Water: Coastal Zone Act Reauthorization Amendments
C. Management Measure for Protection of Surface Water Quality and Instream and Riparian Habitat
Develop and implement a program to manage the operation of dams in coastal areas that includes an assessment of:
- Surface water quality and instream and riparian habitat and potential for improvement and
- Significant nonpoint source pollution problems that result from excessive surface water withdrawals.
This management measure is intended to be applied by States to dam operations that result in the loss of desirable surface water quality, and of desirable instream and riparian habitat. Dams are defined as constructed impoundments which are either:
- 25 feet or more in height and greater than 15 acre-feet in capacity, or
- 6 feet or more in height and greater than 50 acre-feet in capacity.
This measure does not apply to projects that fall under NPDES jurisdiction. This measure also does not apply to the extent that its implementation under State law is precluded under California v. Federal Energy Regulatory Commission, 110 S. Ct. 2024 (1990) (addressing the supersedence of State instream flow requirements by Federal flow requirements set forth in FERC licenses for hydroelectric power plants under the Federal Power Act).
Under the Coastal Zone Act Reauthorization Amendments of 1990, States are subject to a number of requirements as they develop coastal NPS programs in conformity with this measure and will have some flexibility in doing so. The application of management measures by States is described more fully in Coastal Nonpoint Pollution Control Program: Program Development and Approval Guidance, published jointly by the U.S. Environmental Protection Agency (EPA) and the National Oceanic and Atmospheric Administration (NOAA) of the U.S. Department of Commerce.
The purpose of this management measure is to protect the quality of surface waters and aquatic habitat in reservoirs and in the downstream portions of rivers and streams that are influenced by the quality of water contained in the releases (tailwaters) from reservoir impoundments. Impacts from the operation of dams to surface water quality and aquatic and riparian habitat should be assessed and the potential for improvement evaluated. Additionally, new upstream and downstream impacts to surface water quality and aquatic and riparian habitat caused by the implementation of practices should also be considered in the assessment. The overall program approach is to evaluate a set of practices that can be applied individually or in combination to protect and improve surface water quality and aquatic habitat in reservoirs, as well as in areas downstream of dams. Then, the program should implement the most cost-effective operations to protect surface water quality and aquatic and riparian habitat and to improve the water quality and aquatic and riparian habitat where economically feasible.
A variety of approaches have been developed and tested for their effectiveness at improving or maintaining acceptable levels of dissolved oxygen, temperature, phosphorus, and other constituents in reservoirs and tailwaters.
One general method uses pumps, air diffusers, or air lifts to induce circulation and mixing of the oxygen-poor, but cold hypolimnion with the oxygen-rich, but warm epilimnion. The desired result is a more thermally uniform reservoir with increased dissolved oxygen (DO) in the hypolimnion. Reservoir mixing improves water quality both in the reservoir and in tailwaters and helps to maintain the temperatures required by warm-water fisheries.
Another approach to improving water quality in tailwaters is appropriate if trout fisheries are desired downstream. In this approach, air or oxygen is mixed with water passing through the turbines of hydropower dams to increase the concentration of DO. Air or oxygen can be selectively added to impoundment waters entering turbine intakes. Reservoir waters can also be aerated by venting turbines to the atmosphere or by injecting compressed air into the turbine chamber.
A third group of approaches include engineering modifications to the intakes, the spillway, or the tailrace, or the installation of various types of weirs downstream of the dam to improve temperature or DO levels in tailwaters. These practices rely on agitation and turbulence to mix the reservoir releases with atmospheric air in order to increase the concentrations of dissolved oxygen. Selective withdrawal of water from different depths allows dam operators to maintain desired temperatures for fish and other aquatic species in downstream surface waters.
The quality of reservoir releases can also be improved through adjustments in the operational procedures at dams. These include scheduling releases or the duration of shutoff periods, instituting procedures for the maintenance of minimum flows, and making seasonal adjustments in the pool levels and in the timing and variation of the rate of drawdown.
Dam operators such as the Tennessee Valley Authority (TVA) further recognize the need for watershed management as a valuable tool to reduce water quality problems in reservoirs and dam releases. Reducing NPS pollutants coming from watersheds surrounding reservoirs can have a beneficial effect on concentrations of DO and pollutants within a reservoir and its tailwaters.
There is also a need for riparian habitat maintenance and restoration in the areas around the impounded reservoir and downstream from a dam. Reservoir shorelines are important riparian areas, and they need to be managed or restored to realize their many riparian habitat and water quality benefits. Examples of downstream aquatic habitat improvements include maintaining minimum instream flows, providing scouring flows when and where needed, providing alternative spawning areas or fish passage, protecting streambanks from erosion, and maintaining wetlands and riparian areas.
The individual application of any particular technique, such as aeration, change in operational procedure, restoration of an aquatic or riparian habitat, or implementation of a watershed protection best management practice (BMP), will, by itself, probably not improve water quality to an acceptable level within the reservoir impoundment or in tailwaters flowing through downstream areas. The individual practices discussed in this portion of the guidance will usually have to be implemented in some combination in order to raise water quality in the impoundment or in tailwaters to acceptable levels.
One such combination of practices has addressed low DO levels at the Canyon Dam (Guadalupe River, Texas). A combination of turbine venting and a downstream weir was used to increase DO levels to acceptable levels. The concentration of dissolved oxygen in water entering the dam was measured at 0.5 mg/L. After passing through the turbine (but still upstream of the aeration weir), the DO concentration was raised to 3.3 mg/L. The concentration of the same water after passing through the aeration weir was 6.7 mg/L (EPRI, 1990).
Another combination of practices, consisting of a vacuum breaker turbine venting system and a stream flow reregulation weir, has been implemented at Norris Dam (Clinch River, Tennessee). The vacuum breaker aeration system uses hub baffles and appears to be the most successful design (EPRI, 1990). The baffles induce enough air to add from 2 mg/L to 4 mg/L to the discharge, while reducing turbine efficiency less than 0.5 percent. The downstream weir retains part of the discharge from the turbines when they are not in operation to sustain a stream flow of about 200 cubic feet per second (cfs). Prior to these improvements, the tailwaters of the Norris Dam had DO levels below 6 mg/L an average of 131 days per year and DO levels below 3 mg/L an average of 55 days per year. After installation of the turbine venting system and reregulation weir, DO levels were below 6 mg/L only 55 days per year and were above 3 mg/L at all times (TVA, 1988).
Combinations of increased flow, stream aeration, and wasteload reduction (from municipal and industrial sources) were found to be necessary to treat releases from the Fort Patrick Henry Dam (Holston River, Tennessee). An unsteady state flow and water quality model was used to simulate concentrations of dissolved oxygen in the 20-mile downstream reach from Fort Patrick Henry Dam and to explore water quality management alternatives. Several pollution abatement options were considered to identify the most cost-effective alternative. These options included changing wasteloads of the various dischargers, varying the flows from the reservoir, and improving aeration levels in water leaving the reservoir and in areas downstream. The modeling study identified flow regime modifications as more effective in improving DO than wasteload modifications. However, a decision to increase flow from the dam when stream levels are low might result in unacceptable reservoir drawdown in dry years. Although at some projects the increased DO will persist for many miles, improvements that were predicted by aeration of dam releases diminished rapidly at this particular site because they decreased the DO deficit and reduced natural reaeration rates. No wasteload treatments short of total recycle would achieve the 5-mg/L standard under base conditions (Hauser and Ruane, 1985).
