Downstream Fish Passage Technologies: How Well

| 69

4

DownstreamFish Passage

Technologies:How Well

Do They Work?

he implementation of downstream miti-gation for fish passage at hydropowerfacilities has three distinct goals: totransport fish downstream; to prevent

fish from entrainment in turbine intakes; and tomove fish, in a timely and safe manner, through areservoir.1

A range of mitigation methods for down-stream passage and for prevention of turbineentrainment exist, and some have been appliedwith more success than others. The so-called“standard” or “conventional” technologies aremainly structures meant to physically exclude or“guide” fish to a sluiceway or bypass around theproject and away from turbine intakes by meansof manipulating hydraulic conditions. Other“alternative” technologies attempt to “guide”fish by either attracting or repelling them bymeans of applying a stimulus (i.e., light, sound,electric current). Many theories have beenapplied to the design of downstream passage sys-tems and further experimentation is underway insome cases (see box 4-1).

1 The main difference between up- and downstream passage is that upstream moving fish may keep trying until they find a means of pas-sage (i.e., a fishway). A downstream migrating juvenile has one chance to find the proper passage route, otherwise it becomes entrained.

For downstream migrating species, includingthe juveniles of anadromous upstream spawners,it is important that a safe route past hydropowerfacilities be made available. For these fish, ameans of preventing turbine entrainment, via adiversion and bypass system, is often needed(242,243) (see box 4-2). For some resident fish,downstream movement may not be critical ordesirable. Philosophies of protection vary acrossthe country depending on target fish, magnitudeof the river system, and complexity of the hydro-power facility. For example, practitioners in theNorthwest tend to prefer exclusion devices thatphysically prevent entrainment, while those inthe Northeast tend to recommend structuraldevices that may alter flow and rely on fishbehavior for exclusion.2 Much of the variance inprotection philosophy may be linked to differ-ences in target fish in these regions. The North-west hosts a number of endangered or threatenedspecies (mainly salmonids), while the Northeastdoes not have quite the same history of concern.In the Northwest, fish protection is mainlyfocused on salmonids. Downstream migrants

2 The mechanism that causes fish to be guided by angled bar racks is not well understood.

T

70 | Fish Passage Technologies: Protection at Hydropower Facilities

tend to be small and have limited swimming abil-ity. In the Northeast, fish protection is focused ona variety of species. In some cases downstreammigrants are of fairly good size and possessfairly good swimming ability (e.g., Americanshad).

Physical barriers are the most widely usedtechnology for fish protection. These technolo-gies include many kinds of screens (positionedacross entrances to power canals or turbineintakes) providing physical exclusion and protec-tion from entrainment. In some parts of the coun-try, behavioral guidance devices such as angledbar racks (modified versions of conventionaltrashracks) are used to protect fish from turbineentrainment. For both categories of downstreampassage technologies, careful attention to dimen-sions, configurations and orientations relative toflow is required to optimize fish guidance.3

In most cases, structural measures to excludeor guide fish are preferred by resource agencies.Screens and angled bar racks providing structuralmeasures for physical guidance are preferred byresource agencies, however, the screens can beexpensive to construct and maintain. As a result,

3 Fish impingement on screens or trashracks can stress, descale and otherwise injure fish, particularly juveniles (168, 190).

the development of alternatives to these technol-ogies, such as alternative behavioral guidancedevices (e.g., light, sound), continues to beexplored. These devices have not been proven toperform successfully under a wide range of con-ditions as well as properly designed and main-tained structural barriers. Thus, the resourceagencies consider them to be less reliable in thefield than physical barriers. In addition, othermethods for downstream passage are also beingexplored. New turbine designs that will be notonly more efficient but more “friendly” to fishare under proposal. And in the Columbia RiverBasin, a surface collector system which intendsto guide fish past hydropower facilities by betteraccommodating natural behavior is being experi-mented with at a number of sites.

DESIGN OF CONVENTIONAL STRUCTURAL MEASURESProgress in developing effective downstreamfish passage and protection mechanisms hasoccurred over the past 50 years (203,205,221).Physical barrier screens and bar racks and lou-

BOX 4-1: Chapter Findings—Downstream Technologies

■ There is no single solution for designing downstream fish passage. Effective fish passage design for aspecific site requires good communication between engineers and biologists and thorough understand-

ing of site characteristics.■ Physical barrier screens are often the only resource agency-approved technology to protect fish from tur-

bine intake channels, yet they are perceived to be very expensive.■ The ultimate goal of 100 percent passage effectiveness is most likely to be achieved with the use of phys-

ical barrier technologies, however site, technological, and biological constraints to passing fish around orthrough hydropower projects may limit performance.

■ Structural guidance devices have shown to have a high level of performance at a few studied sites in theNortheast. The mechanism by which they work is not well understood.

■ Alternative behavioral guidance devices have potential to elicit avoidance responses from some speciesof fish. However, it has not yet been demonstrated that these responses can be directed reliably; behav-

ioral guidance devices are site- and species-specific; it appears unlikely that behavioral methods will per-form as well as conventional barriers over a range of hydraulic conditions and for a variety of species.

SOURCE: Office of Technology Assessment, 1995.

Chapter 4 Downstream Fish Passage Technologies: How Well Do They Work? | 71

BOX 4-2: Complements to Exclusion, Diversion, and Guidance Technologies

Once fish are diverted by physical screens, angled bar racks, or louvers, a means of passing them aroundhydropower projects is needed. This is achieved through the use of bypasses and sluiceways. These measures

would also be required for any emerging behavioral guidance technologies.

Bypasses

Engineered bypass conduits are needed for downstream-migrating fish at hydropower facilities and are the

key to transporting fish from above to below a hydropower project. Most early downstream mitigation effortsonly marginally improved juvenile fish survival. Today, juvenile bypass structures are more efficient due to les-

sons learned and a better understanding of the interaction of hydraulics and fish behavior (190). In someinstances bypasses must provide efficient and safe passage for both juvenile and adult life stages (175).

Despite efforts at designing mitigation systems for specific sites, efforts may fail due to inadequately

designed fish bypasses (204). Bypass design should be based on the numbers, sizes, and behaviors of targetspecies (204). The entrance to such channels may be their most important feature. Smooth interior surfaces and

joints, adequate width, absence of bends and negative pressures, proper lighting, and appropriate hydraulicgradients should be considered when designing an effective bypass system (239). High-density polyethylene,

PVC, or concrete cylinders are all appropriate bypass materials (175).

Bypass entrances and the velocity of the flow are critical to success. For example, fish may be less likely to

enter a bypass if met with extremely high flows. Typically, bypass entrances consist of a sharp-crested weirconfiguration which causes an increase in velocity. The development of a new weir, which may be able to be

retrofitted at some applications, will result in gradual velocity acceleration intended to be more attractive tofish.a

Bypass outfalls are also critical in achieving safe downstream passage of target fish. The potential for preda-

tion at bypass exits where fish are concentrated is a particular concern (204). Gulls, squawfish, otters, herons,and other predators often congregate at these outfalls. Submerged outfalls may allow for avoidance of strong

currents, bottom injury, and predation by birds; but they may cause disorientation and have debris problems(175,190). Elevated outfalls may greatly subject fish to predation and disorientation, but avoid problems with

debris. Injury and mortality associated with various bypass structures has rarely been studied, although in somecases it has been high.

Sluiceways

Sluiceways are typically used to bypass ice and debris at hydropower projects, but they can also provide anadequate and generally successful means of downstream passage provided fish are able to locate them. Small

hydropower projects often rely on sluiceways for passage. This type of passage may work well for surface ornear-surface oriented fish (i.e., clupeids, salmonids, and some riverine species) but may not work as well for

fish distributed elsewhere in the water column.

Entrance location, adequate flow, and thorough maintenance and debris removal are critical factors tosluiceway success. The sluiceway should be located to one side of the powerhouse, generally at the most

downstream end, with its outfall located so as not to interfere with the attraction flow of the upstream fishway.The greatest problem associated with sluiceways is the potential for predation at the entrance or exit.

aThe NU-Alden Weir was developed by Alden Research Laboratory with funding from Northeast Utilities, Inc. Testing of theweir took place at the Conte Anadromous Fish Research Center during the spring of 1995. Results were promising.



vers have been used to exclude fishes from tur-bine intakes and are considered to be standard,conventional technologies.4 In cases where thereis a large forebay area, water velocities are high,or site specifications are limiting, these types ofsystems may not be feasible, or the costs may beexceedingly high. Physical barrier screens mayprovide nearly 100 percent protection for migrat-ing (target) fish, but for the aforementioned rea-sons, the development of alternative behavioralguidance techniques (e.g., sound and light) hasbeen, and continues to be, pursued in the publicand private sector.

The design of effective structural measures forassisting in downstream passage of juvenile out-migrants and riverine species is dependent onbehavioral criteria, and the knowledge of physi-cal, hydraulic, and biological information whichare critical to success (13,185). Lack of knowl-edge of fish behavior tends to lead to disagree-ment on what the best available method ortechnology for a particular site might be. Thistype of information is also necessary for thedesign of alternative behavioral guidancedevices. For example, the limited swimmingability exhibited at the juvenile stage is a criticaldesign concern. Flow data, species and popula-tion size, and where the target species tend toexist within the water column will help deter-mine location and type of passage system neces-sary.

Downstream passage design must take intoconsideration the lack of or limited swimmingability of outmigrating (anadromous) juvenileand smolt fishes. Other catadromous and riverinespecies may have limited swimming ability aswell, depending on age and size. Where largercatadromous fishes and anadromous adult repeatspawners are concerned, entrainment avoidancemight be more related to behavior than to physi-

4 Bar racks and louvers are considered standard technologies for application in the Northeast, but not in the Northwest.

cal swimming ability. Where hydropowerprojects exist in series, a system of reservoirsmay be created where velocities are low andwater temperatures elevated. These conditionsmay alter fish behavior and slow outmigration ofjuveniles that are dependent on water flow toassist their movement. The series of four dams(McNary, The Dalles, John Day, Bonneville) onthe lower Columbia River, for example, can addup to 20 to 30 days to the travel time for juvenilefish due to alteration in flow conditions (230).

Screens as well as bar racks generally aredesigned to work with site hydraulics to help orencourage fish in moving past or away from tur-bine intakes. Well-designed screen facilities mayresult in a guidance efficiency of over 95 percent(see appendix B) (236,236A). The effectivenessof bar racks is less conclusive. The size and costof screen and bar racks systems depends on thesite. However, water velocities in the forebay ingeneral, and the approach velocity5 in front ofthe system in specific, are of primary concern.The idea is to maintain approach velocitieswithin the cruising speed of all target fishes to bescreened in order to achieve protection (58).

❚ Physical Barriers

ScreensOutmigrating juvenile salmonids depend a greatdeal on hydrology and hydraulics to guide theirmovement. These fish have limited swimmingability and orient themselves into the flow.6

Therefore, downstream protection devices musttake advantage of natural fish behavior. At manyhydropower projects a physical barrier is used inconjunction with a bypass to facilitate passage.The flow characteristics that are generated by theparticular placement of a screen and the physicalparameters of the screen itself help to guide fish

5 “Approach velocity” is the velocity component of flow normal to and approximately three inches in front of the screen face. Fisheriesagencies determine this value based on the swimming capabilities of the smallest and/or weakest fish present (239).

