CHAPTER 7 DESIGN GUIDELINES - King County, Washington · ods, and a brief discussion on preparing design drawings, plans, and specifications. 7.1 STREAMBANK ZONES As discussed in

SECTION

OHW

varies

OVERBANKZONE

BANKZONE

TOEZONE

EXISTINGCHANNEL BED

varies

Figure 7.1 The toe and bank zones.

Rock toe key

While there are many effective means ofbank erosion control available, not all techniqueswork equally well in every situation. Many inef-fective techniques are used as quick solutions tolong-term problems. To choose the best solution,a match must be made between the objectives ofthe project, existing site conditions, possible tech-niques, and fish and wildlife habitat concerns.

Only after the cause of the failure has beenclearly defined should a bank stabilization plan beprepared. Bank stabilization projects fall in twobroad categories: those that correct the problemand those that compensate for it. Even though themost effective way to stabilize a bank is to elimi-nate the cause of the instability, measures to com-pensate for a problem are often used in addition toor instead of correcting the fundamental cause. Itis vital to ensure that the proposed project solves orhelps solve the problem before proceeding withthe project.

As mentioned previously, because of the com-plexity of most bank failures, integrated, interdis-ciplinary, effective teamwork is required at allproject stages. Knowledge of many aspects ofriverine environments is essential if bank stabili-zation projects are to be successful. Again, it isstrongly recommended that a team approach beused when developing or reviewing possible bankstabilization projects. The nature of the projectwill likely dictate the most suitable qualificationsor experience required of the team. At a minimum,the team consisting of an engineer with experiencein river systems, an ecologist knowledgeable infisheries and riparian ecology, and a soil scientistwill generate the most successful projects. Someprojects may require the specialized skills of ageomorphologist, botanist, or landscape architect.

Although the streambank zones above andbelow the ordinary high water mark are treatedseparately in these guidelines for organizationalreasons, it is important that the entire bank beconsidered as a single entity. Toe protection andvegetative components must be incorporated into

a single design with an appropriate transition attheir common boundary. It is also important thatthe geometry and hydraulic characteristics of thestream channel in all three perspectives (cross-section, plan view, and profile) be fully examinedand investigated. This understanding of the streamis essential to achieve a successful integration ofthe project with the natural channel.

This chapter describes basic design consider-ations and criteria for rock, vegetated, and inte-grated (i.e., vegetation, soil, and rock) methods forbank stabilization. Also included in this chapterare suggested habitat components for these meth-ods, and a brief discussion on preparing designdrawings, plans, and specifications.

7.1 STREAMBANK ZONES

As discussed in Chapter 3, streambanks can bedivided into three zones: the toe zone, bank zone,and overbank areas (Figure 3.1 and Figure 7.1).

CHAPTER 7

DESIGN GUIDELINES

Design Guidelines 7-1

Figure 7.1 A bank stabilization project with a rocktoe key.

This section summarizes the characteristics ofthese three zones, focusing particularly on thedesign implications for each zone.

7.1.1 TOE ZONE

The toe zone, which is the area of bank belowthe ordinary high water mark (OHWM), is usuallyinundated and subject to toe erosion and undercut-ting of the bank. Because of the harsh conditionsin this, woody vegetation generally does not growhere; as such, bank stabilization methods that relyprimarily on vegetation are not particularly effec-tive. Methods that are commonly used to stabilizethis zone are rock toe keys, cribwalls, and largewoody debris.

7.1.2 BANK AND OVERBANK ZONES

The bank zone, which is between the OHWMand the top of the bank, is inundated during periodsof moderate (i.e., up to bankfull) flows and ex-posed to periodic erosive currents and debris move-ment. Woody and herbaceous vegetation growwell here. All three bank stabilization methodsmentioned above (rock, vegetative, and integrated)may be used in this zone.

The overbank zone is the area landward of thetop of bank which is subjected to occasional inun-dation during flood flows. Important consider-ations in this zone, where riparian vegetation tran-sitions into upland areas, are wildlife habitat andaccess for project construction and long-term main-tenance.

Most stream channels have complex crosssections. Often there are one or more small chan-nels that concentrate flow during low flow periodsand a larger channel in which flows are confinedmost of the time. Low flow channels are oftenflanked by one or more sand or gravel bars thatmay lack permanent vegetation. Active channelsare generally flanked by sedimentary berms orerosional scarps covered with perennial vegeta-tion. Compound, multi-sloped banks tend to bemore stable than simple, single-sloped banks be-

Design Guidelines7-2

cause berms reduce effective bank height andprovide extra toe support for the upper bank.

7.2. DESIGN OPTIONS AND CRITERIAFOR DIFFERENT METHODS

Bank stabilization methods can be categorizedinto three fundamental types: rock, vegetative,and integrated. Rock methods are those that relyon riprap and/or large boulders to armor the toeand sometimes the bank, or redirect erosive flows.Vegetative methods are those that use plants orplant cuttings to stabilize the bank. Integratedmethods are those that incorporate various materi-als (rock, timber, soil, and plants). In combinationwith these materials, integrated methods may alsoinclude fabrics such as jute or coir mesh.

To help designers in selecting solutions appro-priate for each situation, the following discussionprovides basic descriptions of each method andgeneral selection criteria. Installation procedures,including relative quantities of material requiredand construction techniques, are discussed in Chap-ter 8.

7.2.1 GENERAL DESIGNCONSIDERATIONS

There are many factors to consider when se-lecting a design option. Among these factors arethe stream characteristics (cross-sectional dimen-sions, flow depth, velocity [both magnitude anddirection] and slope of bed or bankline beingprotected). Construction techniques and methodsto minimize adverse impacts to the riparian envi-ronment should also be considered.

Location of the Structure. Most King Countylevees and revetments were constructed alongnatural channel banks to convert as much of thefloodplain as possible for other uses. Recently,recognizing the benefits of floodplain conveyanceand storage, and the drawbacks inherent in en-croachment on the channel, this policy has changed.(See discussion of King County Sensitive AreaOrdinance in Chapter 5). Current practice, when-ever possible, is to set back at least the upper bank

of any new facility from the main channel. The toesection can be built at the location of the existingbank, with a bench constructed at the ordinaryhigh water line, and the upper bank set back.Figure 7.2 illustrates a setback levee with a veg-etated bench. In time, vegetation planted on thebench will extend out over the river to provideshade and cover along the stream margin for fish.

Bank Sloping. Most methods of streambankprotection will require some bank regrading. Steepor undercut banks may require regrading the slopeto 2H:1V or flatter. Because of their unconsoli-dated nature, streambanks with sandy soils mayrequire slopes of 4H:1V or flatter. The applicationof methods that require extensive bank slopingmay be limited by the close proximity of structures(i.e., buildings, roads, utilities), loss of vegetationof significant size (i.e., large trees), land acquisi-


EXISTING LEVEETO BE REMOVED

Existingflood stage

Flood stage aftersetback levee construction

OHW

Flood stage reduction

��

PROPOSEDSETBACK LEVEE

CHANNELBED

BENCH (vegetation on

bench not shown)

tion or easements. In these situations, a rock wall,live cribwall or vegetated geogrid could be used tocreate a steeper slope.

Design Flow. Because structure design is basedon flood velocities and depths, it is necessary toselect one or more design flows to analyze thehydraulics of the reach and find the values of thenecessary variables. A range of flows, up to andincluding the 100-year event, should be examineddepending on the site characteristics, project com-plexity and its associated risks. Of particular inter-est is the bankfull or overtopping event for thestructure in question; this event generates thegreatest velocities and tractive forces.

Design Velocities. Local water velocities (i.e.,velocities at or near the area of erosion), notaverage channel velocities, should be used fordesign. Local velocities along the outside of bends,

Figure 7.2 Setback levee.

for example, can be as much as 50 percent greaterthan the average velocity at that cross-section(Maynord et al. 1989). Analytical methods forestimating velocities in curved channels and/orengineering judgments are used for predicting theeffects of the outside of bends and other hydraulicfactors on the local velocities. Occasionally, thedesigner will be faced with placing protectionalong a straight channel reach. In these cases, thelocal velocity is often less than the average veloc-ity. Methods for estimating local velocities arediscussed further in Appendix C.

Extent of Protection. Many designers mistak-enly extend erosion protection too far upstreamand not far enough downstream, particularly forstructures on the outer banks of bends (Figure 7.3).The highest velocities generally occur at the down-stream end of the bend, and on the outer bank of thestraight section immediately downstream. Often,the erosion potential does not decrease apprecia-bly until the channel straightens and the thalwegcrosses over to the opposite bank for the next bend.The downstream movement of meander bendsshould also be considered. If not properly ori-

susceptible locations up- or downstream of theproject area. The upstream end of the facility inparticular must withstand the greatest forces. Ingeneral, structures should be continued to an areaof reduced velocity. If not, erosion may removebank materials from behind the face of the struc-ture. This greatly weakens the facility and cancause possible failure.

