compendium of case histories on repair of erosion-damaged concrete in hydraulic structures

Keywords: abrasion; abrasion resistance; aeration; cavitation; chemical attack; concrete dams; concrete pipes; corrosion; corrosion resistance; deterioration; erosion; grinding material

ACI 210.1 R-94

Compendium of Case Histories on Repair of

Erosion-Damaged Concrete in Hydraulic Structures

Reported by ACI Committee 210

(Reapproved 1999)

Stephen B Tatro Chairman

Patrick J Creegan Angel E Herrera

James R Graham Richard A Kaden

This report is a companion document to ACI 210R It contains a series of

case histories on hydraulic structures that have been damaged by erosion

from various physical mechanical and chemical actions Many of these

structures have been successfully repaired There were many examples to

select from; however, the committee has selected recent, typical projects,

with differing repair techniques, to provide a broad range of current

exper-ience These case histories cover only damage to the hydraulic surfaces due

to the action of water, waterborne material or chemical attack of concrete

from fluids conveyed along the hydraulic passages In addition to repairs

of the damaged concrete, remedial work frequently includes design

modi-fications that are intended to eliminate or minimize the action that

pro-duced the damage This report does not cover repair of concrete damaged

by other environmental factors such as freeze-thaw, expansive aggregate, or

corroding reinforcement.

Keywords: abrasion; abrasion resistance; aeration; cavitation; chemical attack;

concrete dams; concrete pipes; corrosion; corrosion resistance; deterioration;

erosion; grinding (material removal); high-strength concrete hydraulic structures;

maintenance; outlet works; penstocks; pipe linings; pipes (tubes); pittings; polymer

concrete; renovating; repairs; sewers; spillways; tolerances (mechanics); wear.

CONTENTS

Chapter l-Introduction, p 210.1R-1

Chapter 2-Cavitation-erosion case histories, p 210.1R-2

Dworshak Dam

Glen Canyon Dam

Lower Monumental Dam

Lucky Peak Dam

Terzaghi Dam

Yellowtail Afterbay Dam

Yellowtail Dam

Keenleyside Dam

ACI Committee Reports, Guides, Standard Practices, and

Commentaries are intended for guidance in designing,

plan-ning, executing, or inspecting construction and in preparing

specifications References to these documents shall not be

made in the Project Documents If items found in these

documents are desired to be part of the Project

Docu-ments, they should be phrased in mandatory language and

incorporated into the Project Documents.

James E McDonald Ernest K Schrader

Chapter 3-Abrasion-erosion case histories, p 21O.lR-13

Espinosa Irrigation Diversion DamKinzua Dam

Los Angeles River ChannelNolin Lake Dam

Pine River Watershed, Structure No 41Pomona Dam

Providence-Millville Diversion StructureRed Rock Dam

Sheldon Gulch Siphon

Chapter 4-Chemical attack-erosion case histories, p 210.1R-25

Barceloneta Trunk SewerDworshak National Fish HatcheryLos Angeles Sanitary Sewer System andHyperion Sewage Treatment FacilityPecos Arroyo Watershed, Site 1

Chapter 5-Project reference List, p 210.1R-32

CHAPTER 1-INTRODUCTION

This compendium of case histories provides tion on damage that has occurred to hydraulic structuresand the various methods of repair that have been used.ACI Committee 210 has prepared this report to help oth-ers experiencing similar problems in existing work.Knowledge gained from these experiences may help

informa-ACI 210.1R-94 became effective Nov 1.1994.

Copyright 8 1994, American Concrete Institute.

All rights reserved including rights of reproduction and use in any form or by any means, including the making of copies by any photo process, or by any elec- tronic or mechanical device, printed, written, or oral, or recording for sound or visual reproduction for use in any knowledge or retrieval system or device, unless permission in writing is obtained from tbe copyright proprietors.