6 Some salmonid pre-smolts have good swimming ability (i.e., sockeye, coho, steelhead), while others (i.e., pink and chum) smolt shortlyafter emergence and their swimming ability does not change significantly during smoltification. Not as much is known about how othermigratory species (e.g., American shad, blueback herring) behave during outmigration (187).


to the bypass. The key to successful downstreampassage is to employ the fish’s behavior to guidethem to a safe bypass. The hydraulics of thestructure must be benign enough that the fish canbe guided to safety before they fatigue or areinjured.

Physical barrier screens can be made of vari-ous materials based on the application and typeof screen (i.e., perforated plate, metal bars,wedgewire, or plastic mesh). Screens aredesigned to slow velocities and reduce entrain-ment and impingement (78). Smooth flow transi-tions, uniform velocities, and eddy-free currentsjust upstream of screens are desirable. Adequatescreen area must be provided to create a low flowvelocity that enables fish to swim away from thescreen.

The positioning of the screening device is crit-ical. It must be in appropriate relationship to thepowerhouse to guide fish to the bypass by creat-ing the appropriate hydraulic conditions. Fishthen enter a bypass which either deposits them ina canal that eventually rejoins the main channel,releases them into the main flow downstream ofthe project via an outfall pipe or sluiceway, orleads them to a holding facility for later trans-port. Outfall pipes typically release fish abovethe water’s surface to avoid creation of a hydrau-lic jump or debris trap within the closed pipe.Releasing fish above the water may also alleviatedisorientation and help to prevent schooling.However, predation at the outfall can be a prob-lem and there is no consensus on how to avoidthis, though multiple outfalls might alleviate thesituation in some cases (188).

The screen must be kept clean and clear ofdebris or it will not function properly. Debris iscommonly the biggest problem at any screen andbypass facility. Debris loading can disrupt flow

and create high-velocity hot spots, or causeinjury to fish (238). In addition, a partiallyblocked bypass entrance can reduce the effi-ciency of fish passage and cause injury or mor-tality (190) (see box 4-2). Installation andoperation of a screen cleaning system and regularinspections to ensure proper operation of screensmay be the most important activities to increaseeffectiveness. Mechanical cleaning systems arepreferable over manual ones and often more reli-able, provided they are functioning properly.Very frequent cleaning may be needed wherethere is a lot of debris. California screen criteriarequire cleaning every five minutes. Ideally,screens should be cleaned while in place, andtemporary removal of a screen for cleaning isusually not acceptable (12).

A variety of physical barrier screens has beendeveloped to divert downstream migrants awayfrom turbine intakes.7 Years of design, experi-mentation, evaluation, and improvement havealleviated some problems but others still remain,and no physical barrier is 100 percent effective inprotecting juveniles. Few studies have been ableto demonstrate conclusively a guidance effi-ciency exceeding 90 percent; and although theeffectiveness of these facilities is probably closeto 100 percent at many sites, losses of fish mayoccur due to predation or leakage of fish pastfaulty or worn screen seals (59). However,improvements in screen components have beenmade and designs have begun to reflect newknowledge about hydraulics. Some specifics ofdesign and function of a variety of low-velocity8

physical barrier screens are highlighted below.The drum screen is often found to provide the

best fish protection at sites with high debrisloads. Comprehensive evaluation of large drumscreen facilities has demonstrated nearly 100

7 Between 1985 and 1989, a series of evaluation reports on the performance of diversion screens in use at irrigation and hydroelectricdiversions in the Yakima River Basin, Washington, were jointly produced by the U.S. Department of Energy, Bonneville Power Administra-tion, and Batelle PNW Laboratory. The reports evaluate flow characteristics of the screening facilities. A discussion of these sites is notincluded in OTA’s report; however, they were used by resource agencies in developing screen criteria in the Northwest and therefore thereports deserve mention (244, 245, 246, 247, 248, 249, 250).

8 The physical barrier screens discussed in this section are considered to be low-velocity screens, meaning that they can function at veloc-ities (perpendicular to the screen) between 0.33 to 0.5 feet per second (59).


percent overall efficiency and survival (12). Thedrum rotates within a frame and is operated con-tinuously for cleaning. Debris is carried over thedrum and passed down a channel or into a bypass(175). Drum screens can be expensive to con-struct and install, but relatively economical tooperate; however, application criteria are sitespecific. These screens have been proven to bereliable at sites in California and the PacificNorthwest (204). Relatively constant water lev-els in the forebay are necessary for operation,and maintenance and repairs to seals can beproblematic and costly.

Simple fixed screens can be an economicalmethod of preventing fish entry into waterintakes at sites where suspended debris is mini-mal; however, costs are site specific. Thoughfixed panel screens can and have been built inareas with substantial debris, automatic screencleaners are required. These screens have dem-onstrated greater than 95 percent overall effi-ciency and survival at sites in the ColumbiaRiver Basin (12). Several types of simple fixedscreen are available. The stationary panel screenis a vertical or nearly vertical wall of mesh pan-els installed in a straight line or “V” configura-tion. Fish-tight seals are easily maintainedaround this fixed screen, and the design accom-modates a range of flows and forebay water ele-vations (175).

Inclined plane screens are also stationary, butare tilted from the vertical to divert fish up ordown in the water column to a bypass. A con-ceivable problem with this design is the potentialfor dewatering of the fish and debris bypass routeif water levels should fall below either end of thetilted screen. Also, cleaning is a primary concernfor both stationary panel and inclined planescreens (175). Manual brushing is usuallyrequired to keep surfaces debris-free. The designis practical for water intakes drawing up to 38cubic meters per second (175,204); however,application depends more on the site than on theflow.

Submersible traveling screens (STSs) areexpensive to construct and install, and subject tomechanical failures, although in some cases they

have been considered by the U.S. Army Corps ofEngineers to be the best available technology fordiverting downstream migrating fish in theColumbia River Basin (204). STS configurationsoperate continuously during the four- to nine-month salmonid migration period in the Colum-bia River; they are capable of screeningextremely large flows in confined intakes but donot screen the entire powerhouse flow (175,204).At hydropower facilities where the fish are con-centrated in the upper levels of the water column,good recoveries have been achieved (65). How-ever, intakes at projects in the Basin tend to bevery deep (i.e., greater than 90 feet) and flowsare high. Under these conditions, fish have beenseen to try to move away from STSs, especiallyif they are deeper in the intake. Also, the poten-tial for impingement is greater due to highthrough-screen flow velocities (175). Thesescreens seem to work better for some speciesthan others.

Vertical traveling screens were originallydesigned to exclude debris from water intakesbut were found to be effective at guiding or lift-ing fish past turbine intakes. The screen mayconsist of a continuous belt of flexible screenmesh or separate framed screen panels (baskets).Vertical traveling screens are most effective forsites where the intake channel is relatively deep.If approach velocities are kept within the cruis-ing speed of the target fish, impingement can begreatly reduced (175,204). However, travelingscreens that lift fish are not recommended forfish that are easily injured, such as smoltingsalmonids.

❚ Structural Guidance DevicesFish passage devices designed with the goal ofguiding fish by eliciting a response to specifichydraulic conditions are described below.

Angled Bar or Trash Racks and LouversAngled bar racks and louvers are used to directjuvenile fish toward bypasses and sluiceways athydropower plants. These structural guidancesystems are devices that do not physically


exclude fish from intakes, but instead createhydraulic conditions in front of the structures.Theoretically, fish respond to this condition bymoving along the turbulence toward a bypasssystem. The success of these systems is depen-dent on fish response to hydraulic conditions,which means their performance can be poorunder changing hydraulic conditions and for dif-ferent fishes of non-target sizes and species(12,65).

Angled bar and trash racks have become oneof the most frequently prescribed fish protectionsystems for hydropower projects, particularly inthe northeastern United States (59,243), to pre-vent turbine entrainment of down-migratingjuvenile anadromous species (e.g., alosids andsalmonids) (194,242). Most of the angled barracks installed to date consist of a single bank ofracks placed in front of the turbine intake at a 45-degree angle to flow. Although design can varyfrom site to site, most racks consist of 1-inchspaced metal bars with a maximum approachvelocity of two feet per second (15,59).

The angled bar rack is set at an acute angle toflow and with more closely spaced bars than con-ventional trashracks. It can divert small down-stream migrating fish, and larger fish cannottypically pass through the bars. However, the useof close-spaced bar racks creates the potential forimpingement of fish. This is of greatest concernfor species with weak swimming ability and/orcompressed body shapes (59). Most of the angledbar racks have been installed at small hydro-power projects, the majority of which have notbeen evaluated for their performance in effec-tively diverting fish.

Proper cleaning and maintenance of the barand trash rack systems on a regular basis is a crit-ical element of operational success. Racks can beequipped with mechanical cleaning systems orcan be pulled out of the water for manual clean-ing; trash booms can also be helpful in mitigatingdebris loading. The ideal trash boom is designedto carry debris past the fishway exit to the spill-way or falls and out of the forebay area (15).

A louver system consists of an array of evenlyspaced, vertical (hard plastic) slats aligned across

a channel at a specified angle and leading to abypass (59). The louver system, like the angledbar rack, attempts to take advantage of the factthat fish rely mainly on senses other than sight toguide them around obstacles. Theoretically, asfish approach louvers, the turbulence that is cre-ated by the system causes them to move laterallyaway from it toward a bypass (59).

Louvers have been installed at a small numberof locations, but are not generally acceptable as amitigation technology for protecting fish fromturbine entrainment. If approach velocities do notexceed their swimming ability, fish generallyassume a tail-first position and move parallel tothe line of louvers guided by streamflow andhydraulics toward a bypass (204). However, lou-vers may be considered for sites with relativelyhigh approach velocities, large uniform flow andrelatively shallow depths (204), and for somesites with species requiring lesser levels of pro-tection. Louver efficiency in fish diversion,although high for some species, is relatively lowon average compared to true physical barriers.

Passage of Atlantic salmon smolts at the Ver-non and Bellows Falls hydropower projects onthe Connecticut River was evaluated during thespring outmigration in 1995. A newly designedangled louver system at the Vernon site, whichwas based on hydraulic modeling, is in place toguide fish to a primary bypass chute in the mid-dle of the powerhouse. Smolts are spilled into thetailwater of the project. Preliminary data indicatethat about half the smolts are being guided to theprimary bypass, while the remainder are eithersounding beneath the louvers and passingthrough the turbines, or going through the sec-ondary bypass, or are never making it into theforebay due to downmigration behavior (94).

The system may not be as successful as hopeddue to the fact that the actual hydraulic condi-tions in the forebay of the project are not consis-tent with the modeling. This is mainly a result ofnot replacing certain turbine units adjacent to theprimary bypass. This decision, which was madebased on economics, has led to a less than ade-quate flow regime in the forebay of the project.Data and evaluation have yet to be finalized.