Another method of tieing in is to construct arock deflector at each end of the facility. Thisdeflector acts as a hardpoint that deflects flowsaway from more vulnerable point along the bank.This is particularly valuable at the downstreamend of structures such as rock revetments.

7.2.2 ROCK PROTECTION METHODS

Rock protection methods include toe keys,deflectors, and revetments. These methods arecommonly used where bank materials are weakand water velocities are high.

Rock Toe Key

At sites where toe erosion has been identifiedas the mode of bank failure, stabilization struc-tures should be keyed into the channel bed at thebank toe. While this may be obvious where toeerosion is the major problem, all alluvial streamsscour during flood events unless the bed is ar-mored with large material. If the stream is under-going bed erosion, whether by general degrada-tion or headcutting, structures must be protectedagainst undercutting.

Rivers with highly mobile beds (i.e., largefluctuations in scour depth) may require deep toekey placement. Recent feasibility studies on theTolt River and South Fork Snoqualmie Rivers, forexample, included a recommendation to place thetoe key a minimum of three feet below the lowestrecorded thalweg elevation (Shannon and Wilson1993a; 1993b). Both of these preliminary recom-mendations require detailed scour analyses to de-velop the final toe key design.

For large river environments, Lagasse et al.(1991) recommends placing the riprap a minimum


Figure 7.5 Extent of protection required at a channel bed. (Adapted from Lagasse, et al, 1991.)

1.5 W

Flow

Tangent at midpointon the bank.

1.0 W

W

ented, the structure can deflect flows and createerosion problems on the opposite bank.

Tie In. It is important that the end points of thefacility be tied into a stable bank area. Some bankprotection measures such as riprap structures cre-ate “hardpoints” that can cause erosion at more

Figure 7.3 A schematic of the minimum extent ofprotection required at a channel bend.(Adapted from Lagasse et al. 1991.)

of five feet below the original streambed eleva-tion. Alternatively, the potential bed scour can beestimated, and the toe then placed deeper than thepredicted scour depth. Methods for accuratelypredicting scour have been developed byRichardson, Harrison and Davis (1991) andRichardson, Simons and Julien (1990). Althoughthe methodology specifically addresses scour inthe vicinity of road crossings, it is useful in anyevaluation of bed scour.

Toe key dimensions depend on stream charac-teristics, level of protection, and type of structure.The major consideration in designing a toe key isthe proper sizing of the rock. The rock must belarge enough to remain stable under the flowdepths and velocities to which it will be exposed.Typically, rocks will need to have a minimumdimension of two feet or a minimum weight of atleast five hundred pounds. Over-sizing the rockshould generally be avoided because of increasedcost and difficulty of placement.

The toe key can be difficult to construct inrivers with high banks. In these situations, it isextremely difficult to reach down from the top ofthe bank with a dragline to key rock in at the toe.An alternative is to design a bench at ordinary highwater that can be used as a construction platform.The bench can be left as a permanent feature whichthen allows the upper bank revetment to be setback from the main channel of the river.

The width of the toe key is not as critical as itsdepth. For riprap revetments, the minimum widthof the key should be 1.5 to 2 times the thickness ofthe riprap blanket at the base of the slope. If usedwith cribwalls and vegetated geogrids, the toeneed not extend beyond a line formed by extend-ing the slope angle of the structure to the maxi-mum key depth.

Quarried stone is recommended because an-gular rock tends to interlock, which makes it morestable. Using rounded stream rock is discouragedbecause it is less stable. In stream areas where therock has formed an armor layer, its removal bymining operations may cause local scour prob-lems. Irregular rocks should be placed with thelong axis parallel to the flow. Only hard rocks,such as granite or other volcanic rock that will noterode rapidly, should be used.

Large rock may be added to the toe for habitatpurposes if it does not create currents that causeerosion problems. Rocks create habitat by provid-ing refuge from high flow velocities (a form ofcover) and creating scour holes. Rocks are usuallyplaced within the zone of highest flow velocitiesand can be incorporated into the toe of a protectedslope. Rocks used in this fashion are intended tocreate velocity refuges, rather than scour holes.Habitat elements are discussed in greater detail inSection 7.2.5.

As an alternative to using large rock, it ispossible to use smaller stone wrapped in a naturalor synthetic geotextile material. Because abrasivesediments and debris will wear, snag and tear thesefabrics with time, high flows may remove thissmaller rock. This can potentially undermine thestructure and cause it to fail. The geogrid materialshould have high tensile strength and resist corro-sion and abrasion. The diameter of the rock fillused in the wrap must be greater than the size of thegrid openings but should not exceed six to eightinches. If larger stone is used, there should besufficient small rock to fill voids between the largestones so that the fill cannot shift and allow thestructure to settle over time.

Deflectors

Deflectors are structures that are attached toone bank and project into the flow (Figures 7.4 and7.5). Commonly referred to as spurs or spur dikes,deflectors protect erodible banks by directing theflow toward the middle of the channel. They areuseful in reducing meander migration and watervelocities near the bank. The scour holes that formaround deflectors can provide rearing pools andcover for fish.

The design variables most used for deflectordesign are: orientation angle, effective length,crest height, placement site, construction material,spacing between multiple deflectors, and deflec-tor construction materials. The following is a listof recommendations and options, not strict rulesfor designing deflectors (Conner 1991):

Orientation Angle. Deflectors oriented up-stream create larger and deeper scour holes than



FLOW

Figure 7.4 Downstream oriented rock deflector keyed into a streambank.

perpendicular or downstream oriented deflectors(Klingeman et al. 1984). Deflectors oriented up-stream may also be the most unstable. The eddiesthat form on the upstream face of the deflector mayscour a longitudinal hole along the bed, undermin-ing the structure and causing it to roll forward(Owusu and Klingeman 1984). The eddy formedin the pocket between the upstream oriented de-flector and the bank protect the bank from highervelocities. If the eddy velocity is sufficient totransport local bank materials, it will scour thebank and undermine the structure (Copeland 1983).

Deflectors oriented downstream direct the flowaway from the bank along the deflector (Klingemanet al. 1984). Because the flow deflection angleapproximates the orientation angle, the designercan predict where the flow may impinge on theopposite bank (Klingeman et al. 1984; Reeves andRoelofs 1982). These areas may be protected byriprap, vegetation, or by placing another deflectorto intercept the flow. The downstream orientationcauses less flow deflection, and therefore, little orno scour of the opposite bank (Owusu andKlingeman 1984).

A downstream orientation is recommended atsites where bed and bank stability may be a prob-lem. Additionally, debris and ice are less likely toaccumulate on downstream oriented deflectors(Klingeman et al. 1984). For these reasons, down-stream orientation of these structures is generallyrecommended (Federal Highway Administration1979; British Columbia Ministry of the Environ-ment 1980; Seehorn 1985; Wesche 1985).

Perpendicular deflectors may be the most cost-effective bank protection because the length ofbank protected is directly correlated with the ef-fective deflector length (Copeland 1983). Becauseperpendicular deflectors intercept flow at an abruptangle, they may also be more inclined to fail.Special care should be taken in the design stage toprevent failure of single perpendicular deflectors.The perpendicular design is often used in combi-nation with multiple deflectors to protect a lengthof bank.

Effective Length. The greater the channel con-striction caused by the deflector, the greater thevelocity at the tip of the deflector and the greaterthe scouring potential of the flow. The channel

constriction or solidity is represented by the effec-tive length of the deflector (Le) compared with thechannel width (W).

The deflector must be long enough to deflectthe flow away from the length of bank to beprotected, unless riprap or another secondary bankprotection is used. Several shorter deflectors mayalso be used to protect the same length ofstreambank as one long deflector. Miller and Kerr(1984) found that a deflector could protect thedownstream bank for 2 to 5.5 times its effectivelength, depending on the expansion angle of theflow. Severe channel constriction may cause a


��Crest height

��

Crest slope

SECTION A-A'

SECTION B-B'

��

W

LLe

��

Flow

A

A'

B'

PLAN VIEW

B

Orientation angle

Spacing

Figure 7.5 Schematic diagram of a deflector.(Adapted from Conner 1991.)

sharper expansion angle, and thus decrease thelength of bank protected per length of deflector(Klingeman et al. 1984). Miller and Kerr (1984)found in flume studies that the optimum effectivelength for bank protection was 0.2 of the flumewidth.

Recommendations for effective length in theliterature range from Le/W of 0.25 to 0.8 (Seehorn1985; Wesche 1985; Crispen 1988). Althoughthese authors claim that deflectors may block asmuch as 60 to 80 percent of the flow area, thesedeflectors would likely create adverse effects andfail. Thus, deflectors that block significant portionof the flow area are rarely practical.