Selection of this management measure was based on:
- The availability and demonstrated effectiveness of practices to improve water quality in impoundments and in tailwaters of dams and
- The level of improvement in water quality of impoundments and tailwaters that can be measured from implementation of engineering practices, operational procedures, watershed protection approaches, or aquatic or riparian habitat improvements.
Successful implementation of the management measure will generally involve the following categories of practices undertaken individually or in combination to improve water quality and aquatic and riparian habitat in reservoir impoundments and in tailwaters:
- Artificial destratification and hypolimnetic aeration of reservoirs with deep withdrawal points that do not have multilevel outlets to improve dissolved oxygen levels in the impoundment and to decrease levels of other types of nonpoint source pollutants, such as manganese, iron, hydrogen sulfide, methane, ammonia, and phosphorus in reservoir releases (Cooke and Kennedy, 1989; Henderson and Shields, 1984);
- Aeration of reservoir releases, through turbine venting, injection of air into turbine releases, installation of reregulation weirs, use of selective withdrawal structures, or modification of other turbine start-up or pulsing procedures (Hauser and Ruane, 1985; Henderson and Shields, 1984);
- Providing both minimum flows to enhance the establishment of desirable instream habitat and scouring flows as necessary to maintain instream habitat (Kondolf et al., 1987; Walburg et al., 1981);
- Establishing adequate fish passage or alternative spawning ground and instream habitat for fish species (Andrews, 1988); and
- Improving watershed protection by installing and maintaining BMPs in the drainage area above the dam to remove phosphorus, suspended sediment, and organic matter and otherwise improve the quality of surface waters flowing into the impoundment (Kortmann, 1989).
As discussed more fully at the beginning of this chapter and in Chapter 1, the following practices are described for illustrative purposes only. State programs need not require the implementation of these practices. However, as a practical matter, EPA anticipates that the management measure set forth above generally will be implemented by applying one or more management practices appropriate to the source, location, and climate. The practices set forth below have been found by EPA to be representative of the types of practices that can be applied successfully to achieve the management measure described above.
The systems that have been developed and tested for reservoir aeration rely on atmospheric air, compressed air, or liquid oxygen to increase concentrations of dissolved oxygen in reservoir waters before they pass through the dam. Depending on the method selected, aeration can accomplish thorough mixing throughout the impoundment. However, this practice has not been used at large hydropower reservoirs because of the cost associated with aerating these large-flow reservoirs. Aeration will elevate levels of DO, but also will usually redistribute higher concentrations of algae found in the shallower depths and nutrients that are normally restricted to the deeper waters. It is not always desirable to have waters containing higher levels of algae and nutrients released into portions of the waterway below the dam (Kortmann, 1989). If the principal objective is to improve DO levels only in the reservoir releases and not throughout the entire impoundment, then aeration can be applied selectively to discrete layers of water immediately surrounding the intakes or as water passes through release structures such as hydroelectric turbines.
- a. Pumping and Injection Practices
One method for deployment of circulation pumps is the U-tube design, in which water from deep in the impoundment is pumped to the surface layer. The inducement of artificial circulation through aeration of the impoundment may also provide the opportunity for a "two-story" fishery, reduce internal phosphorus loading, and eliminate problems with iron and manganese in drinking water (Cooke and Kennedy, 1989).
Air injection systems operate in a manner similar to that of pumping systems to mix water from different strata in the impoundment, except that air or pure oxygen is injected into the pumping system (Henderson and Shields, 1984). These kinds of systems are divided into two categories: partial air lift systems and full air lift systems. In the partial air lift system, compressed air is injected at the bottom of the unit; then, the air and water are separated at depth and the air is vented to the surface. In the full air lift system, compressed air is injected at the bottom of the unit (as in the partial air lift system), but the air-water mixture rises to the surface (Figure 6-3). The full air lift design has a higher efficiency than the partial-air lift and has a lesser tendency to elevate dissolved nitrogen levels (Cooke and Kennedy, 1989).
Diffused air systems provide effective transfer of oxygen to water by forcing compressed air through small pores in systems of diffusers to form bubbles (Figure 6-4). One test of a diffuser system in the Delaware River near Philadelphia, Pennsylvania, in 1969-1970 demonstrated the efficiency of this practice. Coarse-bubble diffusers were deployed at depths ranging from 13 to 38 feet. Depending on the depth of deployment, the oxygen transfer efficiency varied from 1 to 12 percent. When compared with other systems discussed below, this efficiency rate is rather low. But the results of this particular test determined that river aeration was more economical than advanced wastewater treatment as a strategy for improving the levels of DO in the river (EPRI, 1990).
Mechanical agitation systems operate by pumping water from the reservoir into a splash basin on shore, where it is aerated and then returned to the hypolimnion. Although these types of systems are comparatively inefficient, they have been used successfully (Wilhelms and Smith, 1981).
Localized mixing is a practice to improve releases of thermally stratified reservoirs by destratifying the reservoir in the immediate vicinity of the outlet structure. This practice differs from the practice of artificial destratification, where mixing is designed to destratify all or most of the reservoir volume (Holland, 1984). Localized mixing is provided by forcing a jet of high-quality surface water downward into the hypolimnion. Pumps used to create the jet generally fall into two categories, axial flow propellers and direct drive mixers (Price, 1989). Axial flow pumps usually have a large-diameter propeller (6 to 15 feet) that produces a high-discharge, low-velocity jet. Direct drive mixers have small propellers (1 to 2 feet) that rotate at high speeds and produce a high-velocity jet. The axial flow pumps are suitable for shallow reservoirs because they can force large quantities of water down to shallow depths. The high-momentum jets produced by direct drive mixers are necessary to penetrate deeper reservoirs (Price, 1989).
Water pumps have been used to move surface water containing higher concentrations of DO downward to mix with deeper waters as the two strata are entering the turbine. Aspirating surface aerators deployed in Lake Texoma (Texas/Oklahoma border) raised the levels of DO in the tailwaters from concentrations of 1.8 mg/L (without aerators) to 2.0 mg/L (with one 5-hp aeration unit in operation) and to 2.6 mg/L (with three 5-hp units operating).
A test of large-diameter axial-flow surface water pumps at Bagnell Dam (Lake of the Ozarks, Missouri) increased DO levels in the reservoir releases from 1.3 mg/L to 3.6 mg/L, before maintenance problems caused a discontinuance of use of the pumps (EPRI, 1990).
Small-diameter surface pumps, operated at the J. Percy Priest Dam (Tennessee), increased the DO levels in the tailwaters to 4.0 mg/L from a background level of 2.7 mg/L (EPRI, 1990).
Oxygen injection systems use pure oxygen to increase levels of dissolved oxygen in reservoirs. One type of design, termed side stream pumping, carries water from the impoundment onto the shore and through a piping system into which pure oxygen is injected. After passing through this system, the water is returned to the impoundment. Another type of system, which pumps gaseous oxygen into the hypolimnion through diffusers, has effectively improved DO levels in the reservoir behind the Richard B. Russell Dam (Savannah River, on the Georgia-South
Carolina border). The system is operated 1 mile upstream of the dam, with occasional supplemental injection of oxygen at the dam face when DO levels are especially low. The system has successfully maintained DO levels above 6 mg/L in the releases, with an average oxygen transfer efficiency of 75 percent (EPRI, 1990; Gallagher and Mauldin, 1987).