210.1R-1

210.1R-2 ACI COMMITTEE REPORT

avoid oversights in design and construction of hydraulic

structures and provide guidance in the treatment of

future problems

Erosion of concrete in hydraulic structures may occur

as a result of abrasive action, cavitation, or chemical

attack Damage may develop rapidly after some unusual

event such as a flood or it may develop gradually during

normal continuous operation or use In most cases where

damage has occurred, simply replacing the eroded

con-crete will ensure immediate serviceability, but may not

ensure long-term performance of the structure

There-fore, repair work usually includes replacing eroded

concrete with a more resistant concrete and additional

surface treatment, modifying the design or operation of

the structure to eliminate the mechanism that produced

the damage, or both A detailed discussion of

mechan-isms causing erosion in hydraulic structures, and

recommendations on maintenance and repair, is

con-tained in ACI 210R

When damage does occur to hydraulic structures,

repair work poses some unique problems and is often

very costly Direct access to the damaged area may not

be possible, or may be limited by time, or other

con-straints In some cases, such as repair to spillway stilling

basin floors, expensive bulkheads and dewatering are

required It may not be possible to completely dry the

area to be repaired or maintain the most desirable

temperature A great deal of planning and scheduling for

repair work are normally required, not only for the

repairs and access, but also for control of water releases

and reservoir levels If time permits, extensive

inves-tigation usually precedes planning and scheduling to

determine the nature and extent of damage Hydraulic

model studies may also be necessary to evaluate possible

modifications in the design or operation of the facility

This compendium provides the history on 21 projects

with hydraulic erosion damage They vary in size and

cover a variety of problems: 8 with cavitation damage, 9

with abrasion-erosion damage, and 4 with erosion

damage from chemical attack Table 1.1 summarizes the

projects Each repair was slightly different Each history

includes background information on the project or

facil-ity, the problem of erosion, the selected solution to the

problem, and the performance of the corrective action

Histories also contain references and owner information

if further details are needed

CHAPTER 2-CAVITATION-EROSION

CASE HISTORIES DWORSHAK DAM

North Fork, Clearwater River, Idaho

BACKGROUND

Dworshak Dam, operational in 1973, is a straight-axis

concrete gravity dam, 717 ft high, 3287 ft long at the

crest, and contains 6,500,000 cubic yards of concrete Inaddition to two gated overflow spillways, three regulatingoutlets, 12 ft wide by 17 ft high, are located in the spill-way monoliths The inlet elevation for each regulatingoutlet is 250 ft below the maximum reservoir elevation.Each outlet jet is capable of a maximum discharge of14,000 fij/s

Outlet surfaces are reinforced structural concreteplaced concurrently with adjacent lean, large aggregateconcrete Coatings to the outlet surfaces were appliedduring the original construction In Outlet 1, the wall andinvert surfaces from the tainter gate to a point 50 ftdownstream are coated with an epoxy mortar having anaverage thickness of % in The same area of Outlet 2 wascoated using an epoxy resin, approximately 05 in inthickness Outlet 3 was untreated

The outlets were operated intermittently at variousgate openings for a period of 4 years between 1971 and

1975, resulting in a cumulative discharge duration ofapproximately 10 months The three outlets were notoperated symmetrically; outlets 1 and 2 were used pri-marily

PROBLEM

Inspection in 1973 showed minor concrete scaling ofthe concrete wall surfaces of Outlets 1 and 2 One yearlater, in 1974, serious erosion had occurred at wallsurfaces of both outlets immediately downstream of thewall coatings, 50 ft from the tainter gate Part of this wallarea had eroded to a depth of 22 in., exposing and evenremoving some No 9 reinforcing bars In the wall sur-faces downstream of Outlet 1 medium damage, up to 1

in depth of erosion, also occurred in over 60 squareyards of surface, bordered by lighter erosion Everyhorizontal lift joint (construction joint) along the path ofthe jet, showed additional cavitation erosion

SOLUTION

Repairs were categorized as three types:

l Areas with heavy damage, with erosion greater than

2 to 3 in., were delineated by a 3-in saw cut andthe interior concrete excavated to a minimumdepth of 15 in (Fig 2.1 and 2.2) Reinforcementwas reestablished and steel fiber-reinforced con-crete (FRC) was used as the replacement material

l Areas with medium damage, where the depth oferosion was less than 1 in., were bush-hammered to

a depth of % to 1 in and dry-packed with mortar.The mortar, if left untreated, would easily havefailed when subjected to the high velocity discharge

l Areas with minor damage, surfaces showing a blast texture, were not separately treated prior topolymer impregnation The entire wall surfaces ofOutlet 1 were then treated by polymer impregna-tion from the downstream edge of the existingepoxy mortar coating to a distance 200 ft down-stream

sand-REPAIR OF EROSION DAMAGED HYDRAULIC STRUCTURES 210.1 R-3 TABLE 1.1 - SUMMARY TABLE OF PROJECTS COMPRISING THIS REPORT

Reference page

210.1R-2

Year

Authority redesign Stilling basin Montana Bureau of Cavitation Various overlays

RecIamation 1966

Espinosa Irrigation

Diversion Dam

1984 Diversion dam New Mexico

1965 Stilling basin Pennsylvania

Soil vation Service Corps of Engineers

Conser-Abrasion

Abrasion

Steel plate armor

Silica fume concrete

210.1R-13

210.1R-15 Kinzua Dam

Los Angeles River

Channel

210.1R-17 7 1940s Channel California

1963 Stilling basin Kentucky

Proposed Channel Colorado

Corps of Engineers Corps of Engineers SoiI Conser- vation Service

Abrasion

Abrasion

Abrasion

Siiica fume concrete Hydraulic redesign High-strength concrete

Pine River Watershed,

Stilling basin Kansas

Diversion dam Utah

Stilling basin Iowa

Syphon outlet Wyoming

Pipeline Puerto Rico

Concrete tanks Idaho

Sewerage California structures

Corps of Abrasion various Engineers

Soil Conser- Abrasion Surface hardener vation Service

Corps of Abrasion Underwater Engineers concrete Soil Conser- Abrasion Polymer-modified vation Service mortar