Despite efforts to monitor performance at anyof these hydropower sites on the ConnecticutRiver, information regarding effects of theangled bar rack and louvers on the overallsalmon population in the Connecticut River hasyet to be generated. Though angled bar and trashracks are frequently used to prevent turbineentrainment, evaluations of performance andeffectiveness are rare. As of the writing of thisreport, 36 trash racks have been installed atprojects in the Northeast (U.S. Fish and WildlifeService (FWS)–Region 5); however, few hadbeen evaluated prior to the spring of 1995.9

Louvers operate most efficiently when theyare designed for larger fish of a specific size(175). Tests of a louver system at the J.E. Skin-ner Fish Protective Facility in Tracey, California,showed good guidance for larger juveniles (i.e.,greater than 70 percent) (100). However, thissame system operated poorly under high debrisconditions. Floating louver systems have shownexcellent promise for protecting fish whichmigrate downstream near the water surface(204). However, excessive entrainment on lou-vers of smaller, weaker fish, including juveniles,has caused louvers to be rejected as a design con-cept at most new hydropower installations (190).

There is a great deal of variation in opinionregarding how well, or why, louvers work. A bet-ter understanding of fish behavior could lead toimproved designs for these structural guidancedevices. Currently, they are recommended foruse by the FWS in the northeastern part of thecountry. They are not in use in the Pacific North-west because they have not been found to pro-vide a high enough degree of effectiveness. Thedegree of protection granted by a louver systemis directly related to the target fish, the degree ofprotection being sought, the approach velocity,and the extent that debris is present or is a prob-lem.

9 The trash rack at the Wadhams Project on the Boquet River in northeastern New York, in place to guide down-migrating Atlanticsalmon smolts, has been evaluated. Others include Cabot Station and the Holyoke Canal Louver on the Connecticut River in Massachusetts,and the Pine Valley Project on the Souhegan River in New Hampshire (195).

OTHER METHODS FOR PROVIDING DOWNSTREAM PASSAGEOther methods for providing downstream fishpassage include pumps, spilling, turbine passage,and transportation.

❚ PumpsThe hydropower industry is currently examiningthe application of fish collection systems, orpumps, to collect and divert fish at intakes (220).There are air-lift, screw impeller, jet, and volutepumping systems. These pumps could be used toforce fish into bypass pipes for downstream pas-sage at hydropower projects. Pump size andspeed, however, may affect fish survival (223).

Fish pumps are not widely used because theycan lead to injury and de-scaling as a result ofcrowding in the bypass pipe and to disorientationonce released back into the river environment,and do not allow the fish to move on their own(196). Historically, the conventional wisdom ofthe resource agencies is to use bypass methodswhich allow fish to move of their own volition.However, a major research effort spearheaded bythe Bureau of Reclamation is underway at RedBluff Diversion Dam on the Sacramento River.Tests are being done to evaluate the usefulness ofpumps to pass juvenile salmonids. Both theArchimedes screw and the Hydrostal-Volutepumps are being tested for the effective and safepassage of fish.

❚ SpillingSpill flows, or water releases independent ofpower generation, are the simplest means oftransporting juvenile fish past (over) a hydro-power project and away from turbines (36).Increased spill to flush fish over a dam can beespecially cost-effective when the downstreammigration period of the target species is short,when migration occurs during high river flows,


or where spill flows are needed for other reasons(e.g., to increase dissolved oxygen levels tomaintain minimal instream flows).

Care should be taken to ensure that spillwaymortality does not exceed turbine passage mor-tality (36,243). Consideration of forebay flowpatterns, location of spillway relative to turbineintake, and positive flow to attract fish to spill-ways are all features of effective spillway pas-sage (175).

Spilling is a particularly controversial issue inthe Columbia River Basin (see box 4-3). TheU.S. Army Corps of Engineers (COE) maintainsthat spilling water to pass juvenile fish has beendemonstrated to be the safest, most effective, andone of the lowest-mortality means of gettingjuvenile anadromous fish past hydropowerprojects in the Columbia River Basin. In addi-tion, it is viewed as the only means of enhancingsurvival without additional flow augmentation ordrawdown (229). However, spilling water toassist fish in downstream passage means lost rev-enue for the hydropower operator. The COE rec-ognizes that spill has its own associated risks(231) and has modified some spillways and oper-ations to reduce problems in the Columbia RiverBasin (49). Passing juvenile fish by spillingwater can result in “gas bubble trauma,” or causepressure-induced injury. According to at leastone study, juvenile anadromous fish that pass ahydropower project by means of spill have a sig-nificantly higher rate of survival (98 percent esti-mated) than do fish that pass through the turbines(85 percent estimated) (229). However, this 85percent turbine survival is through low-headdams with Kaplan turbines; survival is muchlower for high-head dams with Francis turbines(12).

Gas Bubble TraumaAs spill water plunges below the dam the hydro-static pressure causes air, mostly nitrogen gas, to

be entrained in the flows. The pressure at the bot-tom of the stilling basins forces the gases intosolution, creating a supersaturated condition. Theslack water and low flow velocities below thedam slow the escape of the gas back into theatmosphere (23).10 When fish absorb this gas,bubbles can form in the bloodstream. This effect,coupled with the pressure changes experiencedwhen fish plunge with the flow and then return tothe surface, can cause traumatic effects and evendeath. This situation is referred to as gas bubbletrauma.

Since the late 1960s, tests on exposure of adultsalmonids to supersaturated water have beenconducted to determine the effects of exposure.The impact that dissolved gas may have on fishat any given time cannot be simply determinedfrom gas saturation measurements. Thus, moni-toring of migrants for signs of gas bubble traumais an important management tool for determiningif dissolved gas levels are having an impact onpopulations (229).

In June of 1994 the National Marine FisheriesService (NMFS) Northwest regional office con-vened a panel of experts to review the biologicaldata concerning dissolved gas effects on fish.Their findings indicate that a dissolved gas levelof 110 percent can protect fish on purely biologi-cal grounds, whereas levels above 110 percenthave the potential to be damaging (231,234).COE policy calls for keeping gas supersaturationlevels at less than 110 percent in the ColumbiaRiver Basin, the level set by Oregon and Wash-ington State water quality standards (231). Somelaboratory research indicates that total dissolvedgas levels above 110 percent in shallow waterincreases mortality observed in laboratory ani-mals. Yet, field responses may be very different,making it difficult to base in-river managementcriteria on laboratory results.11 The NMFSNorthwest office and the Intertribal Fish Com-mission, which represents tribes in the Columbia

10 In the Columbia River Basin dams were built so that the reservoir of one project backs up on the tailwater of the next project upstream,exacerbating the supersaturation problem.

11 For example, juveniles may dive to greater depths to avoid areas of high dissolved gas concentration.


BOX 4-3: Spilling to Facilitate Fish Passage:Debate Over the Effects on Juvenile Salmonids

Spilling water to pass downmigrating fish is being used as an alternative method for protecting juveniles andenhancing survival at mainstem dams in the Columbia River Basin.a Spilling would occur during high flowperiods when juvenile salmonids are in the midst of their downstream migration. However, there is stilldebate over whether this method might do more harm than good.

A 1995 Spill and Risk Management report prepared for the Columbia River Basin notes that spill passage

and associated damage caused by dissolved gases should not generate greater mortality than that caused byturbine passage. The report goes on to say that there is little doubt that increasing the total dissolved gas levels

in laboratory studies results in increasing the levels of mortality observed in laboratory animals in shallow water.By the same token, the report recognizes that mortality levels experienced in the lab are in conflict with those

that would be observed in the natural environment where fish can sound to a safer depth to avoid injury.

The incidence of Gas Bubble Trauma (GBT) has been observed in juvenile anadromous fish during periods

of high flow and spill during the spring out-migration in the Columbia River Basin.b GBT occurs when gas bub-bles or emboli develop in the circulatory systems and tissues of fishes as a result of supersaturated gas-eous conditions in the tailrace waters of hydropower projects. GBT is considered a physical, not apathological, response to an environmental condition (117). The occurrence of GBT has been shown to bedependent, in part, on water temperature, species, genetic composition, and physiological condition offish, as well as proximity and length of exposure to the total gas pressure (3,117).

The data which have been collected in situ as well as in the laboratory are in conflict with some observations

that have been made in the natural environment. Laboratory experiments have indicated that fish exhibit a highlevel of mortality when exposed to constant supersaturated conditions, but in contrast, observations made in

the wild actually indicate that higher survival rates occur in populations migrating under higher spill/flow/TDGconditions. In some of the laboratory situations fish were held at a constant depth and exposed to a constant

level of TDG. In the natural environment, fish would be sounding to different depths and therefore would proba-bly exhibit a different response. As a result, the usefulness of these tests in the development of a spill manage-

ment plan may be questionable.

The effect of supersaturated conditions on fish is dependent on the depths (i.e., spatial and temporal distri-bution) at which they swim and are present in the water column. Therefore, completing depth distribution studies

would generate helpful information. According to scientists, each meter of depth affords adults a 10-percent reduc-tion in adverse impacts of gas supersaturation. In addition, the length of time it takes for a fish to travel through a

reach of the river, where nitrogen concentrations might be a concern, influences exposure to high levels of dissolvedgas. This is the major factor in determining the impacts a high-level exposure might have on the fish (44a).

These concerns, and mounting political pressure, have led the federal and state governments to set stan-dards for limits on the allowable levels of gas supersaturation in the tailraces of mainstem dams in the Columbia

River Basin. Washington and Idaho have set water quality standards with maximum levels at 110 percent for theColumbia and Snake Rivers, while Oregon has adopted a 105-percent standard. Some have contested that

these standards were set without adequate biological research and information regarding the effects of super-

saturation on fish. In addition, there is concern over the lack of information regarding fish response to the com-bination of supersaturated conditions and reaction with other gases, varying water temperatures, exposure

time, and swimming depth.c In general, a 110-percent standard is considered conservative because thislevel is typically observed, if not exceeded, in the Columbia River Basin with no discernible impacts onfish. Therefore, scientists and resource managers argue that the impacts that supersaturated conditionshave on fish can only be determined by monitoring migrants for signs of trauma, and monitoring naturalenvironmental conditions.

(continued)


River Basin, recently recommended that spillingshould be implemented on a broader scale to sup-port juvenile downstream migration.

❚ Turbine PassageAn explicit assumption behind the design ofdownstream bypass systems at hydropower facil-ities is that fish mortality associated with thebypass will be significantly less than turbinemortality (see figure 4-1; see chapter 2 for an in-depth discussion of turbine entrainment and mor-tality). This assumption is reasonable for manysmall-scale facilities, but is not always borne outat hydropower plants with large, efficient tur-bines (243). For example, studies at Bonneville

Dam on the Columbia River indicate that sub-yearling Chinook salmon suffered more short-term mortality in screen/bypass systems thanwhen passed through turbines, perhaps due topredation at outfalls (242). In a review of studiesat 64 turbine installations, fish mortality rangedfrom zero to more than 50 percent (204). Tur-bine-induced fish mortality may be greatly over-estimated or underestimated (206), and can varyconsiderably from site to site.