The TVA has been testing the use of pure oxygen at the Douglas Dam (French Broad River, Tennessee) since 1988 (TVA, 1988). The absorption efficiencies measured in the downstream tailwaters range from 30 to 50 percent when the diffusers are arranged in a loose arc around the intakes. When the diffusers are placed tightly around the intakes, the efficiency range improves to 72 to 76 percent.
In another test at facilities operated by the Tennessee Valley Authority, diffusers were deployed to inject high-purity oxygen near the bottom of the 70-foot-deep reservoir at Fort Patrick Henry Dam (Holston River, Tennessee) near one of the turbine intakes. Levels of DO in the tailwaters increased from near 0 mg/L to 4 mg/L as a result of operation of this aeration system. Unfortunately, the operation costs of this kind of system were determined to be relatively high (Harshbarger, 1987). However, these results were very site-specific and every site needs to be evaluated for the best mix of solutions.
- b. Turbine Venting
Turbine venting is the practice of injecting air into water as it passes through a turbine. If vents are provided inside the turbine chamber, the turbine will aspirate air from the atmosphere and mix it with water passing through the turbine as part of its normal operation. In early designs, the turbine was vented through existing openings, such as the draft tube opening or the vacuum breaker valve in the turbine assembly. Air forced by compressors into the draft tube opening enriched reservoir waters with little detectable DO to concentrations of 3 to 4 mg/L. Overriding the automatic closure of the vacuum breaker valve (at high turbine discharges) increased DO by only 2 mg/L (Harshbarger, 1987).
Turbine venting makes use of the low-pressure region just below the turbine wheel to aspirate air into the discharges (Wilhelms, 1984). Autoventing turbines are constructed with hub baffles, or deflector plates placed on the turbine hub upstream of the vent holes to enhance the low-pressure zone in the vicinity of the vent and thereby increase the amount of air aspirated through the venting system (Figure 6-5). Turbine efficiency relates to the amount of energy output from a turbine per unit of water passing through the turbine. Efficiency decreases as less power is produced for the same volume of water. In systems where the water is aerated before passing through the turbine, part of the water volume is displaced by the air, thus leading to decreased efficiency. Hub baffles have also been added to autoventing turbines at the Norris Dam to further improve the DO levels in the turbine releases (Jones and March, 1991).
Recent developments in autoventing turbine technology show that it may be possible to aspirate air with no resulting decrease in turbine efficiency. In one test of an autoventing turbine at the Norris Dam (Clinch River, Tennessee), the turbine efficiency increased by 1.8 percent (March et al., 1991; Waldrop, 1992). Technologies like autoventing turbines are very site-specific and outcomes will vary considerably. Achievement of desired DO levels at specific projects may require evaluation of several different technologies.
In addition to the pumping and injection systems for reservoir aeration discussed in the preceding section, another set of systems can accomplish the aeration of water as it passes through the dam or through the portion of the waterway immediately downstream from the dam. The systems in this category rely on agitation and turbulence to mix the reservoir releases with atmospheric air in order to increase the concentrations of dissolved oxygen. Another approach involves the increased use of spillways, which release surface water to prevent it from overtopping the dam. The third approach is to install barriers called weirs in the downstream areas. Weirs designed to allow water to overtop them can increase DO through surface agitation and increased surface area contact. Some systems create supersaturation of dissolved gases and may require additional modifications to prevent supersaturation.
Two factors should be considered when evaluating the suitability of hydraulic structures such as spillways and weirs for their application in raising the DO concentration in waterways:
- Most of the measurements of DO increases associated with hydraulic structures have been collected at low-head facilities. The effectiveness of these devices may be limited as the level of discharge increases (Wilhelms, 1988).
- The hydraulic functioning of these types of structures should be carefully considered since undesirable flow conditions may occur in some instances (Wilhelms, 1988).
- a. Gated Conduits
Gated conduits are hydraulic structures that divert the flow of water under the dam. They are designed to create turbulent mixing to enhance the rest of the oxygen transfer. Gates are used to control the cross-sectional area of flow. Gated conduits have been extensively analyzed for their performance and effectiveness (Wilhelms and Smith, 1981), although the available data are mostly from high-head projects (Wilhelms, 1988). In modeling studies, gated conduit structures have been found to achieve 90 percent aeration and a minimum DO standard of 5 mg/L (Wilhelms and Smith, 1981).
- b. Spillways
The U.S. Army Corps of Engineers has studied the performance of spillways and overflow weirs at its facilities to determine the importance of these structures in improving DO levels. Increases in DO concentration of about 2.5 mg/L have been measured at the overflow weir of the Jonesville Lock and Dam (Ouachita River, Louisiana) (Wilhelms, 1988). Increases in DO concentrations of 3 mg/L have been measured at the overflow weir of the Columbia Lock and Dam (Ouachita River, Louisiana). Passage of water through the combinations of spillways and overflow weirs at these two facilities resulted in DO saturation levels of 85 to 95 percent in downstream waters (Wilhelms, 1988).
- c. Spillway Modifications
At the Tellico Dam (Little Tennessee River, Tennessee), a siphon/underwater barrier dam was installed to improve DO and temperature conditions in the releases. The installed siphon draws about 8 cfs of cool water from the reservoir over the spillway into the Little Tennessee River. During the summer, the water forms a pool behind a 6-ft high underwater barrier dam and creates the temperature and oxygen concentrations needed by striped bass. The fish attracted to the pool provide a desirable sport fishery for the community (TVA, 1988).
The operation of some types of hydraulic structures has been tied to problems stemming from the supersaturation of some types of gases. An unexpected fish kill occurred in spring 1978 due to supersaturation of nitrogen gas in the Lake of the Ozarks (Missouri) within 5 miles of Truman Dam, caused by water plunging over the spillway and entraining air. The vertical drop between the spillway crest and the tailwaters was only 5 feet. The maximum saturation was 143 percent. In this case, the spillway was modified by cutting a notch to prevent water from plunging directly into the stilling basin (ASCE, 1986). At dams along the Columbia and Snake Rivers of the western United States, spillway deflectors have been found to be the most effective means for reducing nitrogen supersaturation (Bonneville Power Administration, 1991). The deflectors are designed to direct flows horizontally into the stilling basin to prevent deep plunging and air entrainment (ASCE, 1986).
Spill at hydroelectric dams is routinely required during periods of high runoff when the river discharge exceeds what can be passed through the powerhouse turbines. The Columbia River of Washington State has a series of 11 dams beginning with the Grand Coulee and ending with Bonneville. The Snake River also has four dams. If all of these dams were spilling simultaneously, the entire river would become and remain highly saturated with nitrogen gas since the water would pick up gas at each successive spilling project. The Corps of Engineers has proposed several practices for solving the gas supersaturation problem. These include (1) passing more headwater storage through turbines, installing new fish bypass structures, and installing additional power units to reduce the need for spill; (2) incorporating "flip-lip" deflectors in spillway-stilling basins (Figure 6-6), transferring power generation to high-dissolved-gas-producing dams, and altering spill patterns at individual dams to minimize nitrogen mass entrainment; and (3) collecting and transporting juvenile salmonids around affected river reaches. Only a few of these practices have been implemented (Tanovan, 1987).
- d. Reregulation Weir
Reregulation weirs have been constructed from stone, wood, and aggregate. In addition to increasing the levels of DO in the tailwaters, reregulation weirs result in a more constant rate of flow farther downstream during periods when turbines are not in operation. A reregulation weir constructed downstream of the Canyon Dam (Guadalupe River, Texas) increased DO levels in waters leaving the turbine from 3.3 mg/L to 6.7 mg/L (EPRI, 1990).