Puerto Rico Chemical attack PVC lining Aqueduct &

Sewer Authority Corps of Chemica l attack Linings Engineers

City of Los Chemical attack Shotcrete and Angeles PVC liners

210.1R-20

210.1R-22 Providence-Millville

Diversion Structure

Red Rock Dam 210.1R-23

Sheldon Gulch Siphon 210.1R-25

Los Angeles Sanitary

Sewer System and

Damage to the epoxy mortar was minimal and located PERFORMANCE

near the outlet gate This area was repaired with new

epoxy

The polymer impregnation process involved drying all

the surfaces to a temperature up to 300 F to drive off

water and then allowing the surface to cool to 230 F

Monomer was then applied to the surface using a vertical

soaking chamber Excessive monomer was drained and

the surface was polymerized by the application of

approx-imately 150 F water

Operation of the outlets from the time of repair in

1975 until 1982 has been minimal averaging 1400 ft3/s

per outlet with peak discharges of 3600 ft3/s per outlet.Durations of usage are not known After 1982 outlet dis-charges increased, with durations exceeding 50 days.Inspections performed in 1976, the year after therepairs, showed no additional concrete damage except forsome minor surface spalling adjacent to a major pre-existing crack in an area of dry-packed mortar The

210.1R-4 ACI COMMlTTEE REPORT

Fig 2.1-Dworshak Dam Detail showing depth of erosion behind reinforcing steel

Fig 2.2-Dworshak Dam Extent of outlet surface preparation prior to concrete and mortar placements

spalled area was patched with epoxy paste, except that

the epoxy paste did not bridge the crack this time Epoxy

resin coating repairs applied to Outlet 2 showed some

failures,

Inspections in 1983 and 1988 showed that epoxy

mor-tar coatings in Outlet 1 continued to perform well Small

areas of damage, typically spalls, are periodically repaired

with a paste epoxy Epoxy resin coatings in Outlet 2 are

repaired more frequently but are performing adequately

Surfaces repaired with FRC and mortar and subsequently

polymer-impregnated showed negligible damage

Poly-mer-impregnated parent concrete shows a typical matrix

erosion around the coarse aggregate to a depth of 1

/ 4-in.,and lift joints exhibit pitting up to 3

/ 8-in deep Surfaces

along lift joints not polymer-impregnated show erosion

up to 3 / 4-in in depth and a general surface pitting greaterthan the companion polymer-impregnated surfaces,

DISCUSSION

Because of variation in the operation of these outlets,both in flow rate and duration, exact time-rate erosionconclusions are difficult to make Recent outlet dischargehas fluctuated annually from moderate flows to none Ingeneral, surfaces that received replacement materials andwere subsequently polymer-impregnated have performedwell Original concrete and new polymer impregnatedconcrete is showing evidence of deterioration, but at arate that is less than the unimpregnated surfaces The

REPAIR OF EROSION DAMAGED HYDRAULIC STRUCTURES 210.1 R-5

best performance was by the original epoxy mortar

coat-ing The epoxy mortar in Outlet 1 continues to display an

excellent surface condition, with no cavitation-generated

pitting The epoxy resin coating in Outlet 2 displays good

performance

In 1988, outlets were modified by adding aeration

de-flectors, wedges 27 in wide by 1.5 in high, to the sides

and bottom of each outlet These deflectors were

de-signed to increase the aeration of the discharge jet and

further reduce the cavitation erosion of the outlet

sur-faces Subsequent deterioration of the outlet surfaces has

not been observed

The polymer impregnating of the concrete surfaces of

the outlets was a very complex system of operations

Suc-cess requires continual evaluation of application

tions and flexibility to react to changes in those

condi-tions Issues relating to safety, cost, and field engineering

add significant challenges to a polymer impregnation

pro-ject It is doubtful that this process would be attempted

today under similar circumstances It is more likely that

the aeration deflectors would be the first remedy

con-sidered since they provide a positive solution to the

problem without the higher risks of a failure inherent in

the polymer impregnation process

REFERENCES

Schrader, Ernest K., and Kaden, Richard A, “Outlet

Repairs at Dworshak Dam,” The Military Engineer, The

Society of American Military Engineers, Washington,

D.C., May-June 1976, pp 254-259

Murray, Myles A, and Schultheis, Vem F.,

“Polymer-ization of Concrete Fights Cavitation,” Civil Engineering,

V 47, No 4, American Society of Civil Engineers, New

York, April 1977, pp 67-70

U.S Army Engineer District, Walla Walla, “Polymer

Impregnation of Concrete at Dworshak Dam,” Walla

Walla, WA, July 1976, Reissued April 1977

U.S Army Engineer District, Walla Walla, “Periodic

Inspection Reports No 6, 7, and 8, Dworshak Dam and

Reservoir,” Walla Walla District, Jan 1985

CONTACT/OWNER

Walla Walla District, Corps of Engineers

City-County Airport

Walla Walla, WA 99362

GLEN CANYON DAM

Colorado River, Northeast Arizona

BACKGROUND

Glen Canyon Dam, operational in 1964, is a concrete

gravity, arch structure, 710 ft high with a crest length of

1560 ft The dam is flanked on both sides by high-head

tunnel spillways, each including an intake structure with

two 40- by 55-ft radial gates Each tunnel consists of a

41-ft diameter section inclined at 55 percent, a vertical

bend (elbow), and 985 ft of near horizontal length lowed by a deflector bucket Water first flowed throughthe spillways in 1980, 16 years after completion of thedam

fol-PROBLEM

In late May 1983, runoff in the upper reaches of theColorado River was steadily increasing due to snowmeltfrom an extremely heavy snowpack On June 2,1983, theleft tunnel spillway gates were opened to release 10,000

ft3 / s On June 5 the release was increased to 20,000 ft3 / s

On June 6 officials heard loud rumbling noises comingfrom the left spillway Engineers examined the tunneland found several large holes in the invert of the elbow.This damage was initiated by cavitation, triggered by dis-continuities formed by calcite deposits on the tunnelinvert at the upstream end of the elbow In spite of thisdamage, continued high runoff required increasing thedischarge in the left spillway tunnel to 23,000 ft3 / s byJune 23 Flows in the right spillway tunnel were held at