Turbine passage exposes outmigrating juve-niles to blades, which can either de-scale or killthem, and distinct pressure changes, which cancause physical injury and/or death. Turbine mor-tality increases with fish size, suggesting thatphysical impact is also important (51,87). At the

It is difficult to monitor the response of fish to supersaturated conditions because mortality may occur beforeany physical characteristics are evident. After death, the external signs of GBT (i.e., large body blisters) may

disappear within 24 hours, leaving dissection the only option by which to make determinations regarding causeof mortality (52). However, swimming performance, physical growth, and blood chemistry can be adversely

affected, leaving weaker fish more susceptible to predation, disease, and migration delay (47).

The National Biological Survey’s research lab in the Columbia River Basin has instituted a Smolt MonitoringProgram (SMP) to be implemented in 1995. The SMP will monitor biological parameters in both the tailwater and

the reservoir of a number of dams on the Lower Snake and Lower- and Mid-Columbia. Ideally, data resultingfrom the SMP will give managers a sense of what the existing levels of supersaturation are so that an appropri-

ate spill management plan can be developed.

Recently, a study of hatchery Chinook test fish (juvenile fall Chinook salmon) being in net-pens below Ice

Harbor Dam on the Snake River resulted in mortality during a study of the effects of high quantities of dissolvednitrogen. While the exact cause of mortality was not known, an uncontrolled spill of heavy spring runoff was

occurring at the dam and all the dead fish had signs of GBT.

Events such as these have kept the debate over spilling to facilitate passage of juvenile outmigrants at a pre-mium. And despite all past studies, there is still great disagreement and many unanswered questions that

remain regarding the level of dissolved gases that can be safely tolerated by juvenile salmonids.

aJuvenile salmon passed via spill as opposed to going through the turbines have a higher survival rate (98 percent) than thoseexposed to turbine passage (85 percent) (Scientific Rationale for Implementing a Summer Program to Increase Juvenile SalmonidSurvival in the Snake and Columbia Rivers, by: Columbia Inter-Tribal Fish Commission, ID Dept. of F&G, OR Dept. of F&W, USFWS,WA Dept. F&W).

b Spilling has been implemented at mainstem COE dams since 1989 under 1989 MOA (protection of juveniles until functionalbypasses are installed) and at Mid-Columbia PUD dams since 1983 under the Mid-Columbia FERC Proceedings. Studies haveshown mortality from turbine passage to be 8 to 32 percent compared to 0 to 4 percent for spillway passage.

c Some research has indicated that swimming stamina is affected at concentrations of 110 percent, growth is affected at 105to 115 percent, and blood chemistry is affected at 115 percent.


BOX 4-3: Spilling to Facilitate Fish Passage:Debate Over the Effects on Juvenile Salmonids (Cont’d.)

80 Fish Passage Technologies: Protection at Hydropower Facilities

I

o I I I I I I 1 1 I I \

Dam Number

Simulation Assumption: Initial population size = 100,OOOTurbine Mortality = 15 percentBypass Mortality = 5 percent per damPredation Mortality = 5 percent per dam

Turbine mortality is only assessed to that part of the group that does not usethe bypass. Bypass mortality is only assessed to that part of the group that usesthe bypass. Predation mortality is assessed to all surviving individuals that passthrough turbines and bypass.

Initial fish Fish surviving

population to next dam

100,000 83,885



edge of the turbine blade are areas of negativepressure that can be strong enough to pull mole-cules of metal from the turbine blades and likewisecan cause damage to fish in the same vicinity.

Various turbine designs have been found to belinked to varying mortality rates for naturally andexperimentally entrained fish.12 Francis turbinesare designed with “fixed” blades to accommo-date a given head, flow, and speed. Kaplan tur-bines have “adjustable” blades which are betterfor low-head operations and seem to be better forfish survivability (i.e., are more “fish friendly”).To evaluate turbine mortality, fish must betagged and released in the intake and then cap-tured in the tailrace. The mark, release, andrecapture technique has been found to be themost effective method of evaluating resultantturbine mortality for salmonid species; however,it has not been proven to be as useful for alosids(51) (see also chapter 2).

Operational factors can also affect turbinemortality rates. Running turbines at maximumoverload during high power demands can resultin higher losses of juveniles (23). In 1967, MiloBell, a hydraulics engineer at the University ofWashington, suggested that the best way toreduce mortality of smolts passing through theturbines was to operate the turbines at maximumefficiency. COE estimates that in most cases inthe Columbia River Basin the expectation forturbine survival is 85 to 90 percent (230).

At Conowingo Dam (hydropower project) onthe Susquehanna River, two old, damaged tur-bines were replaced with new Kaplan-type(mixed-flow) turbines.13 This technology hasbeen marked as more “fish friendly.” The pas-sage of American shad juveniles through the tur-

12 Design changes to reduce turbine mortality include smoothing of conduit surfaces, increasing clearance spaces, decreasing speed ofrotation of turbine blades, reducing the height of the turbine above the tailwater, increasing the depth of the entrance to the penstock, anddecreasing turbine diameter (145).

13 Entrainment survival increased from about 80 percent with the old turbines, to 95 to 98 percent with the new turbines. There are plansto replace the remaining turbines at some point in the future.

bines was evaluated to determine survival rate.The new turbine design is based on a number ofconcepts: it allows for shallow intakes, and asmaller number of blades; it is capable ofincreasing dissolved oxygen in the tailwater; ithas a wide flow range and is non-cavitating;14 italso is greaseless and oil-free. These design con-siderations aim to increase survivability. Otherfactors are equally important to successful pas-sage, such as where the fish exist in the turbine,what the blade strike range is, and what effect thepressure gradient that occurs in the vortexesbetween blades (gap flows) has on the juveniles.Principals in the turbine industry predict thattechnology is moving toward the use of thesevariable speed units.

❚ TransportationTransportation as a means of providing down-stream passage of juvenile fish encompassesboth trap and truck operations and barging.Transporting fish around hydropower facilities isused for a variety of reasons: to mitigate the lossof fish in long reservoirs behind dams; to avoidthe impacts of nitrogen supersaturation that maybe associated with spilling water; to decrease thepossibility of turbine entrainment; and to helpavoid predation problems associated with locat-ing bypass entrances to downstream fish pas-sageways and diversion systems.

The use of transportation to move juvenilesalmonids downstream in the Columbia RiverBasin is to decrease the time it takes for outmi-grants to move through the system.15 However,transportation in the Basin is controversial. Dur-ing high flow periods, the need for transport isdiminished, while during low flows the need for

14 Cavitation occurs when vapor masses collapse on or behind small localized areas of the turbine blade, creating intense negative pres-sure. This results in the loss of metal from the blade. This situation can result in injury to fish and/or oxygen depletion, nitrogen supersatura-tion, other physical stresses, and ultimately mortality.

15 Trucking requires approximately six to eight hours and barging, from the Lower Granite Dam to below Bonneville Dam, about a dayand a half (176).


transportation is favored, in part due to the lengthof time required for the juveniles to movethrough reservoirs (176). During high flowsjuveniles may be bypassed by spilling and maybe able to pass relatively quickly through reser-voirs. However, during times when flows rangesomewhere in the middle, the use of transporta-tion becomes controversial.

In the Columbia River System, juvenilesalmonids are screened from turbine intakes,then loaded onto trucks or barges. After beingtransported downstream, the fish are dischargedbelow the lowest dam, thereby avoiding turbineentrainment and exposure to predators at inter-vening dams. However, juveniles may experi-ence delay in their migration schedule as a resultof transportation, depending on flow rates, pointsof collection, holding time, and points of release.Delay may have a negative impact on physiolog-ical development (i.e., smolting) critical to thesurvival of juvenile salmonids. Fish may also beexposed to diseases, stress, and disorientation.However, the effects of transportation on fishdevelopment and behavior are virtually unknownand little study has been done.

There is strong regional fish agency and tribalsupport for trap and truck operations to movejuvenile fish in the Columbia River Basin, espe-cially during low flow periods. Much work isneeded to improve facilities and operations fur-ther to reduce stress and injury (7).

Barging juvenile fish downstream has drawnmixed reviews although it continues to be sup-ported and promoted by the Army Corps of Engi-neers (230). More barges are scheduled for useduring 1997 in the Columbia River Basin. Barg-ing juveniles has generated support over the useof trucks by virtue of the fact that fish are left inthe water when barges are utilized. However,some controversy remains.

In the Columbia River Basin the focus of thetransportation effort is on increasing smolt sur-vival and improving the numbers of returningadults in future years. Research results are notconclusive regarding the link between transpor-tation and adult returns to spawning grounds(251). There is some evidence that transportation

from rearing to release site does affect salmonhoming, but the extent of the effect is dependenton the status of the salmon (smolt, hatchery resi-dent, or in-river migrant), the method of trans-portation, and the physical distance betweenrearing and release sites (251). However, it hasbeen shown that salmon trucked long distancesdo tend to return to their release site (i.e., belowlowest obstruction on the river), as opposed totheir rearing site (251). Juvenile salmon learn theodors of their home stream, or hatchery, prior toseaward migration and this olfactory memory isessential for the freshwater stages of homing(98). Salmon transplanted prior to smolt stagetend to return to their release site, not their natal(i.e., native) site. Smolts are more likely to returnto the reach of river where they were released(251). Homing patterns may differ depending onwhether fish are transported by truck or barge.

The COE supports the transportation of fish inthe Columbia River Basin. However, due to thelifecycle of salmonids, the length of time spent atsea, and the various obstacles to survival anygiven fish encounters, it is difficult to pinpointcause and effect relationships between theimpacts of either of these methods on population.Although the desirability of transport is contro-versial, there is some agreement that barges arepreferable to trucks; that the release site shouldnot be an estuarine or marine one, but the riveritself; and that fish should be captured after someperiod of migration rather than transported fromthe point of origin; and finally, transportationshould be regarded as experimental (251).

EVOLVING DOWNSTREAM PASSAGE TECHNOLOGIESA number of methods for providing downstreamfish passage are currently under development orbeing experimented with.

❚ Advanced Hydropower Turbine System (AHTS)The heritage of current hydropower turbinedesigns dates from the late 19th and early 20thcenturies, when little was known about environ-


mental conditions and requirements. DOE hastaken a new look at the “turbine system” in aneffort to identify innovative solutions to prob-lems associated with the operation of turbines athydropower projects. DOE and the hydropowerindustry have co-funded the AHTS program.DOE has the lead role in developing and imple-menting the program (26). The hydropowerindustry created a non-profit organization, theHydropower Research Foundation, Inc. (HRFI),which includes 10 utilities that have contributedfunds for the conceptual design phase. HRFI willrepresent and administer industry funds for theprogram. Steering and technical committees con-sisting of representatives for industry, utilities,and other federal agencies are in place to provideprogram direction and technical evaluations.

The purpose of the program is to stimulate andchallenge the hydropower industry to design,develop, build, and test one or more environmen-tally friendly advanced turbine(s). This wouldinvolve the development of new concepts, appli-cation of cutting-edge technology, and explora-tion of innovative solutions (26). Also, theAHTS program will function to develop, con-duct, and coordinate research and developmentwith industry and other federal agencies in orderto improve the technical, societal, and environ-mental benefits of hydropower.