The U.S. Army Corps of Engineers Waterways Experiment Station (Wilhelms, 1988) has compared the effectiveness with which various hydraulic structures accomplished the reaeration of reservoir releases. The study concluded that, whenever operationally feasible, more discharge should be passed over weirs to improve DO concentrations in releases. Although additional field tests are planned, current results indicate that overflow weirs aerate releases more effectively than low-sill spillways (Wilhelms, 1988).
- e. Labyrinth Weir
Labyrinth weirs have extended crest length and are usually W-shaped. These weirs spread the flow out to prevent dangerous undertows in the plunge pool. A labyrinth weir at South Holston Dam (Figure 6-7) was constructed for the dual purpose of providing minimum flows and improving DO in reservoir releases. The weir aerates to up to 60 percent of the oxygen deficit. For instance, projected performance at the end of the summer is an increase in the DO from 3 mg/L to 7 mg/L (or an increase of 4 mg/L) (Gary Hauser, TVA, personal communication, 1992). Actual increases in the DO will depend on the temperature and the level of DO in the incoming water.
The quality of reservoir releases can be improved through adjustments in the operational procedures at dams. These include scheduling of releases or of the duration of shutoff periods, instituting procedures for the maintenance of minimum flows, making seasonal adjustments in the pool levels or in the timing and variation of the rate of drawdown, selecting the turbine unit that most increases DO (often increasing the DO levels by 1 mg/L), and operating more units simultaneously (often increasing DO levels by about 2 mg/L). The magnitude and duration of reservoir releases also should be timed and scheduled so that the salinity regime in coastal waters is not substantially altered from historical patterns.
- a. Selective Withdrawal
Temperature control in reservoir releases depends on the volume of water storage in the reservoir, the timing of the release relative to storage time, and the level from which the water is withdrawn. Dams capable of selectively releasing waters of different temperatures can provide cooler or warmer water temperature downstream at times that are critical for other instream resources, such as during periods of fish spawning and development of fry (Fontane et al., 1981; Hansen and Crumrine, 1991). Stratified reservoirs are operated to meet downstream temperature objectives such as to enhance a cold-water or warm-water fishery or to maintain preproject stream temperature conditions. Release temperature may also be important for irrigation (Fontane et al., 1981).
Multilevel intake devices in storage reservoirs allow selective withdrawal of water based on temperature and DO levels. These devices minimize the withdrawal of surface water high in blue-green algae, or of deep water enriched in iron and manganese. Care should be taken in the design of these systems not to position the multilevel intakes too far apart because this will increase the difficulty with which withdrawals can be controlled, making the discharge of poor-quality hypolimnetic water more likely (Howington, 1990; Johnson and LaBounty, 1988; Smith et al., 1987).
- b. Turbine Operation
Implementation of changes in the turbine start-up procedures can also enlarge the zone of withdrawal to include more of the epilimnetic waters in the downstream releases. Monitoring of the releases at the Walter F. George lock and dam (Chattahoochee River, Georgia), showed levels of DO declined sharply at the start-up of hydropower production. The severity and duration of the DO drop could be reduced by starting up all the generator units within a minute of each other (Findley and Day, 1987).
A useful tool for evaluating the effects of operational procedures on the quality of tailwaters is computer modeling. For instance, computer models can describe the vertical withdrawal zone that would be expected under different scenarios of turbine operation (Smith et al., 1987). Zimmerman and Dortch (1989) modeled release operations for a series of dams on a Georgia River and found that procedures that were maintaining cool temperatures in summer were causing undesirable decreases in DO and increases in dissolved iron in autumn. The suggested solution was a seasonal release plan that is flexible, depending on variations in the in-pool water quality and predicted local weather conditions. Care should be taken with this sort of approach to accommodate the needs of both the fishery resource and reservoir recreationalists, particularly in late summer.
Modeling has also been undertaken for a variety of TVA and Corps of Engineers facilities to evaluate the downstream impacts on DO and temperature that would result from changes in several operational procedures, including (Hauser et al., 1990a, 1990b; Higgins and Kim, 1982; Nestler et al., 1986b):
- Maintenance of minimum flows;
- Timing and duration of shutoff periods;
- Seasonal adjustments to the pool levels; and
- Timing and variation of the rate of drawdown.
Most nonpoint source pollution problems in reservoirs and dam tailwaters frequently result from sources in the contributing watershed (e.g., sediment, nutrients, metals, and toxics). Management of pollution sources from a watershed has been found to be a cost-effective solution for improving reservoir and dam tailwater water quality (TVA, 1988). Practices for watershed management include land use planning, erosion control, ground-water protection, mine reclamation, NPS screening and identification, animal waste control, and failing septic tank control (TVA, 1988).
Another general watershed management practice involves the evaluation of the total watershed and the use of point/NPS source trading. Simply put, this practice involves the evaluation of the sources of pollution in a watershed and determination of the most cost-effective combination of practices to reduce pollution among the various point and nonpoint sources. Podar and others (1985) present an excellent example of point/NPS source trading as applied to the Holston River near Kingsport, Tennessee. Bender and others (1991) used modeling to evaluate the cost-effectiveness of various point/NPS source trading strategies for the Boone Reservoir in the upper Tennessee River Valley.
- a. Land Use Planning
Land use plans that establish guidelines for permissible uses of land within a watershed serve as a guide for reservoir management programs addressing NPS pollution (TVA, 1988). Watershed land use plans identify suitable uses for land surrounding a reservoir, establish sites for economic development and natural resource management activities, and facilitate improved land management (TVA, 1988). Land use plans must be flexible documents that account for the needs of the landowners, State and local land use goals, the characteristics of the land and its ability to support various uses, and the control of NPS pollution (TVA, 1988). The watershed planning section of Chapter 4 contains additional information on land use planning.
- b. Nonpoint Source Screening and Identification
The analysis and interpretation of stereoscopic color infrared aerial photographs can be used to find and map specific areas of concern where a high probability of NPS pollution exists from septic tank systems, animal wastes, soil erosion, and other similar types of NPS pollution (TVA, 1988). TVA has used this technique to survey about 25 percent of the Tennessee Valley to identify sources of nonpoint pollution in a period of less than 5 years at a cost of a few cents per acre (TVA, 1988).
- c. Soil Erosion Control
Soil erosion has been determined to be the major source of suspended solids, nutrients, organic wastes, pesticides, and sediment that combined form the most problematic form of NPS pollution (TVA, 1988). Chapter 4 in this guidance contains an extensive selection of practices aimed at preventing soil erosion and controlling sediment from reaching surface waters in runoff.
- d. Ground-Water Protection
Proper protection and management of ground-water resources primarily depends on the effective control of NPS pollution, particularly in ground-water recharge areas. Polluted ground water has the potential to contribute to surface-water pollution problems in reservoirs. Ground-water protection can be achieved only through public awareness of the problems associated with ground-water pollution and the potential of various activities to contaminate ground water. Identifying the ground-water resources in a watershed and developing a plan for protection of these resources are critical in establishing a good ground-water protection program. TVA (1988) has found that an extensive public outreach program is instrumental in the development of an effective ground-water protection program and in eventual protection of the resource.