6000 ft3 / s to minimize damage from cavitation Spillwaygates were finally closed July 23, and engineers made athorough inspection of the tunnels

Extensive damage had occurred in and near the lefttunnel elbow (Fig 23) Immediately downstream fromthe elbow, a hole (35 ft deep, 134 ft long, and 50 ft wide)had been eroded in the concrete lining and underlyingsandstone foundation Other smaller holes had beeneroded in the lining in leapfrog fashion upstream fromthe elbow

SOLUTION

Therepair work was accomplished in six phases: 1) moving loose and defective concrete lining and founda-tion rock; 2) backfilling large cavities in sandstone foun-dation with concrete; 3) reconstructing tunnel lining; 4)grinding and patching of small defective areas; 5) remov-ing about 500 cubic yards of debris from lower reaches oftunnel and flip bucket; and 6) constructing an aerationdevice in the lining upstream of the vertical elbow.Sandstone cavities were filled with tremie concrete be-fore the lining was replaced About 2000 cubic yards ofreplacement concrete was used The aeration slot wasmodeled in the Bureau of Reclamation Hydraulic Labor-atory to ensure that its design would provide the per-formance required

re-The aeration slot was constructed on the inclined tion of the tunnel approximately 150 ft upstream fromthe start of the elbow A small 7-in-high ramp was con-structed immediately upstream of the slot The slot was

por-4 by por-4 ft in cross section and extended around the lowerthree-fourths of the tunnel circumference (Fig 2.4) Allrepairs and the slot were completed in the summer of1983

PERFORMANCE

Because of the moderate runoff in the Colorado Riversince completion of the tunnel repairs, it has not been

210.1R-6 ACI COMMITTEE REPORT

Fig 2.3-Glen Canyon Dam Erosion of spillway tunnel invert and sandstone foundation

rock downstream of the elbow

necessary to use the large spillway tunnels However,

shortly after completion of the work, another high runoff

period permitted performance of a field verification test

This test lasted 72 hr with a maximum flow during that

time of 50,000 ft3 / S The test was conducted in two

phases with several interruptions in each for examination

of the tunnel Offsets were intentionally left in place to

evaluate whether the aeration slot would indeed preclude

cavitation and attendant concrete damage The tunnel

re-pairs and air slot performed as designed No sign of

cavi-tation damage was evident anywhere in the tunnel

Aera-tion has decreased the flow capacity of the spillway

tunnels by approximately 20 percent of the original flow

capacity

REFERENCES

Burgi, P.H., and Eckley, M.S., “Repairs at Glen

Can-yon Dam,” Concrete International, American Concrete

Institute, MI, V 9, No 3, Mar 1986, pp 24-31

Frizell, K.W., “Glen Canyon Dam Spillway Tests

Model - Prototype Comparison,” Hydraulics and

Hydro-logy in the Small Computer Age, Proceeding of the

Spe-cialty Conference, Lake Buena Vista, Florida, Aug

12-17, 1985, American Society of Civil Engineers, NewYork, 1985, pp 1142-1147

Frizell, K.W., “Spillway Tests at Glen Canyon Dam,”U.S Bureau of Reclamation, Denver, CO, July 1985.Pugh, C.A., “Modeling Aeration Devices for Glen

Canyon Dam,” Water for Resource Development,

Proceed-ings of the Conference, Coeur d’Alene, Idaho, Aug.14-17, 1984, American Society of Cii Engineers, NewYork, 1984, pp 412416

CONTACT

U.S Bureau of ReclamationP.O Box 25007, Denver Federal CenterDenver, CO 80225

LOWER MONUMENTAL DAM

Snake River, Near Kaloutus, Washington

BACKGROUND

Lower Monumental Dam, operational in 1970, consists

of a concrete gravity spillway and dam, earthfii

Original tunnel surfac

Aeration slot

.I8 SECTION A-A Fig 2.4-Glen Canyon Dam Diagram of new tunnel spillway air slot

bankments, a navigation lock, and a six-unit powerhouse

The 86-ft wide by 675-ft long navigation lock chamber,

with a rise of 100 ft, is filled and emptied by two galleries

or culverts, landside and riverside of the lock structure

The landside culvert, which supplies five downstream

lat-erals, crosses under the navigation lock to discharge into

the river The riverside culvert supplies and discharges

water to the upstream five laterals Each lateral consists

of 10 portal entrances approximately 1.5 ft wide by 3 ft

high Plow velocities in excess of 120 ft/s occur in several

of the portals entrances A tie-in gallery exists between

the two main culverts, near the downstream gates, that

equalizes the pressure between the two culverts

PROBLEM

Inspections as early as 1975 revealed that the ceiling

concrete of the landslide culvert was spalled at some

monolith joints to depths of 9 in This may have been

ini-tiated by differential movement of adjacent monoliths

when the lock chamber was filled and emptied Damage

to the invert, at several locations, was irregular, with

erosion a maximum of 18 in deep at the monolith joint,

decreasing to 1 in at a point 10 ft upstream of the joint

Reinforcing steel was exposed Other areas of erosion in

the invert and on wall surfaces were observed, measuring

2 ft square and 2 in deep

Later inspections revealed that portal surfaces nearest

the culverts of the most downstream laterals were

show-ing signs of concrete erosion (Fig 2.5) By 1978, the

por-tal walls, ceiling, and invert had eroded as deep as 3 in

over an area of 5 square ft, exposing reinforcing steel.All four corners of the tie-in gallery experienced ob-vious cavitation damage The damage varied from minorpitting to exposure and undercutting of the 11 / 2-in aggre-gate

