allow continued withdrawals by all users while ensuring adequate water levels in the 11 reservoirs in the basin.

By Edward D. Bruce

Duke Energy provides electricity to areas of North Carolina, South Carolina, Ohio, Indiana, and Kentucky. The 13 powerhouses in the 814-MW Catawba-Wateree Hydroelectric Project in North and South Carolina play a major role in providing electricity for Duke Energy’s customers.

The Federal Energy Regulatory Commission (FERC) operating license for the Catawba-Wateree project was issued in 1958 and will expire August 31, 2008. Duke Energy filed an application for a new license in August 2006. This application included a comprehensive relicensing agreement signed by 70 stakeholders. As part of the relicensing process, Duke Energy developed a drought response plan. This plan sets forth specific steps to be taken in the event of a drought, to ensure adequate water for all the uses of the project.

Understanding the Catawba-Wateree Basin

The Catawba River begins in western North Carolina and flows east and then south into South Carolina, where it joins Big Wateree Creek to form the Wateree River. From upstream to downstream, the 11 reservoirs are: Lake James, Lake Rhodhiss, Lake Hickory, Lookout Shoals Lake, Lake Norman, Mountain Island Lake, Lake Wylie, Fishing Creek Reservoir, Great Falls Reservoir, Cedar Creek Reservoir, and Lake Wateree. Construction of the hydro developments began in the early 1900s. The final development, 350-MW Cowans Ford in North Carolina, was completed in 1963.

The project spans more than 225 miles of river, with a drainage area of 4,750 square miles. The reservoirs are adjacent to several large urban areas, the largest being the Charlotte metropolitan area.

One unique aspect of this hydro project is that the water is used in large quantities for purposes other than electricity generation. The project provides cooling water for Duke Energy’s two nuclear and three coal-fired facilities and drinking water for more than 1.3 million people. During normal flow periods, there is enough water in the reservoirs to meet all of the water uses. However, during sustained periods of very low inflow, water in the reservoirs is needed to provide for all the water needs. Potential water use conflicts can occur.

Experiencing a four-year drought

Until the late 1990s, Duke Energy assumed that the Catawba-Wateree project could supply the region’s water needs during even the worst drought periods. However, beginning in mid-1998, rainfall in the southeast and in the Catawba-Wateree Basin fell as much as 25 to 30 percent below average. This drought continued for four years. In the last two years of the drought, Lake James, the uppermost reservoir, could not be refilled over the winter to target elevations needed to sustain downstream water needs through the summer and fall.

In the late summer of 2002, Duke Energy called a meeting of all the water suppliers that withdrew water from the project reservoirs. The purpose of this meeting was to inform them that if the drought lasted another year, water in some of the reservoirs would be below the level of the water intakes. This situation would have major implications for Duke Energy because about half of the company’s generating capacity in the Carolinas depends on the Catawba-Wateree project for cooling water.

Fortunately, rainfall in the basin in the next 12 months was much higher than normal. The reservoir and groundwater levels returned to normal in 2003.

In February 2003, Duke Energy issued the first stage consultation document for FERC relicensing of the Catawba-Wateree project. At the time, Duke Energy was starting to develop a water quantity modeling and analysis methodology for use during relicensing.

During the spring of 2003, Duke Energy received many study requests from stakeholders in response to the first stage consultation document. There were several requests to study water use and supply and develop a drought response protocol.

Performing a water supply study

One of the 31 studies completed during the relicensing process was a water supply study. The primary goals were to document existing water withdrawals and returns and project future withdrawals and returns for 50 years (the maximum term of a new license for the project). The projected water withdrawals and returns included public water suppliers, industrial uses, agricultural uses, and cooling water evaporation for existing and future thermal power plants. Most of the non-hydro use of the water was for public water consumption and thermal power plant cooling. Figure 1 shows the existing and projected distribution of water use.

Figure 1: Current outflows (at left) from the 814-MW Catawba-Wateree project indicate nearly half (47 percent) of the water is used for power. Projected net outflow (at right) from the project in 2058 shows that much more of the water (51 percent) will be used for public water supply, with corresponding decreases in the other uses.
Click here to enlarge image

Duke Energy then used the water use data developed during the water supply study to populate CHEOPS, a hydraulic water quantity computer model developed by Devine Tarbell and Associates. Duke Energy chose CHEOPS because it accurately models hydroelectric operations and has the ability to address questions concerning drought operations and water supply withdrawals.