- e. Mine Reclamation
Abandoned mines have the potential to contribute significant sediment, metals, acidified water, and other pollutants to reservoirs (TVA, 1988). Old mines need to be located and reclaimed to reduce the NPS pollutants emanating from them. Revegetation is a cost-effective method of reclaiming denuded strip-mined lands, and agencies such as the Soil Conservation Service can provide technical insight for revegetation practices.
- f. Animal Waste Control
A major contributor to reservoir pollution in some watersheds is wastes from animal confinement facilities. TVA (1988) estimated that in the Tennessee Valley, farms produced about six times the organic wastes of the population of the valley. A cooperative program was established to address the animal waste problem in the Tennessee Valley. The results of demonstration facilities in the Tennessee Valley reduced NPS pollution from animal wastes by 25,000 tons in the Duck River basin. The program also had the benefit of reducing the additional input of 1,400 tons of nitrogen and 200 tons of phosphorus to farm fields (TVA, 1988). Refer to Chapter 2 of this guidance for additional information on animal waste control practices.
- g. Failing Septic Systems
Failing septic tank or onsite sewage disposal systems (OSDS) are another source of NPS pollution in reservoirs. TVA has found septic tank failures to be a problem in some of its reservoirs and has identified them through an aerial survey (TVA, 1988). Additional information on OSDS practices can be found in Chapter 4.
Studies like the one undertaken by the U.S. Department of the Interior (USDOI, 1988) on the Glen Canyon Dam (Colorado River, Colorado) illustrate the potential for disruption to downstream aquatic and riparian habitat resulting from the operation of dams.
Several options are available for the restoration or maintenance of aquatic and riparian habitat in the area of a reservoir impoundment or in portions of the waterway downstream from a dam. One set of practices is designed to augment existing flows that result from normal operation of the dam. These include operation of the facility to produce flushing flows, minimum flows, or turbine pulsing. Another approach to producing minimum flows is to install small turbines that operate continuously. Installation of reregulation weirs in the waterway downstream from the dam can also achieve minimum flows. Finally, riparian improvements are discussed for their importance and effectiveness in restoring or maintaining aquatic and riparian habitat in portions of the waterway affected by the location and operation of a dam.
- a. Flow Augmentation
Operational procedures such as flow regulation, flood releases, or fluctuating flow releases all have a detrimental impact on downstream aquatic and riparian habitat. Confounding the problem of aquatic and riparian habitat restoration is necessary for a balance of operational procedures to address the needs of downstream aquatic and riparian habitat with the requirements of dam operation. There are often legal and jurisdictional requirements for an operational procedure at a particular dam that should be considered (USDOI, 1988).
A flushing flow is a high-magnitude, short-duration release for the purpose of maintaining channel capacity and the quality of instream habitat by scouring the accumulation of fine-grained sediments from the streambed. For example, at Owens River in the Eastern Sierra Nevadas, California, a study found that wild salmonids prefer to deposit their eggs in streambed gravel free of fine sediments (Kondolf et al., 1987). Availability of suitable instream habitat is a key factor limiting spawning success. Flushing flows wash away the sediments without removing the gravel. Flushing flows also prevent the encroachment of riparian vegetation. According to a study of the Trinity River Drainage Basin in northwestern California (Nelson et al., 1987), remedial and maintenance flushing flows suppress riparian vegetation and maintain the stream channel dimensions necessary to provide instream habitat in addition to preventing large accumulations of sediment in river deltas. Recommendations for the use of flushing flows as part of an overall instream management program are becoming more common in areas downstream of water development projects in the western United States. For instance, Wesche and others (1987) used a sediment transport input-output model to determine the required flushing regimen for removing fine-grained sediments from portions of the Little Snake River that served as instream habitat for Colorado cutthroat trout. The flushing flows reduced the overall mass of sediment covering the channel bottom and removed the finer grained material, thereby increasing the size of the residual sediment forming the bottom streambed deposits.
However, it is important to keep in mind that flushing flows are not recommended in all cases. Flushing flows of a large magnitude may cause flooding in the old floodplain or depletion of gravel below the dam. Flushing flows are more efficient and predictable for small, shallow, high-velocity mountain streams unaltered by dams, diversions, or intensive land use. Routine maintenance generally requires a combination of practices including high flows coupled with sediment dams or channel dredging, rather than simply relying on flushing or scouring flows (Nelson et al., 1988).
Minimum flows are needed to keep streambeds wetted to an acceptable depth to support desired fish and wildlife. Since wetlands and riparian areas are linked hydrologically to adjoining streams, instream flows should be sufficient to maintain wetland or riparian habitat and function. Flushing and scouring flows may also be necessary to clean some streambeds and to provide the proper substrate for aquatic species.
In the design, construction, and operation of dams, the minimum flow requirements to support aquatic organisms and other water-dependent wildlife in downstream areas should be addressed. Minimum flow requirements are typically determined to protect or enhance one or a few harvestable species of fish (USDOI-FWS, 1976). Other fish, aquatic organisms, and riparian wildlife are usually assumed to be protected by these flows. For instance, when minimum flows at the Conowingo Dam (Susquehanna River, Maryland-Pennsylvania border) were increased from essentially zero to 5,000 cfs, up to a 100-fold increase was noted in the abundance of macroinvertebrates (USDOE, 1991). When minimum flows were increased from 1.0 cfs to 5.5 cfs at the Rob Roy Dam (Douglas Creek, Wyoming), there was a four- to six-fold increase in the number of brown trout (USDOE, 1991).
Flows at Rush Creek on the Eastern slope of the Sierra Nevadas in California have averaged about 50 percent of their prediversion levels (Stromberg and Patten, 1990). Since the construction of the Grant Lake Reservoir, the influence of flow rates and volumes on the growth of riparian trees has been studied. Stromberg and Patten (1990) found that a strong relationship exists between growth rates of riparian tree species and annual and prior-year flow volumes. If the level of growth needed to maintain populations is known, the relationship between growth and flow can be used to determine the instream flow needs of riparian vegetation. Instream models for Rush Creek suggest that requirements of riparian vegetation may be greater than requirements for fisheries.
Seasonal discharge limits can be established to prevent excessive, damaging rates of flow release. Limits can also be placed on the rate of change of flow and on the stage of the river (as measured at a point downstream of the dam facility) to further protect against damage to instream and riparian habitat.
Several options exist for establishing minimum flows in the tailwaters below dams. As indicated in the case studies described below, the selection of any particular technique as the most cost-effective depends on several factors including adequate performance to achieve the desired instream and riparian habitat characteristic, compatibility with other requirements for operation of the hydropower facility, availability of materials, and cost.
Sluicing is the practice of releasing water through the sluice gate rather than through the turbines. For portions of the waterway immediately below the dam, the steady release of water by sluicing provides minimum flows with the least amount of water expenditure. At some facilities, this practice may dictate that modifications be made to the existing sluice outlets to maintain continuous low releases.
Continuous low-level sluice releases at Eufala Lake and Fort Gibson Lake (Oklahoma) improved DO levels in tailwaters downstream of these two dams such that fish mortalities, which had been experienced in the tailwaters below these two dams prior to initiating this practice, no longer occurred (USDOE, 1991).
Turbine pulsing is a practice involving the release of water through the turbines at regular intervals to improve minimum flows. In the absence of turbine pulsing, water is released from large hydropower dams only when the turbines are operating, which is typically when the demand for power is high.