SOLUTION

In 1978, the navigation lock system was shut down fortwo weeks for repairs The major erosion damage to thelandslide culvert was repaired by mechanically anchoredsteel fiber-reinforced concrete The smaller areas ofdamage received a trowel application of a paste epoxyproduct Ceiling damage was backfilled with dry-mixshotcrete Portal and tie-in gallery surfaces receivedapplication of a paste epoxy, troweled to a feather edgearound the perimeter

PERFORMANCE

The mechanically anchored fiber-reinforced concretehas performed well to date No additional erosion hasbeen observed Shotcrete patches to the ceiling adjacent

to the joints show continued spalling, but to a lesserextent than prior to repairs

The repairs to the portal surfaces and tie-in gallerysurfaces performed poorly After 1 year of service, ap-proximately 40 percent of the epoxy paste had failed; andafter 3 years, nearly 100 percent has failed Concreteerosion in these areas has subsequently increased todepths of 6 to 8 in in the tie-in gallery and up to 5 to 6

in on the two most downstream portal surfaces

210.1R-8 ACI COMMITTEE REPORT

DISCUSSION

Recent inspections have shown that the rate of erosion

has decreased The accumulated erosion of concrete from

certain surfaces is significant; however, subsequent

ero-sion is almost negligible Consequently, repair schedules

are not critical

Paste epoxy was applied to the concrete surfaces

tran-sitioning to feather edges along the perimeter of the

patches Cavitation eroded the concrete adjacent to the

feather edges as weIl as eroding the thin epoxy edges

(Fig 2.5) These new voids undermined the new, thicker

epoxy, and at some point caused another failure of the

leading edge As the leading edge void increased in size,

the failure progressed until little epoxy was left in the

repaired area After erosion of the epoxy patch material,

no further concrete erosion has occurred It appears that

the eroded configuration of the surface is hydraulically

stable

Patch-type repair procedures are not sufficient for this

structure because erosion is initiated at the edge of the

new patch Eventual repairs will replace larger areas of

the concrete flow surfaces and will include substantial

anchoring of new materials

U.S Army Engineer District, Walla Walla, “Periodic

Inspection Report No 6, Lower Monumental Lock and

Dam,” Walla Walla, WA, Jan 1977

U.S Army Engineer District, Walla Walla, “Periodic

Inspection Report No 7, Lower Monumental Lock and

Dam,” Walla Walla, WA, Jan 1981

U.S Army Engineer District, Walla Walla, “Periodic

Inspection Report No 8, Lower Monumental Lock and

Dam,” Walla Walla, WA, Jan 1983

CONTACT/OWNER

Walla Walla District, Corps of EngineersCity-County Airport

Walla Walla, WA 99362

LUCKY PEAK DAM

Boise River, Near Boise, Idaho

BACKGROUND

Lucky Peak Dam, operational in 1955, is 340 ft highwith a crest length of 2340 ft The dam is an earth androckfill structure with a silt core, graded filters, and rockshells The ungated spillway is a 6000-ft-long ogee weirdischarging into an unlined channel The outlet worksconsists of a 23-ft-diameter steel conduit that deliverswater to a manifold structure with six outlets Each outlet

is controlled by a 5.25-ft by 10-ft slide gate Individualflip lips were constructed downstream from each slidegate Downstream of the flip lips is the plunge pool, ex-cavated into the basalt rock, with bottom areal dimen-sions of 150 by 150 ft The outlet alignment and designwere determined by hydraulic modeling The sir outletsoperated under a maximum head of 228 ft with a designdischarge of 30,500 ft3 / S and a maximum discharge vel-

ocity ranging between 88 ft/s and 124 ft/s

PROBLEM

The steel manifold gates have a long history of tation erosion problems The original bronze gate sealswere seriously damaged by cavitation after initial use.Flow rates across the manifold gate frames in excess of

cavi-150 ft/s for many days were common The gate seals werereplaced with new seals made of stainless steel andaluminum-bronze The cast-steel gate frames requiredcontinual repair of cavitated areas In 1975 alone, over

2000 pounds of stainless steel welding rod was manuallywelded into the eroded areas and ground smooth Neatcement grout was pumped behind the gate frames to re-establish full bearing of the gate frames with the concretestructure