The utility wanted to use this model to evaluate future license operating scenarios. Figure 2 on page 58 shows the withdrawal amounts used to populate the model. The hydraulic model used an historical 51-year (1953-2003) inflow data set, developed by analyzing historical operating and water withdrawal records, to represent future hydrology in the basin and to introduce random variation patterns of future inflows.

The drought of record inflow occurred at the end of the 51-year data set. The inflow during this period matched up with the projected highest net withdrawal amounts, creating a worst case modeling scenario for water use and inflows.

Another component of the water supply study involved determining the critical elevation of water for each reservoir. Critical elevation was defined as the water level below which any large water intake or any regional power plant intake located on the reservoir would not operate at its design capacity.

In January 2005, Duke Energy formed an ad-hoc study committee to evaluate potential future operating scenarios for the Catawba-Wateree project. The committee developed a water storage flexibility spectrum to illustrate the potential range of future operating options. The license conditions that could be analyzed ranged from baseline to run-of-river conditions. These conditions were deemed to be the limits of what was theoretically possible given the current infrastructure demands on the project. All potential future operating conditions to be analyzed were required to be within those limits.

Early in the relicensing process, stakeholders identified 2,098 individual interests which were consolidated into similar interest categories to develop 225 composite interest statements. Duke Energy then developed a performance measures sheet to quantify how well proposed scenarios met composite interest statements related to water quantity. This sheet grouped the measures into fish and aquatics, recreation, water user, and other interests.

The sheet also included the minimum increment of significant change for each performance measure. This indicated the threshold for that measure such that if the output of two scenarios for a particular criterion does not differ by more than the minimum increment of significant change, there is no significant difference between those two scenarios. For each scenario model run in CHEOPS, the output data was imported into the performance measures sheet. The baseline scenario’s performance measures were always recorded and compared to each of the other scenario runs.

Developing a drought response plan

The stakeholders in the Catawba-Wateree relicensing process realized that a drought response plan would be vitally important to the future of the Catawba-Wateree River Basin. Duke Energy and the stakeholders agreed that a successful drought response plan needed to include actions by other water users, in addition to Duke Energy. Stakeholders also agreed that the plan should be triggered based on more than one monitored drought indicator. Based on these needs, a drought response plan, called the Catawba-Wateree Low Inflow Protocol (LIP), was developed during the relicensing process.

The LIP contains five stages of drought specific to the basin: Stages 0 (watch stage), 1, 2, 3, and 4. Each subsequent stage represents a more severe drought condition. The decision to declare a certain stage of drought depends on three independent triggers:

    – Remaining usable storage in the project reservoirs;
    – Streamflow gages measuring inflow to the project; and
    – The highest U.S. Drought Monitor reading in the basin.

Figure 2: This figure shows data on current and projected water use by other users of the water from the 814-MW Catawba-Wateree Project (besides the hydro stations). This data was fed into the model Duke Energy used to develop a drought response plan.
Click here to enlarge image

Any one of the three triggers can induce a watch stage. To induce other stages, the remaining usable storage trigger and at least one of the other two triggers must be met.

For public water system owners, the LIP calls for voluntary water use restrictions for Stage 1, mandatory restrictions for Stages 2 and 3, and emergency restrictions for Stage 4. This was a monumental task because it required all the public water system owners in the basin to agree to the triggers for voluntary, mandatory, and emergency stages and to agree to the actions required at each stage. One of the highest priorities for water users was to avoid going into and out of mandatory restrictions over short periods of time. This operation is costly because it involves transitioning from voluntary conservation mode to mandatory restrictions, with enforcement provisions such as fines.

Each stage also requires that Duke Energy gradually reduce aquatic and recreation flows, down to “critical flows” in Stage 4. Critical flows are flows that minimally sustain aquatic habitat, water quality, and usability of riverine-located intakes. Stage 4, deemed the emergency stage, indicates that inflows into the project reservoirs are so low that the licensee has little to no control over the reservoir levels. At this stage, a few more weeks of drought conditions mean that reservoir levels will begin uncovering water intakes.

Analyzing the plan

With the CHEOPS model, existing and future water use data, performance measures sheet, and LIP structure, Duke Energy and the ad-hoc modeling committee began soliciting potential future operating scenarios from stakeholders in the summer of 2005. These scenarios were screened using the following criteria as a minimum threshold for acceptance:

    – Maintain reservoirs above critical intake levels (critical elevations);
    – Maintain downstream flows above critical flows; and
    – Does not trigger Stage 4 (emergency) of the LIP.