A study undertaken at the Douglas Dam (French Broad River, Tennessee) suggests some of the site-specific factors that should be considered when evaluating the advantages of practices such as turbine pulsing, sluicing, or other alternatives for providing minimum flows and improving DO levels in reservoir releases. Three options (turbine pulsing, sluicing, and operation of surface water pumps and diffusers) were evaluated for their effectiveness, advantages, and disadvantages in providing minimum flows and aeration of reservoir releases. Computer modeling indicated that either turbine pulsing or sluicing could improve DO concentrations in releases by levels ranging from 0.7 to 1.5 mg/L. (Based on studies cited in a previous section of this chapter, this is slightly below the level of improvement that might be expected from operation of a diffuser system for aeration.) A trade-off can also be expected at this facility between water saved by frequent short-release pulses and the higher maintenance costs due to setting turbines on and off frequently (Hauser et al., 1989). Hauser (1989) found that schemes of turbine pulsing ranging from 15-minute intervals to 60-minute intervals every 2 to 6 hours were found to provide fairly stable flow regimes after the first 3 to 8 miles downstream at several TVA projects. However, at points farther downstream, less overall flow would be produced by sluicing than by pulsing. Turbine pulsing may also cause waters to rise rapidly, which could endanger people wading or swimming in the tailwaters downstream of the dam (TVA, 1990).
A reregulation weir is one alternative that has been used to establish minimum flows for preservation of instream habitat. This device is installed in the streambed a short distance below a dam and captures hydropower releases. Flows through the weir can be regulated to produce the desired conditions of water level and flow velocities that are best for instream habitat. As discussed previously in this chapter, reregulation weirs can also be used in some circumstances to improve levels of dissolved oxygen in reservoir releases.
The installation of such an instream structure requires some degree of planning and design since the performance of the weir will affect both the downstream water surface elevation and the velocity of the discharge. These relationships have been investigated for the Buford Dam (Chattahoochee River, Georgia), where computer simulations of a proposed reregulation weir indicated that a discharge of 500 cfs created the best instream habitat conditions for juvenile brown trout. Instream habitat for adult brook trout, adult brown trout, and adult rainbow trout was most desirable at discharges in the vicinity of 1,000 to 2,000 cfs (Nestler et al., 1986a).
A reregulation weir was also found to be the most cost-effective alternative for providing a 90-cfs minimum flow below the Holston Dam (South Fork Holston River, Tennessee) for maintenance of instream trout habitat (Adams and Hauser, 1990). The weir was investigated as one alternative for establishing minimum flows, along with turbine pulsing and installation of a small generating unit in the existing tailrace that would operate at all times when the existing unit was not operating. The three alternatives were assessed for their effects on river hydraulics and on operation of the hydropower facility.
Small turbines are another alternative that has been evaluated for establishing minimum flows. Small turbines are capable of providing continuous generation of power using small flows, as opposed to operating large turbine units with the resultant high flows. In a study of alternatives for providing minimum flows at the Tims Ford Dam (Elk River, Tennessee), small turbines were found to represent the most attractive alternative from a cost-benefit perspective. The other alternatives evaluated included continuous operation of a sluice gate at the dam, pulsing of the existing turbines, and construction of an instream rock gabion regulating weir downstream of the dam (TVA, 1985).
- b. Riparian Improvements
Riparian improvements are another strategy that can be used to restore or maintain aquatic and riparian habitat around reservoir impoundments or along the waterways downstream from dams. In fact, Johnson and LaBounty (1988) found that riparian improvements were more effective than flow augmentation for protection of instream habitat. In the Salmon River (Idaho), a variety of instream and riparian habitat improvements have been recommended to improve the indigenous stocks of chinook salmon. These include reducing sediment loading in the watershed, improving riparian vegetation, eliminating barriers to fish migration (see sections discussing this practice below), and providing greater instream and riparian habitat diversity (Andrews, 1988).
- c. Aquatic Plant Management
One study of the Cherokee Reservoir (Holston River, Tennessee) reveals the potential importance of watershed protection practices for the improvement of water quality in the reservoir (Hauser et al., 1987). An improved two-dimensional model of reservoir water quality was used to investigate the advantages and disadvantages of several practices for improving temperature and DO levels in the reservoir.
Migrating fish populations may suffer losses when passing through the turbines of hydroelectric dams unless these facilities have been equipped with special design features to accommodate fish passage. The effect of dams and other hydraulic structures on migrating fish has been studied since the early 1950s in an effort to develop systems or identify operating conditions that would minimize mortality rates. Despite extensive research, no single device or system has received regulatory agency approval for general use (Stone and Webster, 1986).
The safe passage of fish either upstream or downstream through a dam requires a balance between operation of the facility for its intended uses and implementation of practices that will ensure safe passage of fish. Rochester and others (1984) provide an excellent discussion of some of the economic and engineering considerations necessary to address the problems associated with the safe passage of fish.
Available fish-protection systems for hydropower facilities fall into one of four categories based on their mode of action (Stone and Webster, 1986): behavioral barriers, physical barriers, collection systems, and diversion systems. These are discussed in separate sections below, along with four additional practices that have been successfully used to maintain fish passage: spill and water budgets, fish ladders, transference of fish runs, and constructed spawning beds.
- a. Behavioral Barriers
Behavioral barriers use fish responses to external stimuli to keep fish away from the intakes or to attract them to a bypass. Since fish behavior is notably variable both within and between species, behavioral barriers cannot be expected to prevent all fish from entering hydropower intakes. Environmental conditions such as high turbidity levels can obscure some behavioral barriers such as lighting systems and curtains. Competing behaviors such as feeding or predator avoidance can also be a factor influencing the effectiveness of behavioral barriers at a particular time.
Electric screens, bubble and chain curtains, light, sound, and water jets have been evaluated in laboratory or field studies, with mixed results. The results with system tests of strobe lights, poppers, and hybrid systems are the most promising, but these systems are still in need of further testing (Mattice, 1990). Experiences with some kinds of behavioral barrier systems are described more fully in the following paragraphs.
Electrical screens are intended to produce an avoidance response in fish. This type of fish-protection system is designed to keep fish away from structures or to guide them into bypass areas for removal. Fish seem to respond to the electrical stimulus best when water velocities are low. Tests of an electrical guidance system at the Chandler Canal diversion (Yakima River, Washington) showed the efficiency ranged from 70 to 84 percent for velocities of less than 1 ft/sec. Efficiencies decreased to less than 50 percent when water velocities were higher than 2 ft/sec (Pugh et al., 1971). The success of this type of system may also be species-specific and size-specific. An electrical field strength suitable to deter small fish may result in injury or death to large fish, since total fish body voltage is directly proportional to fish body length (Stone and Webster, 1986). This type of system requires constant maintenance of the electrodes and the associated underwater hardware in order to maintain effectiveness. Surface water quality, in particular, can affect the life and performance of the electrodes.
Air bubble curtains are created by pumping air through a diffuser to create a continuous, dense curtain of bubbles, which can cause an avoidance response in fish. Many factors affect the response of fish to air bubble curtains, including temperature, turbidity, light intensity, water velocity, and orientation in the channel. Bubbler systems should be constructed from materials that are resistant to corrosion and rusting. Installation of bubbler systems needs to consider adequate positioning of the diffuser away from areas where siltation could clog the air ducts.
Hanging chains are used to provide a physical, visible obstacle that fish will avoid. Hanging chains are both species-specific and lifestage-specific. Their efficiency is affected by such variables as instream flow velocity, turbidity, and illumination levels. Debris can limit the performance of hanging chains; in particular, buildup of debris can deflect the chains into a nonuniform pattern and disrupt hydraulic flow patterns.