The concrete invert and side piers, which separateeach of the six flip lips suffered extensive erosion soonafter the start of operations in 1955 (Fig 2.6) 3 / 4-in.-thicksteel plates were anchored to the piers and invert areasjust downstream of the manifold gates These steel wallplates became severely pitted, as did the downstreamconcrete flip lip invert surfaces In 1968, the damagedplates were again repaired by filling the eroded areaswith stainless steel welding, and grouting behind theplates Deteriorated concrete on the flip lips was re-moved and additional steel plates were installed overthose areas This also failed and repairs commencedagain Deep areas of cavitation damage in the invert andpiers were filled with concrete New 1 / 2-in.-thick plateswere installed These were stiffened with steel beams,welded on 5-ft centers in each direction Deep anchor

REPAIR OF EROSlON DAMAGED HYDRAULIC STRUCTURES 210.1R-9

bars were welded to the plate material to hold them in

place Again, the voids under the plates were grouted

But during the next two years, these repairs also failed

In 1974, it was recommended that the outlet be

re-studied hydraulically That year, remaining plate material

was removed Cavities were found penetrating the invert

and through the piers and into the adjacent outlet invert

These voids were crudely filled with FRC in a “field

expedient” manner Much of this FRC was placed in

standing water with little quality control, while adjacent

bays were discharging

SOLUTION

The side piers were redesigned and replaced to

pro-vide vents that would introduce air to the underside of

the jet just downstream of the gates This modification

was intended to prevent additional invert erosion

How-ever, major modifications to the gates and gate frames

were necessary if cavitation erosion was to be eliminated

These modifications were not made since future

power-house construction would reduce and nearly eliminate

the need to use the outlet, reserving the structure for

emergency and special operations use only Steel lining

on the piers was strengthened and replaced Stiffened

steel plates, 11 / 4-in thick, were installed on the piers and

invert Mortar backfill was pumped behind the invert

plates and new concrete placed between pier plates

PERFORMANCE

After one year of above average usage on bays 3 and

4, cavitation was again observed The side piers just

downstream of the gates showed areas of 1 to 2 square

ft that had eroded through the steel plate and into the

concrete about 6 in No erosion of the invert plates or

the “field expedient” FRC occurred Use of these bays

has almost stopped since the new powerhouse became

operational

DISCUSSION

The introduction of air beneath the jet appears to

have cushioned the effects of cavitation on the flip lip

invert However, pier walls continue to erode at an

extra-ordinary rate The cause lies with the design of the gates

and gate frame It is evident that satisfactory

perfor-mance of the structure can never be achieved until the

gates and frames are redesigned and reconstructed to

eliminate the conditions that cause cavitation

REFERENCES

U.S Army Engineer District, Walla Walla, “Lucky

Peak Lake, Idaho, Design Memorandum 12, Flip Bucket

Modifications,” Supplement No 1, Outlet Works, Slide

Gate Repair and Modification, Walla Walla, WA, July

1986

U.S Army Engineer District, Walla Walla, “Periodic

Inspection Report No 6, Lucky Peak Lake,” Walla

Walla, WA, Jan 1985

U.S Army Engineer District, Walla Walla, “Periodic

Terzaghi Dam, operational in 1960, is 197 ft high with

a crest length of 1200 ft The earth and rockfill ment consisting of an upstream impervious fill, clay blan-ket, sheet pile cutoff, and multiline grout curtain, isfounded on sands and gravels infilling a deep river chan-nel The dam impounds Bridge River flow to form theCarpenter Lake reservoir, from which water is drawnthrough two tunnels to Bridge River generating stations

embank-1 and 2, located at Shalalth, B.C., on Seton Lake.Terzaghi Dam discharge facilities are composed of asurface spillway consisting of a 345 ft long free overflowsection; and a gated section with two 25 ft wide by 35 fthigh gates Two rectangular low level outlets (LLO), each

210.1R-10 ACI COMMITTEE REPORT

Fig 2.7-Terzaghi Dam Downstream detail of constrictor

ring

8 ft wide by 16 ft high are subject to a maximum heat of

169 ft These outlets were constructed in the top half of

the concrete plug in the 32 ft, horseshoe-shaped diversion

tunnel

PROBLEM

The LLOs were operated in 1963 for about 23 days to

draw down Carpenter Lake to permit low-level

embank-ment repairs Severe cavitation erosion of the concrete

wall and ceiling surfaces downstream of bulkhead gate

slots was observed in the north LLO after the water

re-lease

Dam safety investigations in 1985 identified that the

LLOs were required to permit emergency drawdown of

Carpenter Lake for dam inspection and repair, and to

provide additional discharge capacity during large floods

SOLUTION

The repair consisted of three main categories of work

- repair of damage, improvement to reduce cavitation

potential, and refurbishing gates and equipment

Repair of cavitation damage in the north LLO

in-cluded repair of the walls, crown, and gate slots

Improvements to reduce cavitation potential included1) installing 9-in deep rectangular constrictor frames(Fig 2.7) immediately downstream of the operating gates

to increase pressures in the previously cavitated area, 2)backfilling old bulkhead gate slots and streamlining theexisting LLO invert entrances, and 3) installing piezo-meters in the north LLO to provide information on flowcharacteristics of the streamlined LLO during dischargetesting