To save time during the model runs, all potential operating scenarios were screened to eliminate operating scenarios that would not meet the above criteria. The three levels used to screen the scenarios were:

    – Level 1: Used typical (but not most extreme) three-year periods of wet-normal-dry hydrology;
    – Level 2: Used 51-year period of hydrologic record (1953-2003). Considered increases in sedimentation and water withdrawal/return volumes over the term of new license; and
    – Level 3: Conducted sensitivity analyses (i.e., find out what would make the preferred scenario fail and then adjust accordingly). Considered increases in sedimentation volumes and water withdrawal/return volumes over the term of the new license and applied the LIP. The 51-year period of hydrologic record was used.

The scenarios subjected to Level 1 analysis consisted of the two modeled extremes of the flexibility spectrum (baseline and run-of-river), plus scenarios that attempted to optimize specific interests relative to recreation, habitat, water use, and property interests. These interests could include, for example, higher in-stream flows for aquatic habitat and higher lake levels for property interests.

Level 2 analyses were designated as Primary A (combination of habitat and property interests), Primary B (a combination of recreation and water use interests), and Primary C (Primary A with 25 percent less downstream flows).

During the model runs to test the various scenarios, a repetitive trial run process was used to develop a mutual gains scenario. The goal was to determine a scenario that maintained reservoir levels and did not trigger river flows at or below those required for large water intake operations in the riverine sections of the project and did not trigger Stage 4 (emergency) of the LIP. The mutual gains scenario was developed by reviewing the output and performance measures sheets for the primary scenarios and incorporating the interest-based effects of potential operating parameters on water quantity, especially during times of low inflow.

The mutual gains scenario was tested for extremes to see how sensitive the scenario was to changes in water use demand (Level 3 analysis). The sensitivity runs included:

    – Extending the duration of the drought of record;
    – Reducing the inflow hydrology file by 5 percent;
    – Considering the effect of “human intervention” during high inflow periods to test days of flooding;
    – Reducing the frequency of LIP Stage 0;
    – Testing the delay of LIP implementation;
    – Greater than projected growth in water withdrawals (increase the projected average growth rate (AGR) from 1.86 percent to 2.25 percent per year);
    – Projecting no growth in water withdrawals; and
    – Projecting no growth in future interbasin transfers of water.

Modeling results from Level 2 analysis revealed the water volume interdependency of the 11 reservoirs and how sensitive the project is to what happens at any single development. The modeling revealed the importance of the useable storage volume in Lake James (39 feet deep) as one of two significant storage reservoirs in the project, with Lake Norman being the other. Lake James and Lake Norman help regulate reservoir elevations and minimum flow releases from the other nine reservoirs during times of low inflow so that recreation, water supply, and habitat flow releases can be met. The modeling also revealed that the proposed habitat flows from Lake Wylie and Lake Wateree (the most downstream reservoir) were key parameters that had a significant effect on the available storage volume throughout the entire project.

By comparison, recreation releases were generally smaller by volume and had less effect on reservoir levels during drought than did the habitat flows. Modeling results from Level 3 revealed the importance of the LIP and the significance to the trigger points for each stage in extending available volume in the system throughout a potential repeat of the most significant drought in the history of the basin.

The final mutual gains scenario that formed the basis of the comprehensive relicensing agreement met the water quantity goals concerning low inflow periods. In addition, the mutual gains scenario considered sensitivities related to projected sediment increases in the reservoirs and water withdrawal/return volumes.

For the entire period of record, the LIP simulation did not enter into Stage 4 for the mutual gains scenario, and no critical reservoir elevation or critical flows to support habitat needs and large water intakes were violated. The performance measures sheet for this scenario showed some performance measures increasing or decreasing more than the minimum increment of significant change compared to baseline, but the mutual gains scenario compared favorably to the other scenarios.

Mr. Bruce may be reached at Duke Energy, 526 South Church Street, Mail Code EC12Y, P.O. Box 1006, Charlotte, NC 28202; (1) 704-382-5239; E-mail:

Ed Bruce, P.E., senior engineer with Duke Energy, was responsible for the operational modeling and water supply study performed during relicensing of the 814-MW Catawba-Wateree Hydroelectric Project.

µ Peer Reviewed

This article has been evaluated and edited in accordance with reviews conducted by two or more professionals who have relevant expertise. These peer reviewers judge manuscripts for technical accuracy, usefulness, and overall importance within the hydroelectric industry.


Previous articleREW Volume 12 Issue 1
Next articleWind Forecasting Key for More Efficient Turbines

No posts to display