Deflectors can create effective fish habitat byproducing scour holes. To create scour holes thatbenefit fish, deflectors must be long enough tointercept a substantial portion of the flow. Garde etal. (1961) found that in straight reaches, the depthof scour was 0.2 to 0.5 the effective length of thedeflector. Lagasse et al. (1991) provide criteria forpredicting scour depth at deflectors.

Unless desired, the deflector should not be solong that it directs the flow into an erodible oppo-site bank. The opposite bank may be protected ora deflector may be placed downstream on theopposite bank to intercept the flow. If the deflectoris submerged at high flows, it may extend across agreater portion of the channel without causingerosion of the opposite bank.

Crest Height. If the risk of flood damage toadjacent roads is of concern, deflectors should besubmerged at high flows so that they do not catchdebris (Federal Highway Administration 1979;British Columbia Ministry of the Environment1980). Seehorn (1985) suggests that this conditionwill be met if the deflector is no more than 6 to 18inches above average summer low flow level.Deflectors should be no higher than the top of bankand slope downward to the tip to prevent under-mining (Franco 1967). Deflectors with a slopingcrest have to be longer than level crested deflectorsto achieve the same amount of bank protection andbed scour (Klingeman et al. 1984). To maximizedeposition, a series of deflectors can be arrangedso that the crest of each deflector is lower than theone just upstream (Franco 1967).

Deflectors that are designed to be overtoppedat high flows should be shaped so that the flow isnot directed into erodible banks at high flows(Federal Highway Administration 1979). An up-stream oriented deflector or triangular shaped de-flector will shunt high flows toward mid-channeland cause deposition along the bank.

Spacing of Multiple Deflectors. A series ofdeflectors can be used to protect a length of erod-ing bank. The goal is to redirect the thalweg awayfrom the eroding bank. The deflectors should bespaced so that the flow expanding downstream ofone deflector is intercepted by the next and redi-rected toward the opposite bank. To determinedeflector spacing on the outside bank of bends,Miller and Kerr (1984) suggest projecting thetangent of the thalweg from the tip of the upstreamdeflector to the bank downstream. The down-stream deflector should be designed to interceptthis flow. They suggest that the spacing be reducedby 20 percent on sharp bends. The thalweg of thestream straightens at high flows and thus willimpinge on the bank in the downstream portion ofa concave bend (Miller and Kerr 1984). Deflectorsshould therefore be placed closer together in thedownstream portion of a concave bend to protectthe bank at high flows (Reeves and Roelofs 1982).

For the most effective bank protection, deflec-tors should be spaced close enough so that flowwill circulate between the deflectors, creating abuffer zone of eddies that protect the bank fromhigher velocity flow (Copeland 1983). Klingemanet al. (1984) found that the optimum spacing fordeveloping this protective eddy system varies be-tween three and four times the effective deflectorlength and decreases as deflector length increasesbeyond 0.2 of the channel width. Spacing deflec-tors too close reduces sediment deposition be-tween structures (Crispen 1988).

Arrangements of multiple deflectors for vari-ous purposes have been recommended in the lit-erature. Spacing deflectors such that their flowpatterns interact creates more scour and diversityof habitat than they do individually (Heiner 1989).Pairing deflectors opposite each other centers thethalweg and creates a long, deep plunge pool(Federal Highway Administration 1979). Alter-nating deflectors can be used to help re-establish


or create a meander pattern (Wesche 1985; Crispen1988).

Deflector Construction Materials. The mostcommon construction materials used for deflec-tors are rocks and logs. Rock deflectors can beconstructed of two to three rows of interlockingrocks extending out into the flow (Federal High-way Administration 1979; British Columbia Min-istry of the Environment 1980). The largest rockshould be placed at the tip of the deflector as thisis the zone where maximum velocity occurs. Rocksize can be less than that recommended for fishrocksas discussed in Section 7.2.5.

Blunt, wedge-shaped rock deflectors shouldbe used for bank stabilization. These create lessflow disturbance and therefore are less likely tocause scouring of either the bed or banks. Singlelarge boulders, when properly placed, may act asflow deflectors (Oregon Chapter of the AmericanFisheries Society 1988).

Rock deflectors should be embedded in thebed and banks to prevent undermining. The depththat the boulders should be embedded dependsupon how much scour is expected around the baseand root of the structure. Orsborn and Bumstead(1986) recommended rock deflectors be embed-ded for a distance equal to the height of thedeflector.

Rock Revetments

A carefully placed layer of angular rock, gen-erally known as riprap, is a common and effectivemethod of bank protection used on levees andrevetments. While rock offers some resistanceagainst mass-movement, its primary purpose is toprevent loss of bank material by fluvial erosion.Because the system is flexible, riprap can settleand conform to the final streambed contour ifscour occurs. Over time, vegetation may becomeestablished in riprap above the waterline.

Revetments have typically been constructedfrom hand-placed, dumped, or derrick-placed rock.Many types of structural facings have been used.These include: riprap; gabion mattresses; rubbertire networks; articulated, precast concrete blocks;and cellular grids. Because of environmental and

aesthetic concerns, most of these methods havenot gained acceptance in King County. Becausethe most common revetment in King County isrock, the remainder of this discussion will focus onthis material.

Riprap revetments are particularly effective inthe following situations: 1) sharp bends; 2) con-strictions, such as bridges, where velocities in-crease; 3) along the opposite bank at the confluenceof two rivers; and 4) on rivers where debris dam-age may occur.

Limitations. Rock revetments have severallimitations. These include environmental impactssuch as the destruction of fish and wildlife habitat,encroachment into the floodplain, and loss ofaesthetic values. Rock should not be prescribedwithout first carefully considering other alterna-tives. Even where rock is absolutely necessary, anattempt should be made to incorporate vegetation;the structure should be sufficiently set back fromthe channel to enhance rather than degrade ripar-ian environments and instream habitat.

Other factors that may limit the use of rockinclude the availability of suitable-size rocks, thedifficulty and expense of quarrying, transporting,and placing stone, and the large amount of mate-rial needed for deeper streams. While small riprapmay be hand-placed, most is end-dumped or placedby derrick crane or other large equipment.

Gradation. For riprap to function properly, itis essential that it be well-graded. A reasonablegradation will allow the various rock sizes tointerlock and minimize voids in the structure. It isessential that there be no significant gaps (missingsizes) in the gradation. Gaps in the gradationincrease the chance of structural failure if highflows remove smaller rock particles, causing largerparticles to settle. Wittler and Abt (1990) foundthat a relatively uniform gradation can withstandgreater erosive forces. Failure of these rock facili-ties, when it occurs, can be more rapid than thosefacilities having a broader gradation.

Because many standards have been devel-oped, the riprap gradations--along with the me-dian stone size--should be specified in the designplans. The standard gradations generally used inKing County have been those developed by theWashington State Department of Transportation



(DOT). These range in size from quarry spalls,through light loose riprap, to heavy loose riprap.The specific sizes of this material, and the methodsfor computing the size appropriate for a specificsite, are described in Appendix C.

Filter Layer. Most riprap is placed on a filterblanket of smaller sized, graded material. A properfilter layer prevents the loss of finer soil particlesof the bank through the interstices of the ripraplayer. If these finer soil particles are lost, slumpingand failure may occur. The area to be covered witha filter blanket should be reasonably smooth. Aneven thickness of filter material should be placedon the prepared surface. Riprap should be placedcarefully to ensure that the blanket is not rupturedor displaced. For most of the rivers in King County,a filter layer of gravels or quarry spalls is recom-mended. Relationships between sizes of riprapand gradations of adjacent layers have been devel-oped to size the individual rocks in the filter layer.These relationships are discussed in detail in Ap-pendix C.

Geotextile fabrics have sometimes been usedto create this filter. It can be more difficult, how-ever, to key a large rock blanket into fabric thaninto a blanket of smaller rock. Fabric filters aremost useful when the banks consist of fine-grainedalluvium. Banks with extremely fine-grained soilssuch as silt or clay may require both a geotextilefabric and a rock filter.

Bank Slope. Because steep slopes lessen thestability of the total structure, rock should not beplaced on slopes steeper than 2H:1V. Steeperslopes may require a retaining wall or other struc-ture. Maynord et al. (1989) state that stability testshave shown that slope has small effect on riprapstability when side slopes are flatter than 2H:1V.Because high flows can saturate river banks (cre-ating failure in the underlying material), it is vitalthat the revetment face slope does not exceed theangle of repose of the underlying layer.

Because riprap is usually installed at sites ofsevere erosion where the existing side slopes areoften steeper than 2H:1V, substantial site gradingis often required. When needed, the bank slopesshould be laid back away from the channel wherepossible to obtain the appropriate slope angle.Encroachment into the channel is not recom-

mended. Encroachment that can not be avoidedwill require an exemption from the floodplainregulations contained in the King County Sensi-tive Area Ordinance. Grading of the slopes mayincrease right-of-way requirements. It is also ex-tremely difficult in situations where existing build-ings, other structures, or large mature, vegetationare located near the existing top of bank.