Refurbishing gates and equipment included 1) placing leaking gate seals on closure gates; 2) sand-blasting and repainting gates, guides, head covers, and airshafts, 3) cleaning gate lifting rods and replacing bonnetpackings; 4) replacing ballast concrete in north LLOgates and installing ballast cover plates on all gates; and5) refurbishing hydraulic lifting mechanisms of gates

re-Repair concrete was designed to fully bond withexisting concrete Surface preparation included; sawcutting around the perimeter of the damage, chipping toexpose rebar, and installation of grouted dowels Latex-modified concrete was used for all repair work, with steelfiber reinforcement for the cavitation-damaged areas

A total of 26 cubic yards of 3000 psi ready-mixed crete was placed by pumping Maximum aggregate sizes

con-of 3 / 8-in and 3

/ 4-in were used for general repair and vert entrance backfill, respectively

in-The constrictor frames were made from 1

/ 2-in and

3 / 4-in steel plate They were installed in the LLOs bymeans of the following: 1) bolting the constrictor frame

to the existing concrete with a double row of l-in.diameter adhesive anchors at 12-in spacing 2) keying theconstrictor infill concrete into the existing concrete; 3)welding the constrictor frame to the existing gate metal-work in the walls and soffit; and 4) embedding the con-strictor sill shear bar into the existing concrete invert(Fig 2.7)

DISCUSSION

Piezometer readings confirmed that the constrictorframes in the LLOs helped maintain pressures above at-mospheric, indicating that cavitation should not be aproblem in the future

REFERENCES

B.C Hydro, “Terzaghi Dam, Low Level Outlet

Re-pairs-Memorandum on Construction,” Report No EP6,

Vancouver, B.C., Dec 1986

B.C Hydro, “Terzaghi Dam, Low Level Outlet Tests,”

Report No H1902, Vancouver, B.C., Mar 1987.

REPAIR OF EROSION DAMAGED HYDRAULIC STRUCTURES 210.1R-11 CONTACT/OWNER

British Columbia Hydro

Hydrotechnical Department, HED

6911 Southpoint Drive

Burnaby, British Columbia, Canada V3N 4X8

YELLOWTAIL AFTERBAY DAM

Bighorn River, Montana

BACKGROUND

Yellowtail Afterbay Dam, operational in 1966, is a

33-ft-high concrete gravity diversion type structure, 300

ft long, located about 1 mile downstream from Yellowtail

Dam In 1967 following a heavy winter/spring snowpack

in the upstream drainage basin, flood flows passed

through both Yellowtail Dam and the Afterbay Dam

PROBLEM

Divers examined the Afterbay Dam sluiceway and

still-ing basin after the flood flows had passed They found

cavitation damage on the dentates (baffle blocks) and

adjacent floor and wall areas in the spillway stilling basin

Although the cavitation damage was moderate, repairs

were necessary to lessen the likelihood that future

cavitation damage would occur

Damage to the dentates and floor in the sluiceway was

caused by abrasion The relatively low sill at the

down-stream end of the sluiceway was permitting downdown-stream

gravel and sand to be drawn into the stilling area, where

a ball mill-type action ground away the concrete surfaces

In the stilling basin downstream of the reverse ogee

section, cavitation severely eroded the sides of the

den-tates and the adjacent floor areas A similar condition

developed in the sluiceway except that it was caused by

abrasion erosion Since the damage from the two causes

occurred essentially side by side, the situation graphically

illustrated the dissimilar types of erosion resulting from

cavitation and abrasion

SOLUTION

Following the flood, low flows at the dam could be

maintained for only one month That situation required

that all repairs be completed quickly and concurrently In

addition to repairing damaged areas, the downstream sill

in the sluiceway was raised about 3 ft to stop river

gravels from being drawn into the sluiceway Repairs

were completed using a combination of bonded concrete,

epoxy-bonded concrete and epoxy-bonded epoxy mortar,

depending upon thickness of the repair Epoxy used in

this repair was a polysulfide-type material After repaired

materials had been placed and cured, they were ground

to provide a smooth, cavitation-resistant surface

PERFORMANCE

The dam has now been in service about 23 years since

the repairs were made With the exception of a minor

number of spalls, the performance of the repairs hasbeen excellent

REFERENCES

Graham, J.R., “Spillway Stilling Basin Repair Using

Bonded Concrete and Epoxy Mortar,” Proceedings,

Irri-gation and Drainage Specialty Conference, Lincoln, NE,Oct 1971, pp 185-204

Graham, J.R., and Rutenbeck, T.E., “Repair of tion Damaged Concrete, a Discussion of Bureau of

Cavita-Reclamation Techniques and Experiences,” Proceedings,

International Conference on Wear of Materials, St.Louis, MO, April 1977, pp 439-445

CONTACT

Bureau of ReclamationP.O Box 25007, Denver Federal CenterDenver, CO 80225

struc-a high-hestruc-ad spillwstruc-ay locstruc-ated in the left struc-abutment At thisspillway, water enters through a radial-gated intake struc-ture, then passes into an inclined section of tunnel vary-ing in diameter from 40.5 ft at the upper end to 32 ft atthe beginning of the vertical elbow Thereafter, flow fol-lows the 32-ft-diameter tunnel through the elbow and

1200 ft of near horizontal tunnel, exiting into a tion stilling basin-flip bucket, then into the river.During the spring of 1967, heavy rains in the water-shed area of the Bighorn River resulted in high inflowsinto Bighorn Lake behind Yellowtail Dam A total of650,000 acre-ft of flood waters was released through thespillway over a period of 30 days Maximum flow was18,000 ft3 / S