Strobe lights repel fish by producing an avoidance response. A strobe light system at Saunders Generating Station in Ontario was rated 65 to 95 percent effective at repelling or diverting eels (Stone and Webster, 1986). Turbidity levels in the water can affect strobe light efficiency. The intensity and duration of the flash can also affect the response of the fish; for instance, an increase in flash duration has been associated with less avoidance. Strobe lights also have the potential for far-field fish attraction, since they can appear to fish as a constant light source due to light attenuation over a long distance (Stone and Webster, 1986).
Mercury lights are used to attract the fish as opposed to repelling them. Studies of mercury lights suggest their effectiveness is species-specific; alewives were attracted to a zone of filtered mercury light, whereas coho salmon and rainbow trout displayed no attraction to mercury light (Stone and Webster, 1986). Insufficient data are available to determine whether mercury lights are lifestage-specific. The device shows promise, but more research is being conducted to determine factors that affect performance and efficiency.
Underwater sound broadcast at different frequencies and amplitudes has been shown to be effective in attracting or repelling fish, although the results of field tests are not consistent. Fish have been attracted, repelled, or guided by the sound, and no conclusive response to sound has been observed. Not all fish possess the ability to perceive sound or localized acoustical sources (Harris and Van Bergeijk, 1962). Fish also frequently seem to become habituated to the sound source.
Poppers are pneumatic sound generators that create a high-energy acoustic output to repel fish. Poppers have been shown to be effective in repelling warm-water fish from water intakes. Laboratory and field studies conducted in California indicate good avoidance for several freshwater species such as alewives, perch, and smelt (Stone and Webster, 1986), but salmonids do not seem to be effectively repelled by this device (Stone and Webster, 1986). One important maintenance consideration is that internal "O" rings positioned between the air chambers have been found to wear out quickly. Other considerations are air entrainment in water inlets and vibration of structures associated with the inlets.
Water jet curtains can be used to create hydraulic conditions that will repel fish. Effectiveness is influenced by the angle at which the water is jetted. Although effectiveness averages 75 percent in repelling fish (Stone and Webster, 1986), not enough is known to determine what variables affect the performance of water jet curtains. Important concerns would be clogging of the jet nozzles by debris or rust and the acceptable range of flow conditions.
Hybrid barriers, or combinations of different barriers, can enhance the effectiveness of individual behavioral barriers. A chain net barrier combined with strobe lights has been shown in laboratory studies to be 90 percent effective at repelling fish. Combinations of rope-net and chain-rope barriers have also been tested with good results. Barriers with horizontal components as well as vertical components are more effective than those with vertical components alone. Barriers having elements with a large diameter are more effective than those with a small diameter, and thicker barriers are more effective than thinner barriers. Therefore, diameter and spacing of the barriers are factors influencing performance (Stone and Webster, 1986). With hanging chains, illumination appears to be a necessary factor to ensure effectiveness. Their effectiveness was increased with the use of strobe lights (Stone and Webster, 1986). Effectiveness also increased when strobe lights were added to air bubble curtains and poppers (Stone and Webster, 1986).
- b. Physical Barriers
Physical barriers such as barrier nets and stationary screens can prevent the entry of fish and other aquatic organisms into the intakes at a generating facility. However, they should not be regarded as having much potential for application to promote fish bypass at hydroelectric dams for two reasons. First, the size of the mesh and the labor-intensive maintenance required to remove water-borne trash lower the feasibility of their use. Second, these barriers do little to assist fish in bypassing dams during migration (Mattice, 1990).
- c. Fish Collection Systems
Collection systems involve capture of fish by screening and/or netting followed by transport by truck or barge to a downstream location (Figure 6-8). Since the late 1970s, the Corps of Engineers has successfully implemented a program that takes juvenile salmon from the uppermost dams in the Columbia River system (Pacific Northwest) and transports them by barge or truck to below the last dam. The program improves the travel time of fish through the river system, reduces most of the exposure to reservoir predators, and eliminates the mortality associated with passing through a series of turbines (van der Borg and Ferguson, 1989). Survivability rates for the collected fish are in excess of 95 percent, as opposed to survival rates of about 60 percent had the fish remained in the river system and passed through the dams (Dodge, 1989). However, the collection efficiency can range from 70 percent to as low as 30 percent. At the McNary Dam on the Columbia River, spill budgets are implemented (see below) when the collection rate achieves less than 70 percent efficiency (Dodge, 1989).
- d. Fish Diversion Systems
Diversion systems lead or force fish to bypasses that transport them to the natural waterbody below the dam (USEPA, 1979). Physical diversion structures deployed at dams include traveling screens, louvers, angled screens, drum screens, and inclined plane screens. Most of these systems have been effectively deployed at specific hydropower facilities. However, a sufficient range of performance data is not yet available for categorizing the efficiency of specific designs in a particular set of site conditions and fish population assemblages (Mattice, 1990).
Angled screens are used to guide fish to a bypass by guiding them through the channel at some angle to the flow. Coarse-mesh angled screens have been shown to be highly effective with numerous warm- and cold-water species and adult stages. Fine-mesh angled screens have been shown in laboratory studies to be highly effective in diverting larval and juvenile fish to a bypass with resultant high survival. Performance of this device can vary by species, approach velocity, fish length, screen mesh size,screen type, and temperature (Stone and Webster, 1986).
Angled rotary drum screens oriented perpendicular to the flow direction have been used extensively to lead fish to a bypass. They have not experienced major operational and maintenance problems. Maintenance typically consists of routine inspection, cleaning, lubrication, and periodic replacement of the screen mesh (Stone and Webster, 1986).
An inclined plane screen is used to divert fish upward in the water column into a bypass. Once concentrated, the fish are transported to a release point below the dam. An inclined plane pressure screen at the T.W. Sullivan Hydroelectric Project (Willamette Falls, Oregon) is located in the penstock of one unit. The design is effective in diverting fish, with a high survival rate. However, this device has been linked to injuries in migrating fish, and it has not been accepted for routine use (Stone and Webster, 1986).
Louvers consist of an array of evenly spaced, vertical slats aligned across a channel at an angle leading to a bypass. They operate by creating turbulence that fish are able to detect and avoid (Stone and Webster, 1986).
Submerged traveling screens are used to divert downstream migrating fish out of turbine intakes to adjoining gatewell structures, where they are concentrated for release downstream (Figure 6-9). This device has been tested extensively at hydropower facilities on the Snake and Columbia Rivers. Because of their complexity, submerged traveling screens must be continually maintained. The screens must be serviced seasonally, depending on the debris load, and trash racks and bypass orifices must be kept free of debris (Stone and Webster, 1986).
- e. Spill and Water Budgets
Although used together, spill and water budgets are independent methods of facilitating downstream fish migration.
The water budget is the mechanism for increasing flows through dams during the out-migration of anadromous fish species. It is employed to speed smolt migration through reservoirs and dams. Water that would normally be released from the impoundment during the winter period to generate power is instead released in the May-June period when it can be sold only as secondary energy. This concept has been put into practice in some regions of the United States, although quantification of the benefits is lacking (Dodge, 1989).
Spill budgets provide alternative methods for fish passage that are less dangerous than passage through turbines. Spillways are used to allow fish to leave the reservoir by passing over the dam rather than through the turbines. The spillways must be designed to ensure that hydraulic conditions do not induce injury to the passing fish from scraping and abrasion, turbulence, rapid pressure changes, or supersaturation of dissolved gases in water passing through plunge pools (Stone and Webster, 1986).