The first phase involves conceptual engineer-ing designs submitted by the industry to a technicalreview committee. The second phase involvesbuilding and testing fully engineered models of themost promising designs. The third phase will con-sist of building and testing prototypes of the mostpromising models in actual operating hydropowerplants. Each phase will be independent of the othersand will follow in succession as the previous phaseis completed. The program will be subject to ongo-ing evaluation by HRFI and DOE.

The AHTS Program completed Phase I during1995. Two firms, Voith Hydro, Inc., and AldenResearch Laboratory, Inc., have been selected for

negotiations toward possible contracts. Phase IIis scheduled to be initiated in the latter part of1995.

The U.S. Army Corps of Engineers (COE) isalso working to develop an advanced turbinedesign that would be more “fish-friendly” bydetermining the mechanisms which affected fishsurvival. Like the DOE effort, the COE isattempting to come up with new turbine designsto increase survival of downstream migrants(34). The COE program is more oriented towardrelatively minor modifications of existing tur-bines in the Columbia River Basin; the DOE pro-gram is focused on developing new designs thatwould be applicable across the United States.

❚ Eicher ScreenThe Eicher Screen was developed in the late1970s by biologist George Eicher in an effort todevelop a better means of bypassing fish safelyaround a turbine. The elliptical screen design fitsinside the penstock at an angle and can functionin flow velocities up to 8 feet per second (fps)(262).16 Non-penstock designs are also possible(54). The screen’s ability to function at relativelyhigh velocities is what distinguishes it from con-ventional screens, which tend to operate at chan-nel velocities of about 1-2 fps (262).

Eicher Screens are relatively less expensiveand have smaller space requirements than mostbarrier screens (175). The system is about 50 per-cent cheaper to install than conventional, low-velocity screening systems, and involves ascreened area about one-tenth that of conven-tional systems. The other benefits of employingthis screen are that it takes up no space in theforebay area, has low operating costs, no risk oficing, and is not dependent on forebay water lev-els. In addition, because the screen operates athigh velocities, there is less chance that it willharbor predators (262)

The approach velocity into the screen violatesmost state and federal screening criteria. EPRI

16 Both the Eicher Screen and the modular inclined screen are considered to be high-velocity screens. This type of screen is supposed tofunction (i.e., safely pass fish) at 8 to 10 feet per second or up to 3 feet per second perpendicular to the screen (59).


supported the University of Washington test ofthe screen’s efficiency. The studies were per-formed under the assumption that the swimmingability and stamina of the fish were inconsequen-tial to the functionality of the screen.17 Tests per-formed in the laboratory as well as in twoprototypes in the field have produced data to sup-port this assumption. Prototype testing has beenperformed at two hydroprojects, the ElwhaHydropower Project near Port Angeles, Wash-ington, and the Puntledge Project at B.C. Hydroon Vancouver Island.

EPRI tested a refined screen design at theElwha Project with promising results. The Elwhatests evaluated screen performance under a rangeof velocity conditions. EPRI’s tests used hatch-ery-raised smolts which were marked and thenreleased into the forebay. After traveling into thepenstock and being guided by the screen, the fishwere bypassed to a collection tank where theywere measured, counted, and classified byamount of de-scaling or injury they had suffered.According to EPRI, the screen had nearly perfectdiversion efficiency (99 percent) for some spe-cies and life stages, indicating its potential forprotecting downstream migrating fish (263).

Diversion efficiency was lower and mortalityhigher for fry of some species and the statisticalvalidity of this non-peer reviewed study has beenquestioned (12). If the screen can pass differentsizes and species of fish it could have wide appli-cation in the hydropower industry. Additionally,EPRI funded a series of hydraulic model testsduring 1992 to evaluate the applicability ofhydraulic data from Elwha to other sites and toevaluate potential for further improvement of theflow distribution via porosity control. To com-plete these tests, a model of the intake, penstock,and Eicher Screen was constructed at AldenResearch Laboratory. The tests evaluated 1) thepossibility that hydraulic conditions at Elwhawere influenced by the bend in the penstock

17 These criteria are not applicable to this type of pressure screen, because the relative flat slope coupled with the high transportationvelocity over the smooth surface funneling into the bypass means that the fish are involuntarily swept into the bypass seconds after passingover the screen (53).

leading up to the screen; and 2) the potential forcreating a more uniform velocity distributionover the length of the screen (263).

The hydraulic model studies indicated that thevelocity distribution at Elwha was not signifi-cantly influenced by the upstream bend in thepenstock (263). The other tests showed that pas-sage survival rates exceeding 95 percent can beachieved for fish in the 1.5 to 2.0 inch range atvelocities up to 7 fps, while smaller fish can beprotected using lower design velocities andcloser bar spacing (263). At the Puntledge, Brit-ish Columbia project, evaluations indicate 99.2percent successful guidance of coho yearlingsthrough the new Eicher Screen (211).

In general, the Eicher Screen has multiple pos-itive operating characteristics. For instance, it isbiologically effective for target fish; the totalcosts of installation are usually less than for othertypes of screen; it is unaffected by changes in theforebay elevations; it takes up no critical space;operation and maintenance costs are negligible;the relatively high velocities at which it can beused make it adaptable to almost all penstock sit-uations (53).

Research and evaluation of the Eicher Screenhas led to approval at specific sites from agencypersonnel who were not otherwise convinced inthe early stages. Agency approval of use at othersites will depend on documentation that thedesign performs well for target fish at velocitiespresent at the site.

❚ Modular Inclined Screen (MIS)EPRI has developed and completed a biological(laboratory) evaluation of a type of high velocityfish diversion screen known as the ModularInclined Screen (MIS). This screen is designed tooperate at any type of water intake with watervelocities up to 10 fps (221). The MIS consists ofan entrance with trash rack, stop log slots, aninclined wedgewire screen set at a 10- to 20-


degree angle to flow, and a bypass for directingdiverted fish to a transport pipe.

This modular screening device is intended toprovide flexibility of application at any type ofwater intake and under any type of flow condi-tions (221). Installation of multiple units at a spe-cific site should provide fish protection at anyflow rate (220) Currently, no fish protectiontechnology has proven to be highly effective atall types of water intakes, for all species, and atall times (i.e., seasonal variability (65)).

To determine viability of the MIS, a testingprogram to evaluate biological effectiveness wasundertaken by EPRI at the Alden Research Labo-ratory (ARL) in Holden, Massachusetts. Evalua-tions at ARL have focused on the designconfiguration which yields the best hydraulicconditions for safe passage and shows biologiceffectiveness for diverting selected species to thebypass (58).

Mark, release, and recapture tests were under-taken with 11 species including walleye, trout,alosid, and salmon smolts. These species werechosen because they are representative of thosefish that are of greatest concern at water intakesacross the country, based on a review of turbineentrainment and mortality studies that have beenconducted in recent years (62). The tests wereconducted with two screen conditions: cleanscreen (i.e., no debris) and incremental levels ofdebris accumulation. Three replicates were con-ducted at each of the five test velocities and con-trol groups were used to determine mortality andinjury associated with testing procedures. Con-trol fish were released directly into the net penand recovered simultaneously with the test fish.

To assess effectiveness, four passage parame-ters were calculated for each combination of spe-cies, module water velocity, and test condition(i.e., clean screen and debris accumulation) thatwas tested. Success was measured by determin-ing percent of fish diverted live, adjusted latentmortality, adjusted injury rate, and net passagesurvival (221).

According to the 1992 EPRI report, the resultsof the tests “clearly demonstrate that the MIS hasexcellent potential to effectively and safely

divert a wide range of fish species at waterintakes.” The results showed that nearly 100 per-cent of the test fish were diverted live and thatthe adjusted latent mortality was less than 1 per-cent, although this was variable depending onspecies and velocity (58). Fish were safelydiverted over a range of velocities (e.g., 2 to 10fps) with minimal impingement, injury, andlatent mortality; and debris accumulation did notappear to affect fish passage up to certain levelsof debris-induced head loss (221). Also, EPRInoted that it was possible that the testing proce-dures (i.e., transport, marking, fin clipping, net-ting from pen or bypass) may have contributed tothe observed mortality.

ARL has developed a prototype of the MISwhich will be field evaluated in the spillwaysluicegate at Niagara Mohawk’s Green Islandfacility on the Hudson River in September of1995. The prototype MIS test is important in thedevelopment and acceptance of the technology.However, resource agencies will be unlikely toapprove full-scale applications of the MIS with-out additional testing (12). Resource agencies areparticularly troubled by operational aspects ofhigh-velocity turbine screening. These screensonly collect fish when water is flowing overthem. Hydropower operational changes may benecessary to ensure adequate flow to the screens,especially during periods when many hydro-power projects are filling reservoirs and not pro-ducing much power (12).

❚ HydrocombineA hydrocombine design of a hydropower facilityis one where the spillway is situated over the tur-bine intakes. This design was employed at Dou-glas County PUD’s Wells Dam (hydropowerproject) on the Columbia River as a result of thewide success of ice and trash (debris) sluicewaysin passing juvenile fish. Evaluations of thehydrocombine design showed that it too waseffective in providing passage for juvenilesalmonids. As a result, Wells Dam became themodel for research on the “attraction flow” or“surface collection” concept of downstream fish


passage and sparked investigation into the poten-tial for use elsewhere.

The theory was that this combination systemcould improve salmon survival by taking advan-tage of natural behavior and accommodating themajority of juveniles that moved downstream inthe upper portion of the water column. Providinga means of passage over a surface-level spillwayas opposed to forcing juvenile fish to dive to tur-bine intakes is more in line with natural behaviorof outmigrating juveniles. A bypass with verticalslot barrier is placed in the spill intakes, whichcreates an attraction flow for outmigrating juve-niles. Once the fish are entrained in the flow,they enter the bypass and are diverted past thedam instead of passing through the turbines(242). The hydrocombine was shown to producea 90 percent success rate for juvenile fish passingthrough the Wells project (42).

The success of such a system might decreasethe need for spilling, as well as the possibility ofelectricity rate increases. However, the results atWells Dam were not easily explained. As is thecase for many evolving fish passage technologies,there is often a lack of information regarding whythey work. As a result, a prototype was installed atChelan County PUD’s Rocky Reach Dam andGrant County’s Wanapum Dam. The configura-tions of the Wells, Rocky Reach, and Wanapumprojects are significantly different; however, thesurface collection concept is the same. Results arenot yet available on either of these evaluations, butthis research has sparked the development of theCOE’s Surface Collection Program.

❚ Surface CollectorSurface-oriented bypasses could prove to beeffective in improving juvenile salmon survivalin the Columbia River Basin (232).18 There is amajor effort underway in the Pacific Northwestspearheaded by the COE to develop a surfacecollector design (39,77). The thrust of theresearch is to better understand the biologicaland physical principles that are at work at the

18 For a more indepth discussion of the surface collector see appendix A.

Wells Dam, where a hydrocombine design is inuse, and apply them to the surface collector designto provide a safer means of passage for juveniles.This “attraction flow” concept may provide down-stream-migrating juveniles with an alternate, morepassive route through hydropower facilities than ispossible with other methods (42).