Shape of Rock. As noted in the toe key discus-sion earlier, angular rock is preferred to roundedrock because the stones fit together to provide amore solid blanket. Because this strengthens therevetment so that it acts like one structure ratherthan a collection of independent stones, it raisesthe threshold velocity for incipient motion andsubsequent failure. Quarry rock is preferred tonatural river rock as it is generally angular; riverrock is usually rounded and unacceptable as riprap.While the rock should be angular, ideally it shouldbe as nearly rectangular as possible. The ratio ofthe longest to the shortest dimension should be nomore than 3.5:1 (USACOE 1991).

Toe key. As discussed earlier, lack of a suffi-cient toe key is a common cause of bank and riprapfailures. Because tractive forces are greatest in thiszone, a well-constructed toe is essential. If the toeis not sufficiently deep and effectively keyed intothe streambed below the anticipated scour line, theentire structure may be undermined.

Height of Riprap Face. In determining theheight of the riprap face, a factor of safety (relatedto water surface elevation) should be incorporatedinto the design. For installations that are com-prised of riprap only (as opposed to vegetative orintegrated methods), Lagasse et al. (1991) recom-mends extending the riprap to a minimum of twofeet above the design water surface elevation.Riprap should extend up the bank far enough toprovide adequate protection against scour by de-bris, flowing water, or wave action.

In situations where the natural channel hasbeen constricted, the designer will often find thatthe capacity of the channel is insufficient to con-vey the design storm. In these cases, either rockprotection should be provided to the top of thebank or construction a setback levee should beconsidered. If sufficient space is available, a set-back levee is the preferred alternative.


When flood containment is a project objective,the design water surface elevation should be ad-justed to account for the superelevation resultingfrom centrifugal forces at bends. Chow (1959)provides methods for computing superelevation.Freeboard should be added to this superelevationestimate. The Corps and FEMA both may requirethree feet of freeboard above the 100-year watersurface as a factor of safety for levees (see Section5.8 for further discussion).

Vegetation. Vegetation on the face of the riprapstructure can be an important component of bankstability. In the past, maintenance of riprap struc-tures often involved periodic removal of all veg-etation under the assumption that this would im-prove access and visibility for inspecting facili-ties, and that large vegetation, if uprooted, couldseverely damage the riprap face.

Recently, regulatory agencies, such as theWashington Departments of Ecology, Fisheries,and Wildlife, have required the incorporation of

vegetation in levee and revetment faces. For this tobe successful, the vegetation must come into con-tact with the soil underneath or within the ripraparmor. Depending on the season, irrigation mayalso be necessary.

Other Rock Structures

Many other types of rock structures have beenused successfully for bank protection. These in-clude turning rocks, tie-backs, and rock-filltrenches. Because a detailed discussion of thesestructures is beyond the scope of this document,they will be described only in general terms. Thereader is referred to Orsborn and Bumstead (1986)and Lagasse et al. (1991) for further information.

Turning Rocks. Turning rocks are rows ofboulders placed in a bend starting at the upstreamoutside bank and angled toward the inside bank toreduce erosion (Figure 7.6). Turning rocks reduce

Figure 7.8 Turning Rocks

Flow

2/3 - 3/4 flow

1/4 - 1/3 flow

L

W

Detail of turning rock

Flow

Turning rocks

Additionalrows of rocks,as needed

Erodingbank PLAN VIEW

Pointbar

Figure 7.6 Turning rocks used to reduce erosion on the outside of a bend.

the spiral currents that erode the outside bank of abend. They also help dissipate energy to reducestream power locally and provide cover for fish.

Turning rocks are placed in a downstreamdiagonal across the stream to deflect the flow.Each successive rock divides the flow and deflectsthe majority in the desired direction. The next rockdownstream is placed to intercept the deflectedflow from the upstream rock and so on. Thelongest axis of each rock should be positioned at aslight angle to the flow (Orsborn and Bumstead1986).

Turning rocks can be used to direct the flowaway from an unstable bank or to direct the thal-weg down a selected part of the channel. Severalrows may be needed depending on the length andradius of the curve (Orsborn and Bumstead 1986).Because turning rocks may not be adequate to turnthe flow by themselves, especially in deep streamsor rivers, other structures such as deflectors maybe needed to provide adequate protection fromerosion.

Tie-Backs. Tie-backs are individual sectionsof riprap or other structural protection placedperpendicular into an eroding bank to preventflanking by floodwaters (Figure 7.7) . Dependingon the design, they are placed against the bank, and

either lay flush against it or protrude slightly intothe stream channel. Tie-backs protruding into thechannel create hardpoints that provide some en-ergy dissipation. In this way, protection fromerosion is provided without hardening the entirestreambank. Because much of the existing bankline is left undisturbed, a favorable environmentfor the establishment of native vegetation remains.Tie-back revetments are created by connecting thehardpoints with a rock revetment or toe key.

These structures are not effective for extremelyhigh velocity flows. They are most useful in rela-tively straight reaches where the primary erosionthreat is a meandering thalweg. On the outside ofbends, a revetment or a series of rock deflectors isusually more appropriate.

Rock-filled Trenches. A rock-filled trench isplaced parallel to the bankline such that the rockcan fill scour holes and/or scalloped banks aserosion progresses. The trench is dug behind thebank of the channel and filled in with riprap. Thetrench is then covered with a layer of soil andreplanted. This method does not modify the chan-nel and yet provided the riprap trench will halterosion if it occurs. This method provides addi-tional protection when greater security is required.


Figure 7.10 Tie-back trench to prevent flanking. (Adapted from Richardson et al, 1991.)

FLOW

Tie-back trenchfilled with riprap

Riprap

Figure 7.7 Tie-back trench and revetment to prevent flanking. (Adapted from Richardson et al. 1991.)


Another application of a rock-filled trench isto construct it within the channel itself, immedi-ately adjacent to the toe of the bank. As bed scouroccurs, the rock settles in to the degrading toe.Protection from undermining is provided as longas the eventual depth of scour does not exceed thecapacity of the quantity of rock used. In rapidlyeroding river environments (outside of bends,etc.), a keyed-in toe is preferable to a rock-filledtrench type.

7.2.3 VEGETATIVE METHODS

Vegetative methods include herbaceous groundcovers, rooted stock, live stakes, fascines, brushmattresses, and brush layers. While the root sys-tems of these components increase the “structuralintegrity” of a bank with time, their initial value isin protecting the bank surface. These methodsusually can be installed with minimal instreamdisturbance.

Evaluation factors for selecting the appropri-ate plant species and method of application in-clude slope, aspect, soil characteristics, drainage,elevation and tolerance of the plant species toinundation. Much of this is discussed in Chapter 6.Ideally, the selection of vegetation should be re-stricted to native species that are suited to the siteconditions.

Plants should be chosen based on their adapt-ability and tolerance to soil moisture levels, espe-cially on very wet or very droughty sites. The plantassociations in Table 6.4 include many speciessuited to particular conditions. Planting plansshould be designed using subsets of this list orother species as appropriate, depending on site-specific conditions and stock availability.

Herbaceous Ground Cover

Herbaceous ground covers include grasses andother non-woody vegetation. Although they lacksome benefits of woody vegetation (e.g. cover forfish), herbaceous vegetation is useful in somesituations. Ground cover provides temporary ero-

sion control until woody vegetation becomes es-tablished or where cover on bare ground or soilimprovement is desired. Sod-forming grasses andlegumes, especially if left unmowed, can protectbanks of small streams where flow velocities arelow.

Grass species recommended for western Wash-ington streambanks are listed in Table 7.1. Al-though individual species are listed in this table, amixture of species may be more successful anddesirable than a monoculture. Erosion controlseed mixtures are commercially available, and canbe tailored for site conditions. Seed mixtures shouldinclude annual and perennial species and speciesthat will enrich the soil (e.g., legumes). The needfor fertilizers should be evaluated and appropriatekinds and amounts applied. Local Soil Conserva-tion Service personnel are a valuable source ofinformation about specific plant requirements.Wasser (1982) provides an excellent summary ofgrasses, forbs, shrubs, and trees useful for reveg-etation projects in western states.

The use of grass and forb turfs to protectstreambanks is constrained by velocities of designflows. At present, information on design veloci-ties for native grasses is not readily available.Examples of maximum allowable design veloci-ties for other selected grasses are listed in Table7.2.

While unmowed turf also can provide habitatfor small mammals and ground dwelling birds,thick turf or grasses may hinder the establishmentof woody vegetation by competing for water andnutrients. Thick turf may also encourage popula-tions of small rodents that girdle trees and shrubswhen feeding on bark.