In the Columbia River basin (Pacific Northwest), the Corps of Engineers provides spill on a limited basis to pass fish around specific dams to improve survival rates. At key dams, spill is used in special operations to protect hatchery releases or provide better passage conditions until bypass systems are fully developed or, in some cases, improved (van der Borg and Ferguson, 1989). The cost of this alternative depends on the volume of water that is lost for power production (Mattice, 1990). Analyses of this practice, using a Corps of Engineers model called FISHPASS, show that the application of spill budgets in the Columbia River basin is consistently the most costly and least efficient method of improving overall downstream migration efficiency (Dodge, 1989).
The volume of a typical water budget is generally not adequate to sustain minimum desirable flows for fish passage during the entire migration period. The Columbia Basin Fish and Wildlife Authority has proposed replacement of the water budget on the Columbia River system with a minimum flow requirement to prevent problems of inadequate water volume in discharge during low-flow years (Muckleston, 1990).
- f. Fish Ladders
Fish ladders are one type of structure that can be provided to enable the safe upstream and downstream passage of mature fish. One such installation in Maine consists of a vertical slot fishway, constructed parallel to the tailrace, which allows fish to pass from below the dam to the headwaters (ASCE, 1986). The fishway consists of a series of pools, each 8.5 feet by 10 feet in size, which ascend in 1-foot increments through the 40-foot rise from the
tailwater area to the headwater area. When there is no flow in the spillway, fish can pass downstream through an 18-inch pipe. Flow is provided in the tailrace during fish migration season. Fish prefer to travel in these fishways at night under low illumination (Larinier and Boyer-Bernard, 1991).
Information on the effectiveness of these types of structures is scarce and inconclusive, according to a study by the General Accounting Office (GAO, 1990). GAO noted that many studies of bypass facilities have emphasized data collection to document the number of juvenile fish entering the bypass structures and the condition of the individuals after passage is completed. Only two studies were identified in which bypass methods were compared with alternative methods to identify the most successful approaches. The observations collected at Lower Granite Dam and at Bonneville Dam (Columbia River) indicate a higher survival rate for young fish passing through turbines than for those passing through a bypass structure.
- g. Transference of Fish Runs
Transference of fish runs involves inducing anadromous fish species to use different spawning grounds in the vicinity of the impoundment. To implement this practice, the nature and extent of the spawning grounds that were lost due to the blockage in the river need to be assessed, and suitable alternative spawning grounds need to be identified. The feasibility of successfully collecting the fish and transporting them to alternative tributaries also needs to be carefully determined.
One strategy for mitigating the impacts of diversions on fisheries is the use of ephemeral streams as conveyance channels for all or a portion of the diverted water. If flow releases are controlled and uninterrupted, a perennial stream is created, along with new instream and riparian habitat. However, the biota that had been adapted to preexisting conditions in the ephemeral stream will probably be eliminated. One case where an ephemeral stream was used to convey water and create alternative instream habitat for fish is along South Fort Crow Creek, in Medicine Bow National Forest, Wyoming. After 2 years of diversion, the amount of stream channel on an 88-km reach had increased 32 percent. Some measure of the success with which alternative instream habitat has replaced the original conditions can be seen in the total area of beaver ponds, which doubled within 2 years of completion of the project (Wolff et al., 1989).
- h. Constructed Spawning Beds
When the adverse effects of a dam on the aquatic habitat of an anadromous fish species are severe, one option may be to construct suitable replacement spawning beds (Virginia State Water Control Board, 1979). Additional facilities such as electric barriers, fish ladders, or bypass channels will have to be furnished to channel the fish to these spawning beds.
a. Costs for Minimum Flow Alternatives
In a comparisons of costs of minimum flows alternatives at South Fork Holston River, Adams and Hauser (1990) describe costs for a variety of practices, including an estimated total direct cost of $539,000 for a reregulating weir and $1,258,000 for a small hydro unit.
b. Costs for Hypolimnetic Aeration
The diffused air system is generally the most cost-effective method to raise low DO levels (Henderson and Shields, 1984; Cooke and Kennedy, 1989). However, the costs of air diffuser operation may be high for deep reservoirs because of hydraulic pressures that must be overcome. Any destratification that results from deployment of an air diffuser system will also mix nutrient-rich waters located deep in the impoundment into layers located closer to the surface, increasing the potential for stimulation of algal populations. The mixing must be complete to avoid problems with algal blooms (Cooke and Kennedy, 1989).
Fast and others (1976) and Lorenzen and Fast (1977) discuss costs of hypolimnetic aeration. The following are capital cost items for aeration systems: air lift devices, the compressor, the air supply lines, and the diffusers. The costs for these items are dependent on aerator size, which in turn is dependent on the need for oxygen in the reservoir impoundment (McQueen and Lean, 1986). Cooke and Kennedy (1989) reported side stream pumping costs (adjusted to 1990 dollars) were $347,023 (capital costs) and $167,240 (yearly operation and maintenance costs). Partial air lift system costs (adjusted to 1990 dollars) were reported by Cooke and Kennedy (1989) as $627,150 (capital costs) and $105,257 (operation and maintenance costs). Capital costs for full air lift systems ranged (in 1990 dollars) from $250,860 to $585,340, and operation and maintenance costs (in 1990 dollars) were reported as $44,862 (Cooke and Kennedy, 1989). In the opinion of Cooke and Kennedy (1989), the full air lift system is the least costly to operate and the most efficient. Furthermore, there is the potential for surface water quality problems caused by the supersaturation of nitrogen gas with the use of the partial air lift system (Fast et al., 1976). Accordingly, the full air lift system seems to be the overall best choice for aeration, based on cost, efficiency, and environmental impacts.
c. Costs for Diffusers
A cost-effective means of achieving better water quality for reservoir releases is to aerate discrete layers near the intakes to avoid any unnecessary release of algae and nutrients into tailwaters below the dam. In another test at facilities operated by the Tennessee Valley Authority (TVA), diffusers were deployed at the 70-foot depth of Fort Patrick Henry Dam near one of the turbine intakes. Levels of DO in the tailwaters increased from near zero to 4 mg/L as a result of operation of this aeration system. Unfortunately, the operation costs of this kind of system were determined to be relatively high. An operation system to increase the DO in the discharge from both hydroturbines at Fort Patrick Henry Dam to 5 mg/L would have an initial capital cost of $400,000 and an annual operating cost of $110,000 (Harshbarger, 1987).
The TVA has determined that approximately $44 million would be required to purchase and install aeration equipment at 16 TVA facilities (TVA, 1990). The aeration of reservoir waters, combined with other practices such as turbine pulsing, would result in the recovery of over 180 miles of instream habitat in areas below TVA dams. An additional $4 million per year in annual operating costs would also be required.
d. Costs for Aeration Weirs
The estimated costs for an aeration weir constructed downstream of the Canyon Dam (Guadalupe River, Texas) were $60,000. However, the construction of this device occurred at the same time as other construction at the facility, resulting in a reduction in overall project costs (EPRI, 1990).
e. Costs for Fish Bypass System
The Philadelphia Electric Company installed a fish lift system on the Conowingo Dam, located on the Susquehanna River at the head of the Chesapeake Bay. The fish lift system has the capacity of lifting 750,000 shad and 5 million river herring per year. The system was completed in 1991 at a total cost (adjusted to 1990 dollars) of $11.9 million (Nichols, 1992).