Surface collector prototypes are being evalu-ated at The Dalles and Ice Harbor Dams by thePortland and Walla Walla Districts of the COE,respectively. Various configurations of thedesign are being tested. The attraction flow pro-totype consists of a 12-foot-wide by 60-foot-highsteel channel attached to the forebay face of thepowerhouse (42) perpendicular to flow in theforebay. The goal is to guide fish hydraulicallydirectly into the collectors, and then pump themto a bypass which moves them around the dam.

Hydroacoustics will be used to monitor fishmovement and behavior in and near the collector.An adaptation of the new surface collectordesign is in operation at Bangor Hydro’s WestEnfield project on the Penobscot River andEllsworth project on the Union River, althoughdebris blockage has been a problem at both sites.The results of the 1995 testing at Wanapum Damcould potentially add much to what is knownabout downstream fish passage and design athydropower facilities. Also, results of the proto-type tests would hopefully be transferable toother powerhouses at projects on the Columbiaand Snake Rivers (42).

❚ Barrier NetsMost technologies proven to be effective indownstream mitigation at hydropower intakesrely on large screening structures designed toprovide a very low approach velocity. For manyprojects, such technologies are not financiallyfeasible. For others, screens are inappropriate forother reasons. In these cases, the use of barriernets may provide a cost-effective means of pro-tecting fish from entrainment. In general, barriernets have not been utilized in situations where


both downstream passage and protection fromentrainment are desirable.

Barrier nets of nylon mesh can provide fishprotection at various types of water intake,including hydropower facilities and pumpedstorage projects. Nets generally provide protec-tion at a tenth the cost of most alternatives; how-ever, they are not suitable for many sites. Theirsuccess in excluding fish from water intakesdepends on local hydraulic conditions, fish sizeand the type of mesh used. Barrier nets are notconsidered to be appropriate at sites where theconcern is for entrainment of very small fish,where passage is considered necessary, and/orwhere there are problems with keeping the netclear of ice and debris (213). It may not be prac-tical to operate nets in winter due to icing andother maintenance problems. Thus nets may notoffer entrainment protection in winter at somesites.

Nets tend to be most effective in areas withlow approach velocities, minimal wave actionand light debris loads. Biofouling can reduce per-formance, but manual brushing and special coat-ings can help alleviate this problem. Anevaluation was underway during the spring of1995 at the Northfield Pump Storage Project onthe Connecticut River in Massachusetts. Thestudy has yet to be completed. There have beenproblems with debris loading and net inversionwhen flow in the river is reversed due to pump-back at the project.

The Ludington Pumped Storage Plant, one ofthe world’s largest pumped storage facilities,located on the eastern shore of Lake Michigan,has had a 13,000-foot-long barrier net installedaround the intake since 1989. Barrier net effec-tiveness, described as the percentage of fish pro-hibited from entering the barrier net enclosure,was estimated at about 35 percent in 1989, butsubstantially increased to about 84 percent in1994 after significant improvements were made

(90). This seasonal barrier appears to be effectivefor target fish (90).

ALTERNATIVE BEHAVIORAL GUIDANCE DEVICES19

Behavioral guidance technologies include any ofthe various methods that employ sensory stimulito elicit behaviors that will result in down-migrating fish avoiding, or moving away from,areas that potentially impair fish survival. In allcases, the purpose is to get fish to leave a particu-lar area (e.g., a turbine intake) and move some-where else. The nature of the response may belong-term swimming in response to a continuousstimulus where the fish has to move some dis-tance (e.g., a sound that is detected for anextended period of time and from which the fishcontinues to swim), or it may be a “startleresponse” that gets a fish to turn away and thencontinue in a different direction without furtherstimulation. Any stimulus that produces a startleresponse or frightens a fish from a particularplace (essentially exclusion) is not a suitabledeterrent unless there is a component to theresponse that moves the fish in a specific direc-tion that leads to safety as opposed to swimmingaway from the stimulus in a random direction(202).

Fishes, as well as other vertebrates, are capa-ble of detecting a wide range of stimuli in theexternal environment (76). The modalities mostoften detected include sound, light, chemicals,temperature, and pressure. Some fishes candetect electric currents and possibly other stimulithat fall outside of human detection capabilities.

For the most part, behavioral barriers have notbeen approved of and accepted for use by theresource agencies because they have not beenshown to achieve a high enough level of protec-tion (220). In some cases, progress has beenmade in developing technologies that can guidefish, possibly at a lower cost than physical barri-ers. Some in the industry would like to see sub-

19 This section is drawn largely from A.N. Popper, “Fish Sensory Responses: Prospects for Developing Behavioral Guidance Technolo-gies,” an unpublished contractor report prepared for the Office of Technology Assessment, U.S. Congress, June 1995.


stantial investment in developing thesetechnologies for use at sites where complete pro-tection is not required, or as a means of improv-ing the effectiveness of an existing physicalprotection device (220).

Behavior-based technologies are touted asbeing less expensive than physical screeningdevices and easier to install than more conven-tional methods. Another presumed benefit is thatthese technologies can be used with little distur-bance to the physical plant or project operation.Lastly, developers of these technologies claimthat although they have not yet achieved 100 per-cent effectiveness, they have shown that variousbehavioral methods do guide fish, and that guid-ance can be improved upon with research andexperimental application.

❚ LightsMany species of fish have well-developed visualsystems. Light has a high rate of transmission inwater and is not masked by noise. At the sametime, the usefulness of light depends upon theclarity of the water as well as upon the contrastbetween the artificial and ambient light.

The visual system of fishes is highly adaptedto enable different species to see in environmentsthat range from shallow waters of streams togreat ocean depths (142). These adaptationsinclude, for example, the shape of the lens, thedistance between lens and the photoreceptorlayer, the ability to adjust the eye to see objectsat different distances (“accommodate”), andother aspects of the optics of the eye. As oneexample, diurnal fish living in shallow watersoften have yellow corneas (and sometimes yel-low lenses). This serves as an optical filter toscreen out some of the shortwave light which isfound in such waters, and which can scatteraround the eye and decrease visual acuity. Spe-cial adaptations may also be found in the setup ofthe photoreceptors.

While most fish see reasonably well, problemswith use of light include transmission character-istics being very dependent on water turbidity,and variable attenuation of different wave-

lengths. Also, the effectiveness of light is likelyto vary between day and night when the ratiobetween the stimulus light (e.g., strobe) andbackground illumination (e.g., daylight) differs(152).

Two types of lighting are the most widelyused in experiments—mercury and strobe. Of thetwo, experimental results suggest that strobelights (pulsing light) are the more successful inaffecting fish movements, although mercury illu-mination was useful in a number of instances(61,101,163), including attracting and holdingblueback herring at the Richard B. Russell Damto keep them from entering undesirable areas(165,178). At the same time, light may attractsome species and repel others living in the samehabitat (25,76).

One of the earliest studies on use of lights(sealed beams) was by Brett and MacKinnonwho provided data on a limited number of ani-mals moving down a canal away from a lightsource (25). The fish were restrained in a particu-lar region of the canal with nets. The results werenot extensive, but two findings are of interest.First, some species swam away from the lightwhile others did not, suggesting different behav-iors by different species. Second, flashing lightswere more effective at eliciting a response thancontinuous light, a harbinger for the use of strobelights. Response differences to the same lightsource between species have been documentedby others and are not surprising. These differ-ences raise the issue, germane to all stimuli andnot just light, that the stimulus has to be closelyfit to the species being studied.

Strobe light has been extensively evaluated asa fish deterrent in both laboratory and field situa-tions (59). Deterrence has been shown with anumber of species, but the lights have workedmost extensively and effectively with Americanshad juveniles (220). Successful fish deterrencewith strobe lights has often been site specific,which indicates that hydraulic and environmentalconditions and project design and operation haveinfluence on the effect the lights have on species(59). The lack of conclusive results may also be


attributed to inadequate sampling methodologyand design.

Field tests conducted at the York Havenproject on the Susquehanna River demonstrateda strong avoidance response to strobe lights byjuvenile American shad (62,63). The system wasdesigned to repel fish away from the turbineintakes and through the sluiceway. The systemproved to be effective (94 percent). However, thestudy pointed out the need for establishing rela-tionships between behavioral fish bypass sys-tems and site-specific hydraulics in an effort tomaximize bypass efficiency (59). Hydraulic andenvironmental factors had primary influenceover the occurrence, distribution, and behavior ofshad (152). The influence of these factors hasdefinite bearing on the success of the system. Asa result, it was concluded that the proper combi-nation of physical and hydraulic conditions mustexist in the area of the lights and the bypass sys-tem in order to achieve the desired level of effec-tiveness (152). Additional work is underway toverify response of various species.

The use of mercury lights to attract or repelvarious species including salmonids and clupeidsis reviewed by EPRI (57). The results suggestthat such illumination can be used with a numberof species to move fish away from intakes,although the results are quite variable betweensites and species. Such illumination may be moreeffective at night than during the day (not anunreasonable situation considering the contrastbetween the stimulus and ambient illuminationdiffers greatly at night). Incandescent illumina-tion has been tried as a method to modify behav-ior (57), but with no clear success.

Studies conducted at the York Haven projecton the Susquehanna River indicate that mercurylights can be highly effective in attracting giz-zard shad, and several studies have successfullyimproved bypass rates of salmonid species usingmercury or incandescent lighting (57). The rela-tively inexpensive nature of mercury lights is adriving force of research. However, additionalresearch is necessary to determine the feasibilityof using sound as part of a directional bypasssystem (57).

❚ Acoustics (Sound)Sound has many characteristics that make it suit-able for use in the possible modification of fishmovement, especially over longer distances orwhen visibility is marginal. Sound travels at ahigh rate of speed in water, attenuates slowly, ishighly directional, and is not impeded by lowlight levels or water turbidity (201). Moreover,many species of fish are able to detect sounds(69). From the standpoint of directionality, atten-uation characteristics (especially with depth), thelack of effect of turbidity, and suitability duringthe day and night, other potential signals are notas versatile as sound. At the same time, highnoise levels, such as at turbine intakes, may pre-vent fish from hearing artificially generatedsounds in such environments, while high-inten-sity sounds (produced by any source) might havedeleterious effects on fish.

Many fish species are known to use sound aspart of their behavioral repertoire for intra-spe-cific communication. Sounds produced by fishfor communication are generally low-frequency(usually below 500 Hz) and broad-band(159,181). More recently, it has become apparentthat fish are also likely to use sound to get a gen-eral “sense” of their environment, much as dohumans. These sounds may include those pro-duced by surf, water moving against objects inthe environment, or wind action on the surface ofthe water (207). In addition, there is some evi-dence that fishes may respond to sounds that areproduced in association with human-made struc-tures, such as bypass screens and other signalsproduced as a byproduct of hydropower projects(6,164), although little is known about the actualbehavioral responses to these sounds.

It is important to understand that detection ofvibrational signals (which includes sounds) byfishes involves two sensory systems, the ear andthe lateral line. Together, these are often referredto as the octavolateralis system (182). Both sys-tems use similar sensory hair cells as the trans-ducing structure for signal detection and bothrespond to similar types of signals. However,from the perspective of modifying the behavior


of fish with sound, it is probably unimportantwhich sensory system per se is involved with theresponse; however, the distance over which stim-uli can affect each sensory system differs.