Rooted Stock

Rooted stock is any tree, woody shrub, orherbaceous plant with established roots. This in-cludes rooted cuttings, balled and burlapped, bare-root, and containerized plants. This material isused either alone or with other methods to provide

PERMISSIBLE VELOCITY (fps)Type of Grass Slope Erosion Easily

Range (%) Resistant Soils Eroded Soils

Bermuda grass Cynodon spp. 0-5 8 65-10 7 5

over 10 6 4

Buffalo grass Buchloe spp. 0-5 7 5Smooth brome Bromus inermis 5-10 6 4

Blue grama grass Bouteloua gracilis over 10 5 3

Alfalfa Medicago sativa 0-5 3.5 2.5


Recommended Seeding Rates for Given Site Conditions1

Species Shallow or Dryland Irrigated or PalustrineDroughty Sub-irrigate2 orWetland

tall fescue Festuca arundinacea 18 18 18

creeping red fescue F. rubra 8 8 8

sheep fescue F. ovina 5 5

bentgrass Agrostis spp. 1 1

perennial ryegrass Lolium perenne 15 15

tiny white clover Melilotus alba 10-15

big trefoil Lotus crassifolius 4 4

white or Dutch clover Trifolium repens 2 2

red clover T. pratense 4 4

sicklekeeled lupine Lupinus albicaulis 10 10 10

1. Minimum seeding rates in lbs/acre.2. Includes sites receiving extra moisture from runoff, snowmelt, stream water, etc.

Table 7.1 Grasses and ground covers recommended for use on and adjacent to channel banks in westernWashington. (Adapted from SCS 1986.)

Table 7.2 Example maximum allowable design velocities for channels vegetated with selected grasses.(From Simons, Li, and Associates 1982.)


leafy cover and root strength because it sends rootsinto the surrounding soil in weeks rather thanmonths that cuttings may take. It may be placedanywhere on the bank where it will not be removedby erosive flows.

Rooted stock should be used for planting dur-ing the growing season when unrooted cuttingsmay not survive. It is also useful where soils aredroughty, nutrient poor, where rooting of cuttingsis doubtful or when cuttings of desired species areunavailable. Species that do not root readily fromcuttings such as conifers can also be incorporatedinto designs in this manner. Rooted plants may beadded where understory vegetation already existsand larger shade-providing plants are desired.Rooted stock provides immediate vegetative coverand habitat improvement.

Spacing of rooted stock is dependent on theeventual size of selected species. Depending onthe root distribution needed, plants may be spreadevenly across the site for uniform cover or clumpedfor a more natural appearance. The plants vary insize from small (inches) to large (10 or 12 feet tall).Containerized stock has a relatively high cost perplant. Even with established roots, rooted stock at

some sites may require irrigation for one or moreseasons.

Live Stakes and Slips

A quick and effective means of securing avegetative cover for control of soil erosion andshallow sliding is planting unrooted cuttings. Livestakes are woody plant cuttings that can root, andare large and long enough to be tamped into theground as stakes (Figure 7.8). Live stakes aregenerally cut from wood that is two or more yearsold. Slips are similar to live stakes, but smaller insize. Slips, which are cut from first or second yearwood that is still soft and flexible, are not strongenough to act as anchors.

Cuttings from plant species that root easilywill grow if planted under favorable conditions. Itis well known, for example, that most willows,many poplars, and cottonwoods, readily growfrom cuttings set in moist soil (Gray and Leiser1982). Even in very unfavorable conditions (e.g.,deep shade), live stakes will often grow vigor-ously for a few years before they die out. During

Figure 7.2 Live stakes.

Live stake priorto installation

Note: Shown after one growing season.

Figure 7.8 Live stakes.

Figure 7.3 Fascines.

Fascine or wattle bundle

varies

Live or dead stake

Shallow trenchwith fascine

AFTER TOPSOILCOVERING

AT TIME OFINSTALLATION

this time, they will stabilize and modify the soil,and serve as pioneer plants until other plants canbecome established. Live stakes are effective incamouflaging an open area after one or two grow-ing seasons. Over time, the cover provided by livestaking creates riparian and wildlife habitat.

Live stakes are useful alone or when used withstraw, jute mesh, and coir (a coconut fiber mesh)for providing surface protection and in controllingsmall rills and gullies. They are also effectivewhen construction time is limited, an inexpensivemethod is necessary, the problem is very simple,or when work in the channel is not allowed ordesirable. Slips are useful for small projects withsimilar characteristics and sites with soft, moistsoil. While live stakes and slips require moist soilto root, excessive water will result in rotting.

The density of the installation ranges from twoto four live stakes per square yard. Live stakesshould be spaced approximately every two feet ina random to triangular pattern. For slips, higherdensity (about 12 cuttings per square yard) at onefoot spacing is recommended. Live staking re-quires a moderate to large volume of live plant

material. Live stakes can be interplanted withrooted stock.

Fascines

Fascines are sausage-like bundles of long, livecuttings tied together and secured to the bank withlive and dead stakes (Figure 7.9). They are placedon the bank face in shallow trenches and lightlycovered with soil. These are also called wattles orcontour wattles.

Fascines are useful for areas of general scourwhere the banks can be sloped back. They workparticularly well in straight sections where flowvelocities are low. Fascines offer inexpensive andimmediate protection from erosion, especially fromoverland flows moving downslope. They usuallydo not require work in the channel.

Fascines work well to reduce erosion on shal-low gully sites and help in controlling surfaceerosion by reducing the slope into a series ofsmaller slopes. They are an effective stabilizationtechnique once installed and even more so whenroots become established. Fascines help hold soil


Figure 7.9 Fascines.


on streambank faces by creating mini-dam struc-tures or terraces. The erosion control capabilitiesof this system can be enhanced by using straw, juteor coir mesh to cover surface areas betweenfascines. These materials provide stable growingsurfaces that help the invasion of surroundingriparian vegetation.

The fascines should be spaced three to five feetapart on contour, parallel to the stream (Schiechtl1980). Installation should begin at the wetted edgeof OHW and continue up slope.

Brush Mattresses

A brush mattress is a combination of units thatcover the streambank to provide immediate pro-tection. The units used in this method are live anddead stakes, fascines, and a mattress-like branchcover (Figure 7.10).

Used alone, brush mattresses provide somebank protection and erosion control; they canresist temporary inundation, but not undercutting(Gray and Leiser 1982). Structural measures suchas toe keys or revetments may be necessary if bankundercutting is occurring. Brush mattresses areuseful where banks can be graded to a 3H:1V or4H:1V slope. Construction of these units createssmall disturbance. Because established brush mat-tresses reduce local velocities, they are usefulwhere debris and sediment need to be captured.Brush mattresses provide immediate protectionagainst flowing water and establish a dense naturalriparian zone in one or two seasons. The capabili-ties of this system increase with age.

Brush mattresses are generally used to coversix to ten feet of vertical bank face. This methodrequires a very large amount of live material.

Figure 7.10 Brush mattress with a fascine.

Live branches

Live or dead stake

Wire or jute rope

Fascine(optional)

Figure 7.4 Brush mattress.

Note: Topsoil cover not shown.


Brush Layers

Brush layers are alternating layers of soil andlive branches on successive horizontal rows orcontours in the streambank. The buried portion ofthe branches root to form a permanent reinforcedinstallation, while the tips produce vegetative topgrowth (Figure 7.11).

Brush layers are useful in bank protectionprojects requiring fill, or as a rehabilitation mea-sure for seriously eroded and barren banks (Grayand Leiser 1982). They produce an immediatebarrier that repairs gully erosion and local scourholes. They effectively repair holes in earthenembankments. While the construction of brushlayers generally does not require work in thechannel, earthwork related to the installation ofthe layers may cause some disturbance. Brushlayers, however, rapidly produces habitat coverand a stable vegetated bank.

Installation of brush layers is best during lowflow conditions. This method requires a relativelylarge amount of live branches.

7.2.4 INTEGRATED METHODS

Integrated methods incorporate vegetation,soil, timber and rock. These methods include jointplanting, vegetated geogrids, live cribwall, andtree revetments.

Joint Planting

This method consists of live stakes drivenamong rock riprap. It increases the effectivenessof the armored system by forming a root mat andreinforced filter system in the base upon which theriprap has been placed (Figure 7.12). It also helpscollect sediment and debris. Because joint plant-ing creates no channel disturbance, it is usefulwhere rock work has to be accomplished in thesummer. Once the rock is in place, live staking canbe done later without further disturbance to thechannel.

This method improves areas where riprap isalready in place and habitat, recreational, or aes-

Figure 7.12 Brush layers.

Live brush

AFTER TOPSOILCOVERING

AT TIME OF INSTALLATION

Figure 7.11 Brush layers.


thetic values are desired. It enables a streambankto become more natural looking and function as avegetated riparian zone. In time, roots will add tothe strength of the riprap protection.

The thickness of the existing rock layer is amajor consideration in applying this technique. Toachieve successful rooting, live stakes must bedriven through the rock voids and into the under-lying soil layer. Joint planting is more labor inten-sive than ordinary live staking; it also requires amoderate to large volume of live material. A plantloss of 30 to 50 percent is common with thismethod especially in revetments with very thicklayer of riprap (Schiechtl 1980; Christensen andJacobovitch 1992). Irrigation during the first grow-ing season can enhance plant survival.

Vegetated Geogrid

Vegetated geogrids are similar to brush layersexcept that natural or synthetic geotextile materi-

als are wrapped around each soil lift between thelayers of live branches (Figure 7.13).

Vegetated geogrids are useful where slopescannot be cut back or in bank locations requiringaddition protection against strong erosive flows.The level of protection afforded by geogrids isgreater than solely vegetative methods but may beless than rock methods. Vegetated geogrids areuseful where fill is needed to repair local or gen-eral scour. They may be used to abate bank failurecaused by toe erosion when combined with struc-tural toe protection. If constructed with adequatesoil compaction, geogrids can be constructed witha steep face and thus are valuable for repairs atsites where the banks can not be sloped back.

Vegetated geogrids immediately reinforce thebank. While the benefits are similar to those ofbrush layers, vegetated geogrids can be placed ata steeper angles. Vegetated geogrids capture sedi-ment that rapidly rebuilds and stabilizes the bank.They produce rapid growth for habitat and be-comes very natural in appearance and function.

Figure 7.11 Joint planting

Live stakes

Riprap

Live stake

Figure 7.12 Joint planting.


Figure 7.13 Vegetated geogrid.

Excellent overhanging material is provided im-mediately for aquatic habitat, and cover increasesover time.

Unless rooted stock is used, geogrids are bestinstalled while plants are dormant. At sites such asupper bank areas, irrigation during the first grow-ing season may enhance growth and survival.Plants may be installed during the growing seasonif the plants are watered during the planting pro-cess.

Live Cribwall

A live cribwall is a rectangular framework oflogs or untreated timbers, rock, and woody cut-tings (Figure 7.14). Live cribwalls are useful whenspace is limited and slopes cannot be cut back.They may be installed with finished streamsideslopes as steep as 1H:10V. They are effective inrepairing eroding banks in outside meanders orother areas where the currents are strong and fast

flowing. It is useful for large areas of scour and toabate toe erosion when rock is placed in front ofthe structure. Because this method requires fillmaterial, it is useful for restoring lost banks. If thearea to be stabilized requires a larger-sized andmore complex cribwall, the advice of an engineerknowledgeable in these designs should be con-sulted.

Cribwalls can provide excellent overhangcover material for aquatic habitat. The log ortimber framework provides immediate bank pro-tection, while the plants provide long-term dura-bility.

Cribwalls need not be built to the top of theexisting bank. Other methods, such as fascines orbrush layers, work well on upper banks. Cribwallsshould be built during low flow conditions as theyoften require work in the channel bed. This methodrequires a moderate amount of live material. Regu-lar inspection is necessary the first year to identifyand correct potential washout problems.

Figure 7.13 Vegetated geogrid.

Fill material

Geotextilematerial

Live brush

Exposed face of geotextile material


Figure 7.14 Live cribwall.

Rock fill

Live branches

Tree Revetment

A pervious tree revetment, made from wholetrees cabled together and held in place with rockand deadman anchors buried in the bank, is arelatively inexpensive, semi-permanent form ofprotection (Figure 7.15). Tree revetments are usedwhere protection from bank scour and undercut-ting is needed. Additional protection can be ob-tained by jamming large branches or small treesbehind the cabled trees. The stability of the bankabove the tree revetment can be increased by usingtree and shrub plantings.

Trees with a trunk diameter of 10 to 12 inchesor larger are required for good barriers on largestreams or rivers. Smaller trees (two to four inchtrunk diameter) may be used on smaller streams.The most effective species are those with bushytops and durable wood, such as Douglas fir orwestern red cedar.

Tree revetments have a limited life and mustbe replaced periodically. Loss of trees through

damage or deterioration will expose the bank tothe current. If the revetment is not repaired, thebank will continue to undercut and erode.

Aesthetically, this method is acceptable innatural settings. As it collects sediment and beginsto revegetate, it becomes more natural in appear-ance and function. The rate of silting which occursin the revetment area is dependent on the type andamount of sediment being transported by the streamand the type of trees used.

7.2.5 FISH HABITAT COMPONENTS

Bank stabilization projects and other instreammodifications can alter fish habitat by changinglocal depths and velocities, resulting in local scouror deposition at the stream bed or banks. Fishhabitat may benefit from these changes if theyresult in spawning gravel recruitment and createresting areas in feeding zones. Bank stabilizationprojects may also improve overwintering condi-

7-22 Design Guidelines

Figure 7.15 Tree revetment. (Adapted from Henderson and Shields 1984.)

tions by increasing the available interstitial spaceand reducing water velocities. They can also pro-vide cover either as water depth, overhangs, orvisually isolated areas. The benefit of these effectson fish habitat will depend on whether or not thefish populations are limited by other factors in thestream basin.

The bank stabilization methods discussed ear-lier in this chapter can be improved by addingfeatures designed specifically to benefit fish. Thissection describes two such features: large woodydebris (e.g., trees or rootwads) placed into thebank; and boulders or boulder clusters (sometimescalled fishrocks) placed in the channel.

To successfully match the requirements ofvarious fish species and life stages to a project

design, an expert in these topics should be con-sulted to assist with incorporating habitat compo-nents into bank stabilization projects. The typeand location of habitat components added to ariver should be based on the stream gradient,channel geometry, basin hydrology, intendedfunction(s) of the structure, and available materi-als and site accessibility (Heiner 1989). Streamgradient, channel geometry, structural dimensionsand spacing, and discharge determine the forcesexerted on the habitat component. The project’shabitat functions combined with the available ma-terials and site access determine which habitatcomponent that are feasible. These and other vari-ables are discussed further below.

Deadman anchorTrenched-in cables

Varies

Figure 7.15 Tree revetment. (Adapted from Henderson and Shields , 1984).

Overlapvaries

Anchor rock

Flow

Discussion of and design criteria for full span-ning structures, such as log and rock weirs that arecommonly used for instream habitat modifica-tions, is beyond the scope of these guidelines.While appropriate and needed in some situations,these structures are generally not constructed solelyto achieve bank protection.

Large Woody Debris

Large woody debris is any large piece ofwoody material (generally defined as 0.5 feet indiameter and at least 10 feet long) that intrudes oris imbedded in the stream channel. Woody mate-rials affect local flow velocities, streambed andstreambank stability, and local stream morphol-ogy. Very large tree trunks or roots lodged againstreadily erodible streambanks, for example, cansignificantly increase localized streambank scour.The accumulation and burial of large amounts ofwoody debris in the streambed can dramaticallyincrease the stability of the reach against the mo-bilization of streambed sediments during higherflows. Woody materials lodged at various anglesto the flow can efficiently redirect the current toenhance streambank stability. The presence ofwoody elements benefits fish habitat by greatlyincreases the complexity of the currents, the en-trainment and distribution of sediments, and localstream morphology.

When incorporating woody elements into bankstabilization projects, it is necessary to identify thedesired engineering performance and the desiredhabitat benefits. Each project must be specificallytailored to meet the engineering objectives identi-fied for the reach and the habitat requirements ofthe target species.

When selecting a design, it is very important toconsider the factors that influence the relativepermanence of wood in river systems. These in-clude the type of wood, its size and shape, itsexposure to the forces exerted by moving water,and its resistance to movement because of wedg-ing or embedding with adjacent materials. Woodymaterials can be obtained as cut logs, cut stumpsand rootwads, or tree trunks with roots attached.Each has particular advantages and disadvantages.

Disadvantages are largely associated with the levelof difficulty encountered when attempting to an-chor each element in place.

The longevity of any wood will be greatlyenhanced if it remains fully saturated (i.e., “water-logged”). The maximum decay rate occurs withalternate wetting and drying, or consistently dampcondition, rather than full saturation. Wood variesby species in its durability and decay resistantproperties. Cottonwood and alder, even in thelarge sizes needed for installations along majorrivers, are the most rapidly decaying local treespecies. While maple will also decay fairly quickly,it is more durable than the other deciduous treespecies. It is unlikely that deciduous woods can berelied on to survive for more than 5 or 10 years atbest; water saturated maple may effectively doublethese estimates.

For maximum longevity, it is best to use moreresistant coniferous species whenever possible.Of the conifers, hemlock is poorly suited becauseof its rapid decay rates. While very durable, Sitkaspruce is often very difficult to locate and com-paratively expensive because of its desirable lum-ber qualities and locally increasing scarcity. Dou-glas fir has excellent durability, especially whenmaintained in a saturated condition; it is also themost abundant of the commercially managed soft-woods. Western red cedar, however, is the mostdesirable of all native local species because of itsnatural rot-resistant properties. Douglas fir willgenerally survive for at least 25 to 50 years, withcedar lasting twice this length of time. Such lon-gevity puts these species within the normal esti-mates of the functional design lifetime expectedfor conventional riverbank stabilization installa-tions.