A variety of different studies have been con-ducted using sound in attempts to affect move-ment patterns of fish. For the most part, thesestudies have concentrated on various species ofsalmonids and clupeids, although work has beendone with regard to a variety of other species.The range of techniques used has also variedquite widely, as have the sound sources and thefrequencies employed. Results are also quitevariable and range from totally unsuccessful incontrolling behavior to demonstrating potentialusefulness for a few species under certain condi-tions.

Various species of clupeids (herrings and theirrelatives) have been studied by a number ofinvestigators. A major thrust of this work hasbeen to modify the swimming behavior of ale-wives and American shad so that they are keptfrom entering turbine intakes at dams. Someinvestigations have proven unsuccessful, whileothers have achieved some success.

The most compelling studies to date on clu-peids in the United States involve the use ofultrasound to modify the swimming behavior ofAmerican shad and other species at a variety ofsites. These transducers produce high-frequency(approximately 120 kHz) signals that appear toproduce an avoidance response in juvenileAmerican shad, causing them to move awayfrom the sound source. Field studies have dem-onstrated that the effectiveness of sound can bealtered by environmental conditions such aswater temperature or site hydrology. Moreover,sound may be more effective at certain times in

the life cycle of clupeids than at other times, andat certain times of the day or night, possiblydepending upon the particular species beingstudied.20

Early studies on controlling the migration ofsalmonids with sound across a range of frequen-cies generated mixed and somewhat unclearresults (33,156,254). One study showed that ani-mals in a lab setting would respond to certainwavelengths, but there was no apparent responsein a river (254). In another study, attempting toguide trout into a channel using plates set intovibration at 270 Hz, there was some evidence ofsuccess. However, there was no statistical analy-sis, and the limited amount of data does not sug-gest that results were replicated or that othercompounding factors were taken into consider-ation (254).

Hawkins and Johnstone found that Atlanticsalmon would respond to sounds from 32 to 270Hz with best sensitivity from about 100 to 200Hz (99).21 More recently, studies on Atlanticsalmon by Knudsen et al. (128) support the find-ings of Hawkins and Johnstone (99) that this spe-cies only detects very low-frequency sounds.Using a behavioral paradigm, Knudsen and hiscolleagues (126,128) measured the responses ofsalmon to tones from 5-150 Hz. The bestresponses were in the 5-10 Hz range. They alsodetermined that the juvenile salmon would showan avoidance response (in a pool of water) to 10Hz signals but not to 150 Hz signals, althoughavoidance to the 10 Hz signal would only occurif the fish were within 2 m of the soundsource.22,23

Knudsen et al. tested this hypothesis and dem-onstrated that low-frequency sounds could beused to modify salmonid movements in a field

20 Blaxter and Batty (1987) show that the responses to sounds of clupeids changes in the light and in the dark (22).21 Hawkins and Johnstone (1978) trained fish using classical conditioning so that they would show a decrease in heart rate whenever a

sound was presented (99).22 Unconditioned startle responses were also investigated by Stober (217) on the cutthroat trout (Salmo clarki). Stober found that a num-

ber of specimens (but not all) would show an unconditioned startle response to sounds up to 443 Hz, although no response was found below50 Hz. He also showed rapid habituation, as reported by Knudsen et al. (128).

23 While exact distances are different from those reported by VanDerwalker (254), the order of magnitude of the distance from the sourceat which salmonids will respond to sound is the same. These results strongly support the suggestion that the response of salmonids to signalsis when they are close to the sound source and very far into the acoustic nearfield.


experiment (126). They were successful in get-ting salmon to change direction and swim awayfrom a sound source. The stimulus was onlyeffective when the fish were within a few metersof the source (within the acoustic nearfield). Forsuch a system to be fully effective in rivers or bya dam, a large number of projectors would beneeded to insure that fish were properly ensoni-fied.

A similar study reported effective, statisticallysignificant guidance (80-100 percent diversionfrom the entrance to an intake canal for downmi-grating steelhead trout smolts and Pacific Chi-nook salmon smolts) for a patented sound systemnow available from the Energy Engineering Ser-vices Company (EESCO). Natural sounds of var-ious salmonids were recorded, and modifiedforms of the recorded sounds were played backto affect fish movements (141). Results sug-gested that the fish could be as much as 70 feetfrom the projector and the sound would still elicita response. These results have yet to be repli-cated and the study only provided minimal infor-mation as to the nature of the specific soundsused to modify fish movements.

Results from preliminary tests of the EESCOsystem on the Sacramento River in 1993 wereinconclusive (46,94a), largely due to the prelimi-nary nature of the study and problems in experi-mental design. Studies are continuing at theGeorgiana Slough on the Sacramento (171). Theresults of testing that took place during the springof 1994 at Georgiana Slough were encouraging(50 percent overall) and statistically significant(95 percent level) (100).

Infrasound testing is currently underwaywithin the Columbia River Basin as part of theColumbia River Acoustic Program.24 Two typesof sources are being tested, both of which gener-ate infrasound. They differ in one component ofthe sound field they generate. The infrasoundsource, patterned after that used in Norway, isreferred to as an “acoustic cannon” because it

24 Information on the Columbia River Acoustic Program taken from Tom Carlson, Pacific National Laboratory, comments to the Officeof Technology Assessment, August 1995 (39).

generates a sound field with large particlemotion. The acoustic cannon has a 19-cm-diame-ter piston with a displacement of 4 cm (39). Star-tle followed by avoidance has been shown undercontrolled laboratory and field conditions forChinook and steelhead juveniles and smolt. Theother sound source, EESCO technology, gener-ates a sound field with little particle motion. Thissource has a moving coil with a diameter ofapproximately 8 cm with a displacement ofapproximately 0.08 cm (39). The Acoustic Pro-gram has not conducted nor do they know of anycontrolled laboratory behavioral tests of fishresponse to the EESCO technology source.Experience to date indicates that large particlemotion is required to elicit avoidance responsesby salmonids.

Few other fish groups have been tested in asystematic way to determine if they would avoidlow-frequency sounds (69,181). There are, how-ever, remarks in the literature regarding avoid-ance responses of a number of species, and lackof avoidance or any sort of responses by otherspecies. The Empire State Energy ElectricResearch Corporation (ESEERCO) (65a)reported laboratory studies of behavioralresponses to low frequencies by striped bass,white perch, Atlantic tomcod, golden shiner, andspottail shiner (51a, 201a, 201b). Despite somelimitations, the studies demonstrate that whiteperch and striped bass would show an avoidanceresponse to broad-band sounds of below 1,000Hz at sound levels of 148 and 160 dB (re: 1 µPa)during the day, but they showed only a weakresponse at night to sounds as high as 191 dB (re:1 µPa). The other species only showed a weakavoidance response during the day.

Considerable study and data are needed to elu-cidate the mechanisms through which certainfish receive sound. No matter what the actualstimulus, it is of considerable interest that soundcan affect the behavior of certain species eitherby causing a startle response or actually causing


fish to swim away from the source of the sound.It must be kept in mind that a startle responsealone is not sufficient for controlling movementof a fish. Instead, whatever the stimulus, it mustelicit sustained movement of the fish in a specificdirection so the fish avoids the area of danger.

❚ Electric FieldsThere are several recent reports in the gray litera-ture that describe the use of electric fields toguide fish behavior. To date, the results fromthese experiments are equivocal as to their suc-cess in controlling downstream migration of sev-eral different species (20,106).25 A couple ofsignificant points, however, arise from consider-ation of these studies. First, electric fields arepotentially dangerous to other species that mayenter the water in the area of electric field. Sec-ond, the electric fields are restricted to regionsbetween electrodes. Thus, they are most effectivein shallow streams and relatively narrow regionswhere sufficient field strength can be set upbetween opposing electrodes.26

In general, evidence supporting the effective-ness of electrical barriers at supporting the down-stream passage of fish is not available (220).Effectiveness will vary depending on site-spe-cific parameters and species/size-specificresponses. Several problems have been identifiedwith their application, including fish fatigue andthe relationship between fish size and suscepti-bility to electrical fields (59).

A combination electric screen and infrasoundsystem has been extensively tested by Simrad inScotland over the last two years (39). It is novelin the sense that the electric portion of the behav-ioral barrier is used primarily to reinforce theresponse to infrasound by migrating salmonids.The infrasound sources used are large particlemotion sources.

25 The results of testing done during the spring of 1995, of an electrical barrier, at RD 108 (Wilkens Slough) on the Sacramento Riverwere inconclusive (100).

26 There is literature from the manufacturer of electrical guidance systems, Smith-Root, Inc., that their devices can also be used to protectturbine intakes and in other environments than streams. However, this reviewer has seen no analysis, peer-reviewed or “gray” literature, thatevaluates the success of these systems beyond those described in the cited references.

❚ Bubble CurtainsBubble barriers were used by Brett and MacKin-non in an attempt to guide fish, with no apparentsuccess (25). Other researchers suggested thatsuccess with air bubbles may have been associ-ated with the sound that they produce and notnecessarily with the bubbles (107,131). Rugglespoints out that air bubbles are effective for somesaltwater species and possibly for some speciesin streams, but not in rivers. Patrick et al. reportthat air bubbles were effective in producingavoidance behavior in laboratory experimentswith gizzard shad, alewife, and smelt (172). Theyalso reported that avoidance increased when airbubbles were combined with strobes. However,these studies have apparently not been followedup with field experiments. Patrick et al. foundthat air bubbles were most effective when therewas some illumination. They also pointed outthat the basis for fish response was not known,but may have been visual or sound-associated, assuggested by Kuznetsov (131).

Air-bubble curtains have not proven to beeffective in blocking or diverting fish in a varietyof field applications, nor is there data available toindicate potential effectiveness (220). There aresmall-scale studies of water jet curtains in vari-ous field applications; however, mechanical andreliability questions have prevented furtherstudy. Hanging-chain curtains have shown somesuccess in preventing fish passage under labora-tory conditions. Lab results have not been dupli-cated in the field and research has ceased (220).

❚ Hybrid BarriersSome study has been done to evaluate the effec-tiveness of using behavioral barriers in variouscombinations to increase overall effectiveness,yet the results have been equivocal (220). Manyof the field evaluations have been conducted for


application at hydropower projects, including atest combining strobe lights and ultrasonics toguide down-migrating juvenile American shad atthe York Haven Dam hydropower project on theSusquehanna River.

PERSPECTIVES ON TECHNOLOGIESIn an effort to minimize expenditures while stillmeeting protection goals, the hydropower indus-try is looking to implement low-cost fish protec-tion. New behavioral guidance technologies maybe less expensive than conventional fish protec-tion methods (for downstream passage); how-ever, the agencies approach application ofbehavioral technologies with caution and con-sider them to be “experimental.” Therefore, theindustry is reluctant to invest in these technolo-gies for fear that they will simply have to replacethem with more conventional technologies. Thisleads to frustration for the technology vendors.