Individual logs or aggregates of woody mate-rial can increase local and/or reach-specific ratesof erosion by deflecting or re-directing flows.Single logs, for example, are frequently placed sothat they extend into the river at a downstreamangle. When placed in relatively shallow flows,the result most often obtained is increased turbu-lence with higher velocities both flowing aroundthe log and redirected into the bank just immedi-ately downstream of the log.


Conversely, placing logs at an upstream anglewill deflect flows at right angles to the log awayfrom the bank and toward the center of the stream(Figure 7.16). Deeper, higher-velocity flows willincrementally scour a pool around and under theend of the log. This effect, when distributed overa series of logs placed along an outer meanderbend, can effectively shift the deeper, faster flowsaway from the toe of the riverbank slope. Reloca-tion of the thalweg of the river, even by a modestdistance, can markedly enhance the longevity andperformance of riverbank stabilization measures.

Boulders/Boulder Clusters (Fishrocks)

Fishrocks are large, irregular boulders used tocreate fish habitat by producing a diversity ofvelocities and depths. Additional cover and rear-ing areas are provided by the deep water, airbubbles, and turbulence around the rocks. In addi-tion, individual boulders or groups of bouldersmay aid bank stabilization efforts by deflectingflows away from unstable banks.

Fishrocks can be used to achieve many differ-ent objectives, depending on where they are placedand how they are arranged (Figure 7.17). Al-though boulder placements are the simplest fishhabitat structures, the hydraulics surrounding themare complex. Careful planning and the use ofhydraulic criteria to choose and place boulderswill enhance their success.

The first step in predicting the results of rockplacements is to analyze the existing flow patternsand streamlines in plan and profile view and thenvisualize how the rock will change them. This canthen aid in estimating the pattern and extent ofscour and deposition.

Clusters of boulders have several advantagesover single boulders. Boulder clusters providegreater stability, have a greater diversity of depthand velocities, and trap woody debris more effi-ciently than single boulders. Additional cover forfish is provided in the spaces between the boul-ders.

Irregularly-shaped, angular boulders of du-rable rock should be used. Abrupt boulder edges


Figure 7.16 Bank protection using large woody debris.

Toe of bank

OHWM Top of bank

Rock toe

Log with roots securedwithin rock toe.

FLOW

Figure 7.16 Bank protection using large woody debris.

ALIGNMENT OF LARGE WOOD IN THIS CONFIGURATION NO LONGER USED DUE TO SAFETY CONCERNS FOR RECREATIONAL RIVER USERS. (APRIL 2009).


PLAN VIEW

PLAN VIEW

Deposit Scour

Suspendssediment

Flow

Flow

Figure 7.17 Boulder clusters. (Adapted from Orsborn et al. 1985.)

7-26 Design Guidelines

Table 7.3 Recommended rock sizes for fishrocks. (Federal Highway Administration 1979.)

Channel Width (ft.) Water Depth (ft.) Length of Rock (ft.)(summer flow)

< 20 1.0 to 2.5 2.0 to 4.020 to 40 1.0 to 3.0 3.0 to 8.040 to 60 1.5 to 4.0 4.0 to 12.0

create turbulent eddies and enhance the scouringpotential of the rock (Cullen 1989). Angular rockis also less likely to roll in the current (Moreau1984). The irregularities will provide additionalhiding cover for fish (Crispen 1988).

The boulders should not be so large that theycause bank erosion or overtopping. Maximumscour occurs when the water level is at the top ofthe rock (Fisher and Klingeman 1984). The Or-egon State Highway Division (1976) recommendsthat no more than one third of the channel area beblocked. The Federal Highway Administration(1979) cautions that no more than one fifth of thesummer low flow area be blocked unless thestream gradient is greater than three percent. Theserecommendations may be conservative and over-simplified in that the stability of the rock and thechannel depends on many more factors than sim-ply flow blockage.

The rocks should be large enough not to bewashed away during high flows. The size of rockrequired depends on stream size, flow characteris-tics, substrate stability, and rock shape. Crispen(1988) suggests that fishrocks should be at least aslarge as rocks naturally maintained in the stream.

The Federal Highway Administration (1979)recommends using 2-foot diameter, 1000-poundrocks in velocities of up to 10 feet per second and4-foot diameter rocks in velocities of 10 to 13 feetper second. In addition, they provide size guide-lines based on channel width and depth (Table7.3). Other suggestions from various sources sug-gest that boulders should be at least 2.0 to 6.5 feetlong or have a volume of greater than 0.65 cubicyards (British Columbia Ministry of the Environ-

ment 1980; Ward and Slaney 1981; Moreau 1984;House and Boehne 1985; Wesche 1985).

Highway agencies, concerned with protectionof the roadway, advise that if potential flooding orbank erosion is of concern, high flows shouldovertop boulders to clear trapped debris (OregonState Highway Division 1976; Federal HighwayAdministration 1979). If the risk of damage toproperty and improvements is low, debris shouldbe left around fishrocks for additional cover.

Boulders with blunt faces create large up-stream scour holes and then tend to tip into thishole (Fisher and Klingeman 1984; Cullen 1989).Rocks should be selected and positioned so thatthis does not occur. Increased stability can beachieved by placing boulders in clusters. Seehorn(1985) recommends the use of boulder clustersover isolated boulders. While the scour patternaround boulder clusters depends on the clusterpattern, it often resembles a horseshoe.

When placed in riffle and glide areas, boulderscan create pocket-like pools that provide restingareas for rearing or migrating fish (Rosgen andFittante 1986). Avoid placing rocks at sites withfine or unstable beds that scour readily (BritishColumbia Ministry of Environment 1980; Wardand Slaney 1981). As with blunt-faced boulders,rocks placed in these sites tend to sink into theirown scour holes unless the substrate is armored(Fisher and Klingeman 1984; Cullen 1989). Simi-larly, avoid placing rocks in depositional areaswhere the rocks may become buried by sediment(Ward and Slaney 1981). Occasionally, rocks maybe used in such areas to encourage deposition.


contain a site plan, grading plan, layout, irri-gation, planting, and detail sheets. The graph-ics must be complete, accurate, and very easyto read. They should be accompanied bydetailed written instructions, called construc-tion specifications, which supplement andclarify the graphic on the drawings. An ex-ample of construction specifications is pro-vided in Appendix D.

In addition to the elements described above(specific quantities, densities, species), finaldesign plans should include: a key to symbolsfor deciduous and coniferous trees and shrubs,paving, riprap, grass, and forbs; scale (graphicor numerical value: graphic scales have theadvantage of changing at the same rate as therest of the page if enlarged or reduced); and aNorth arrow (Figure 7.18). A complete list ofinformation that should be included in thefinal project drawings and construction docu-ments is provided below:

• Overall location plan showing accessto site from local highways (drawn atany appropriate scale).

• A drawing of the entire bank or repairarea, locating each type of treatment.

• Right-of-way and easement areas.• An elevation drawing of the repair,

identifying whether the elevation istruly vertical or parallel to the slope.Include topographic information (onefoot intervals) of the existing andproposed contours.

• Existing river protection facilities andchannel hardpoints such as large rocksor bedrock areas.

• Roads, gutters, swales, and otherphysical features.

• Existing property improvements (e.g.,homes and other buildings) and utilitiesincluding septic drainfields.

• Temporary construction staging areafor material stockpiling and equipmentstorage.

• Top and toe of bank and water levels(ordinary high water and design flows).

• Proposed areas of cut and fill.

7.2.6 SUMMARY OF DESIGNCONSIDERATIONS

To help designers select solutions appropriatefor each situation, Table 7.4 summarizes eachmethod including relative quantities of material,major limitations, and costs. Installation proce-dures are discussed in Chapter 8. This table shouldbe used in conjunction with the numerous designfactors discussed in the previous chapters.

7.3 DESIGN DRAWINGS, PLANSAND SPECIFICATIONS

Conceptual designs explore ideas and rela-tionships among functions, activities, and spaces.They serve as a basis for further development, asa means of conveying graphic concepts to otherindividuals working on a project, and a mediumfor feedback during the design development stage.They are rough sketches lacking detail, servingmainly to get preliminary ideas on paper. Thereare no right or wrong conventional symbols.

During design development, the rough sketchesand ideas developed in the conceptual designphase are tested and refined. The designer evalu-ates possibilities identified in the conceptual phaseand rejects, adds to, or modifies them. This phaseresults in drawings that include specific informa-tion on spatial organization, material, and sizes offeatures. Called “presentation drawings”, theseare used to communicate ideas and obtain feed-back for later design refinement. Presentation draw-ings should be fairly realistic and self-explana-tory, with limited text for labels. The level of detailprovided is intermediate between conceptual andfinal plans. The plan view (straight down fromabove) is the most commonly used projection inthe design development stage. Section-elevationsor cross-sections are also useful to show the verti-cal arrangement of the streambank.