❚ Resource Agencies

National Marine Fisheries ServiceThe NMFS national office seeks a high level ofeffectiveness for new technologies before theagency will approve application in the field, andin some cases regional offices have releasedposition statements regarding fishery protectionand hydropower. These statements do not applyon a national level, but they do have the potentialto be precedent-setting. The Northwest andSouthwest offices have specific guidelines fordeveloping, testing, and applying alternative fishprotection technologies (see appendix B). NMFSregional offices in the Northwest and Californiastrongly prefer physical barrier screens, whichcan completely exclude fish, for use at hydro-power projects over other structural or behav-ioral guidance devices. In addition, the agencyrequires that experiments evaluating a new tech-nology should parallel the development of a con-ventional (technology) solution.

NMFS maintains that it is critical to requiretechnology developers and the hydropowerindustry to abide by this high standard in order to

uphold the agency’s primary charge to protectfish and because so many fish populations havereached a “crisis” status (257). It is this argumentthat forms the basis of NMFS support of the useof physical barrier screens for fish protectionfrom turbine entrainment. The agency may bemore comfortable with the use of these barriersbecause they physically block or physicallydivert fish, but also because the technologieshave evolved over a fairly long period duringwhich much was learned about how to optimizeperformance and make adjustments based on sitecriteria and biological considerations. In additionto NMFS’s Southwest and Northwest regionaloffices, Washington State’s Department of Fish-eries and Wildlife and the California Departmentof Fish and Game have released statementsregarding screening criteria for salmonids(237,238,239).

U.S. Fish and Wildlife ServiceIn the northeastern United States, FWS may bewilling to consider the application of “experi-mental” devices as an interim or complementarymeasure, depending on the situation and the spe-cies. However, FWS has no formal policy orposition statement regarding the acceptability ofexperimental fish passage technologies. Theagency accepted the use of these technologies incertain limited circumstances, but these weresite-specific decisions based on professionaljudgment, project specific characteristics, andthe significance of the resource at risk (150).

Determinations are a reflection of expert opin-ion and best professional judgment about whatmight work best at a given site. The possibility ofachieving 100 percent efficiency with a passagetechnology, or reducing entrainment to zero per-cent, is unlikely. However, given the status of anincreasing number of threatened and endangeredspecies, the agency may be willing to approvethe application of a technology that fails to reacha 100 percent performance goal, but provides agood level of protection, in situations where thedevelopment of a physical barrier screen orstructural guidance device may take years toachieve.


In the West, FWS is generally inactive onscreening, but is involved to a degree in experi-menting with alternative guidance devices. Theagency has developed interim screen criteria forone species, and supports the use of technologiesthat provide the highest degree of protection pos-sible for target fish at all intakes.

The agency prefers the use of physical barrierand structural guidance devices over alternativeexperimental guidance devices. However, thereis some concern within the agency that constantpressure from vendors to utilize alternativedevices has led to concession in certain cases.The agency is especially concerned that once anexperimental measure is in place at a site it willremain as the long-term protection measureregardless of whether performance is less thanwhat would be expected from a conventionaltechnology. Many agencies view experimenta-tion as a delaying tactic. Although experimenta-tion can be very costly over time (possiblymatching the cost of a conventional approach),yearly expenditures are often much lower thanthe capital outlay to install a conventional tech-nology.

❚ Hydropower IndustryThe industry’s goal is to provide effective fishprotection and to minimize costs, which can be achallenge especially at large hydropowerprojects. The industry claims to be facing diffi-cult economic times, which may be exacerbatedby the possibility of deregulation. This mood hasforced the industry to come out against expendi-tures for what they refer to as seemingly “unnec-essary” items such as fish passage and protectionmitigation technologies.

The possibility of deregulation has alsocaused the industry to reassess its role in thealternative energy market. NHA views hydro-power to be the cleanest, most efficient, and mostdeveloped renewable energy source. As a result,some industry representatives balk at the federalresearch and development investment to advanceand perpetuate other renewable energy sources(i.e., wind, geothermal, solar, etc.), as opposed to

investing in the further improvement and effi-ciency of hydropower on a broad scale. Theindustry claims that it cannot afford to bear thecosts associated with research and developmentof fish passage technologies and that this supportshould come from the federal government.

❚ Technology VendorsVendors of new behavioral and guidance tech-nologies are frustrated by the reluctance ofresource agencies to approve their use despitewhat some consider convincing results in thefield (27,50). Technology developers claim thatthese alternatives to conventional fish protectiontechnologies will work for a fraction of the costof conventional screening mechanisms. Theagencies continue to question “how well” thetechnologies work, and NMFS requires thathydropower operators also pursue a parallel trackwith an accepted technology (e.g., design a phys-ical barrier or other interim measure technologyor method) while an alternative is being devel-oped or tested at a site (174).

Though there is some discussion of allowingthe use of behavioral technologies to enhancephysical barriers or as interim protection devices,the agencies are unwilling to allow these technol-ogies to be utilized as the sole line of defense infish passage mitigation in the absence of scientif-ically rigorous demonstrations of effectiveness.This frustrates the vendors who argue that nosuch evaluation exists for physical barriers andthat behavioral and alternative guidance devicesare being held to a standard that other conven-tional technologies were not during previousyears.

CONCLUSIONSPhysical structures, including barrier screens,angled bar racks, and louvers, that are designedto suit fish swimming ability and behavior, aswell as site conditions, remain the primary down-stream mitigation technologies at hydropowerfacilities. There is general consensus amongpractitioners that the conventional technologieseffectively protect downmigrating fish. Barrier


screens have an appeal in that they are perceivedto be absolute in their operation. According tosome resource agencies, under certain conditionsthey may be the only viable technology. How-ever, the high costs associated with these tech-nologies are often barriers to their use. As aresult, much of the fish passage research pres-ently being done is focused on further develop-ing behavioral guidance devices. Some of thiseffort might be directed toward the installation ofphysical structures because the resource agencieshave identified the need to provide protectionnow while research on behavioral and alternativeguidance devices is taking place.

Extensive descriptions of downstream fishpassage mitigation measures are available(16,59,65). Numerous and varied measures havebeen used to reduce turbine entrainment, includ-ing fixed and traveling screens, bar rack and lou-ver arrays, spill flows, and barrier nets, andalternative behavioral devices. However no sin-gle fish protection system or device is univer-sally effective, practical to install and operate,and widely acceptable to regulatory agencies(37).

With a few interesting exceptions, there is nobehaviorally-based technology that is operation-ally successful in guiding fish. There is potentialfor use of strobe illumination with a number ofspecies, as well as use of infrasound withsalmonids (and possibly other species) and use ofultrasound with clupeids and cod. These investi-gations need to be continued and include bothbasic biology and investigation of field applica-tions of these signals. Very little work is beingdone with electrical stimuli and bubble barriers,and these do not appear to have been broadlysuccessful in earlier studies. There is some evi-dence (165,178) that combinations of sensorystimuli (e.g., light and sound) might be a produc-tive possibility that needs further exploration.

There are major discrepancies in the views ofresource agencies and technology vendors aboutthe potential value and performance of alterna-tive behavioral guidance devices. Part of the dis-crepancy in interpreting performance data hasarisen from lack of a standard approach to testing

and evaluation of the technologies. Vendors willwork closely with clients and consultants butrarely involve the agencies in the early stagesand the decisionmaking process. In addition,though some behavioral guidance devices havebeen shown to elicit an avoidance response infish at certain sites, there are inconsistencies insubsequent years of testing. This type of resulthas caused the resource agencies to question thevalidity of the assumptions and criteria on whichthe studies, and the evaluation, are based. It iscritical to keep in mind that results and methodsdeveloped for large western hydropower facili-ties may not be applicable to much smaller facili-ties in the Northeast and Midwest. At the sametime, methods that do not work at the largerfacility may be very useful and appropriate formuch smaller facilities. In effect, it may beimportant to have research programs directed atdifferent “classes” of sites—such as large hydro-power projects, small hydropower projects,bypasses, etc.

Most of the research on fish exclusion systemshas not been reported in the peer-reviewed scien-tific literature, but appears in progress reports forfunded installations, and may be overly optimis-tic. Often research is not described in sufficientdetail to allow thorough analysis of the results.Thus, it becomes difficult, if not impossible, toassess the effectiveness of many of the tech-niques described or the results reported. Someexperimental results seem at odds with others,and care must be taken in interpreting this infor-mation (204). Conclusions reached should beviewed as tentative.

Many of the earlier studies are weak withregard to behavioral analysis. Methods of analyz-ing the behavioral responses of fish (e.g., meth-ods of observation of fish in experimental pens)have often been poorly described. Also, inappro-priate methods have been used in some cases.This has led some to believe that experimentersdid not use appropriate observational techniques(e.g., “double-blind” experiments where theobservers were unaware of the presence of asound stimulus when reporting the behavior of afish). Moreover, the applicability of techniques


across species, or to the same species under dif-ferent environmental or physical conditions (ageand size), is not well understood. Researchers forthe most part have failed to ask very basic ques-tions about the general behavior of fishes under avariety of conditions and information whichcould be useful in developing bypass systems.

Statistical analyses of behavioral responsesare often inadequate and thus it is hard to assessthe effectiveness of a technique. An issue thatoften arises appears to be differences in waysthat various investigators have used statistics tointerpret data. What may appear to be a positiveresponse in one statistical analysis may appear tobe nonsignificant in another.

Additional studies, with very specific direc-tions, are needed to advance behavioral guidancetechnologies. A key need is to develop a basicunderstanding of the mechanism(s) by whichstimuli elicit responses. In particular, it is notknown how very high-frequency sounds aredetected by clupeids, and basic information toanswer that (and other) questions could helpmarkedly in the design of more suitable controlsystems. Knowledge of the mechanisms of sig-nals detection, the normal behavioral responsesto signals, and the range of signals to which a

fish will respond are critically important in help-ing design appropriate control mechanisms.27

Even basic information on the general behav-ior of fish is often lacking. Thus, it becomesimpossible to predict how a fish might alter itsbehavior when it encounters a hydropower facil-ity or water bypass, how it might respond to vari-ous sensory stimuli (e.g., light or sound),including noxious stimuli, and whether certainsensory stimuli are within the reception capabili-ties of a particular species. Without such basicdata it is very difficult to design a truly effectivemeans for controlling fish behavior.

An interdisciplinary approach to investigatingthe potential for improving fish passage isneeded. Studies should be designed with closecollaboration between fisheries biologists havinginterest and expertise in the needs for fish pas-sage and basic scientists knowledgeable in thebehavior and sensory biology of fishes. Otherimportant specialists would likely includehydraulic engineers and hydrologists, who wouldbring special knowledge of currents and otheraspects of the problem to the discussion, andengineers involved in designing and maintainingbarriers to fish movement. To date, there hasbeen little interaction along these lines.

27 An example of this is found in the work controlling the movement of Atlantic Salmon by Knudsen et al. (126,128). Their experimentaldesign for their field work was clearly based upon their first studies on hearing capabilities (128), as well as the earlier studies of Hawkinsand Johnstone (99).

Downstream Fish Passage Technologies: How Well

Documents