Final designs are prepared for the people whoactually install the project. These should clearlydescribe the exact sizes, shapes, quantities, types,and locations of all project elements. This infor-mation is used to prepare bids, as well as in theactual construction process. The drawings should


Met

hod

Rec

om

men

ded

Posi

tion o

n b

ank

Quantity

of

live

Majo

r lim

itations

Rec

om

men

ded

use

Cost

seaso

n t

o in

stall1

mate

rial r

equir

edor

dra

wback

s

brus

h la

yers

whe

n pl

ants

abov

e or

dina

ryve

ry la

rge2

cons

truct

ion

site

s re

quiri

ng fi

ll;m

oder

ate3

are

dorm

ant

high

wat

erdu

ring

low

flow

scou

r ho

les

brus

h m

attre

ssw

hen

plan

tsab

ove

ordi

nary

very

larg

e2 ; 10

tono

t use

ful o

n ste

ephi

gh v

eloc

ity o

r2-

5 hr

/sq.

yd.

are

dorm

ant

high

wat

er20

lbs

vege

tatio

nba

nks,

con

struc

tion

high

gra

dien

t,to

insta

llpe

r sq

. yd.

3du

ring

low

flow

whe

re s

lope

s ca

nbe

cut

bac

k

fenc

ing

any

seas

onin

ove

rban

k ar

eas

none

does

not

in c

onju

nctio

nva

ries

with

stabi

lize

ban

kw

ith o

ther

met

hods

type

; $5-

15/f

t.

join

t pla

ntin

gw

hen

plan

tsab

ove

ordi

nary

mod

erat

e to

larg

eth

ick

ripra

p m

ayw

here

rip

rap

is5-

7 sq

. yd.

per

are

dorm

ant

high

wat

erbe

diff

icul

t to

stake

requ

ired

or a

lread

yho

ur, i

nclu

ding

in p

lace

cutti

ng a

ndtra

nspo

rt3

live

crib

wal

lw

hen

plan

tsfro

m b

elow

cha

nnel

mod

erat

e2m

ay r

equi

re w

ork

limite

d sp

ace,

stro

ngva

ries

are

dorm

ant

bed;

veg

etat

ion

abov

ein

cha

nnel

bed

;cu

rren

t, ste

ep b

anks

ordi

nary

hig

h w

ater

cons

truct

ed u

sual

lydu

ring

low

flow

fasc

ines

(wat

tles)

whe

n pl

ants

abov

e or

dina

rym

oder

ate2

whe

re d

issi

patio

n an

dlo

w c

ost3 ;

$0.

91ar

e do

rman

thi

gh w

ater

dire

ctio

n of

wat

erto

$2.

44 p

er ft

4

mov

emen

t dow

nslo

peis

nee

ded

live

stake

s,w

hen

plan

tsab

ove

ordi

nary

mod

erat

e to

larg

ela

bor

inte

nsiv

e;ar

eas

with

min

or s

cour

,cu

tting

s: $

0.35

-cu

tting

sar

e do

rman

thi

gh w

ater

for

stake

s, s

mal

lvu

lner

able

tofe

rtile

soi

ls; s

ites

with

0.71

/sq.

yd.

4

for

cutti

ngs

vand

alis

mlim

ited

equi

pmen

tac

cess

herb

aceo

ussp

ring

or e

arly

abov

e or

dina

rysm

all (

hydr

osee

d)se

e ta

bles

7.1

whe

re la

rge

or w

oody

varie

s; $

0.01

-gr

ound

cov

ersu

mm

er; l

ate

high

wat

eran

d 7.

2ve

geta

tion

is n

ot0.

15/s

q. y

d.4

sum

mer

to e

arly

fall

desi

rabl

e; w

here

fast

cove

r is

nee

ded

rock

rip

rap

any

seas

on (m

aybe

low

ord

inar

yno

nege

nera

lly n

ot u

sefu

lw

here

toe

arm

or is

varie

s; $

25-3

0/cu

.be

lim

ited

byhi

gh w

ater

; or

on s

teep

slo

pes

need

ed; h

igh

velo

city

yd. p

lace

dpe

rmit

prov

isio

ns)

high

er o

n ba

nk(g

reat

er th

an 2

H:1

V)

area

s; fl

ood-

fight

ing

as r

equi

red

emer

genc

ies

Tabl

e 7.

4Su

mm

ary

of d

esig

n co

nsid

erat

ions

for

vario

us b

ank

stab

iliza

tion

met

hods

.


Tabl

e 7.

4Su

mm

ary

of d

esig

n co

nsid

erat

ions

for

vario

us b

ank

stab

iliza

tion

met

hods

, con

tinue

d.

Met

hod

Rec

om

men

ded

Posi

tion o

n b

ank

Quantity

of

live

Majo

r lim

itations

Rec

om

men

ded

use

Cost

seaso

n t

o in

stall1

mate

rial r

equir

edor

dra

wback

s

root

ed s

tock

fall

to s

prin

gab

ove

ordi

nary

varie

ssu

mm

er in

stalla

tions

whe

n la

rge

plan

ts fo

rva

ries

with

siz

ehi

gh w

ater

may

req

uire

irrig

atio

nco

ver

is d

esire

dan

d sp

ecie

s

tree

reve

tmen

tan

y se

ason

at o

rdin

ary

varie

sm

ust b

e re

plac

edfo

r pr

otec

tion

from

varie

shi

gh w

ater

perio

dica

llyba

nk s

cour

and

unde

rcut

ting

vege

tate

dw

hen

plan

tsve

geta

tion

abov

eve

ry la

rge

requ

ires

wor

k in

limite

d sp

ace,

ste

epsli

ghtly

mor

e th

ange

ogrid

are

dorm

ant

ordi

nary

hig

h w

ater

chan

nel b

ed,

slope

s th

at c

an’t

bega

bion

s; le

ast

cons

truct

ion

durin

gcu

t bac

k, h

igh

velo

city

expe

nsiv

e re

tain

ing

low

flow

wal

l3

1.M

ost v

eget

ativ

e m

etho

ds h

ave

high

er s

urvi

val i

f con

struc

ted

whi

le p

lant

s ar

e do

rman

t and

whe

nth

e m

ater

ial i

s fre

shly

cut

; man

y w

ill su

rviv

e at

oth

ertim

es, h

owev

er, i

f ade

quat

e m

oistu

re is

ava

ilabl

e or

pro

vide

d an

d th

e m

ater

ial i

s no

t dam

aged

by

cons

truct

ion

activ

ities

.

2.D

icke

rson

199

2.

3.Sc

hiec

htl 1

980.

4.W

eave

r and

Mad

ej 1

981.


Figure 7.18 Examples of symbols for plans and specifications.

Figure 7.18 Examples of symbols for plans and specs

0 100 200 300

ONEACRE

Deciduous trees

Evergreen trees Plant groups

Shrubs

Rocks

Scale and north arrow North arrows

Evergreen shrubs


• Existing trees and major plant materialswith clear indication of whether each itemis to remain or be removed.

• Configuration of the design solution aroundexisting (remaining) trees and othervegetation.

• The zone around existing trees to beprotected during construction.

• Areas of bank failures and extent ofproposed design.

• Sources of plant materials.• Plant names and sizes.• Location of temporary irrigation systems if

applicable.• Power and water source and point of

connection if applicable.• Name and phone number of contact person

representing the project sponsor.• Any special conditions unique to the site.


RECOMMENDED REFERENCES FORADDITIONAL INFORMATION

Gray, D.H. and A.T. Leiser. 1982. BiotechnicalSlope Protection and Erosion Control. VanNostrand Reinhold Company. New York,N.Y.

Lagasse, P.F., J.D. Schall, F. Johnson, E.V.Richardson, J.R. Richardson, and F.Chang. 1991. Stream Stability at HighwayStructures. Pub. No. FHWA-IP-90-014.Hydraulic Engineering Circular No. 20.Federal Highway Administration.

Richardson, E.V., L.J. Harrison, and S.R. Davis.1991. Evaluating Scour at Bridges.Hydraulic Engineering Circular No. 18.Federal Highway Administration.

Richardson, E.V., D.B. Simons, and P.Y. Julien.1990. Highways in the River Environment.Federal Highway Administration.

Schiechtl, H. 1980. Bioengineering for landreclamation and conservation. Universityof Alberta Press. Edmonton.

Simons, Li and Associates. 1982. EngineeringAnalysis of Fluvial Systems. Fort Collins,Colo.

Documents

CHAPTER 7 DESIGN GUIDELINES - King County, Washington · ods, and a brief discussion on preparing design drawings, plans, and specifications. 7.1 STREAMBANK ZONES As discussed in