20 ft wide and 6 to 8 ft deep After the tunnel was watered, it was found that a small concrete patch placedduring construction had failed therebv causing the dis-

de-210.1R-12 ACI COMMlTTEE REPORT

continuity in the flow that triggered the cavitation

SOLUTION

The tunnel liner was repaired using several systems

depending on the size and depth of the damage Areas

where the damage extended through the lining into the

foundation rock were repaired with high quality

replace-ment concrete Major areas of damage where the erosion

did not penetrate through the concrete lining were

repaired with bonded concrete Shallow-damaged

con-crete was repaired with epoxy-bonded concon-crete and

epoxy-bonded epoxy mortar Surfaces were ground where

necessary to bring tolerances into conformance with

specifications requirements Finally, tunnel surfaces

below spring line were painted with an epoxy-phenolic

paint, to help seal the surface and bond any aggregate

particles that may have been loosened

In order to avoid recurring damage, an aeration device

was model tested in the laboratory and then constructed

in the tunnel a few ft upstream of the point of curvature

of the vertical elbow This aeration slot measured 3 ft

wide and 3 ft deep and extended around the lower three

quarters of the tunnel circumference It was designed to

entrain air in the flow for all discharges up to 92,000

ft3

/ S, without the slot filling with water A 27-in-long

ramp was constructed upstream of the slot which raised

the upstream face of the slot 3 in at the tunnel invert

Under most flow conditions the bottom of the jet was

forced away from the tunnel floor surface The jet

re-mained free for a considerable distance downstream, all

the while drawing air into the jet from the aeration slot

Aeration has reduced the discharge capacity by

approxi-mately 20 percent

PERFORMANCE

It has now been 23 years since the tunnel was repaired

and the aeration slot installed, but flows in the river have

never been sufficient to require use of the spillway

How-ever, a controlled prototype test with flows to 16,000 ft3/s

was conducted in 1969 and 1970 As a result of this test,

less than one percent of the concrete repairs failed and

no cavitation damage was observed, even in areas

down-stream from discontinuities To ensure that the tunnel

will always be ready for the next flow, there is a regular

maintenance program to repair ice damage and remove

calcium carbonate buildups

REFERENCES

Borden, R.C., et al., “Documentation of Operation,

Damage, Repair, and Testing of Yellowtail Dam

Spill-way,” Report No REC-ERC-71-23, Bureau of

Reclama-tion, Denver, CO, May 1971

Colgate, D., and Legas, J., “Aeration Mitigates

Cavitation in Spillway Tunnel,” Meeting Preprint 1635,

National Water Resources Engineering Meeting, Jan

24-28, 1972, Atlanta, GA, American Society of Civil

Engineers, New York, NY, 29 pp

CONTACT

U.S Bureau of ReclamationDenver Office, Code D-3700P.O Box 25007, Denver Federal CenterDenver, CO 80225

The sluiceway downstream of the gate slot has an ogeesection designed very conservatively for 65 percent of thedesign head Upstream of the gate sill the profile is afairly broad three-radius compound curve Accordingly,

no negative pressures should occur anywhere on the crestunder free discharge operation

PROBLEM

Cavitation damage has occurred on the sluiceway crestnear the gate slots on all four bays The damage ex-tended from inside the upstream portion of the gate slot

to a point about 4 ft downstream, extending at an angle

of about 30 degrees to the direction of flow (Fig 2.8).All attempts to repair the eroded concrete with epoxymixtures and steel fiber-reinforced concrete (FRC) in

1973, 1975, and 1977 were unsuccessful Continued itation soon pitted the repaired areas which later pro-gressed to development of major voids By 1980 approxi-mately 80 percent of the previous repair had eroded.During a high water inspection in 1986, sluiceway No 2was flow tested for 4 hr at gate openings of 4, 8, 12 and

cav-16 ft and full opening Characteristic noises of cavitationbubble collapse could be heard intermittently at all gatesettings The highest rate of cavitation activity wasobserved to be with gate openings from 4 to 12 ft.The deepest erosion usually occurred just outside thegate slot with depths ranging from about 8 to 14 in.Downstream of the badly eroded area, the concrete atthe invert was observed to be roughened for another 2 ft.The maximum width of the eroded area varied from 18

to 24 in

The cavitation erosion at the foot of the gate slotdamaged not only the concrete invert but also the lowerpart of the steel liner within the gate slot and an area ofthe wall immediately downstream of the liner The 1986study concluded that the severe concrete erosion at andjust downstream of the gate slots was due to 1) cavitationcaused by vortices originating in the upstream corners ofthe gate slots at small, part-gate operation; and 2) lack ofrounding and lack of offset of downstream edge of gateslot

REPAIR OF EROSION DAMAGED HYDRAULIC STRUCTURES 210.1R-13

SOLUTION

Initially, it was recommended that 1) eroded areas

should be filled with concrete and armored with steel

plates, and; 2) field tests should be conducted to identify

cavitation zones Later, the recommendation was changed

to backfill cavitated areas with aggregate, high

strength (6000 psi) concrete The bond between the

back-fill and the original sluiceway concretes was enhanced by

epoxy bonding agent The top surface of the new patch

and the surrounding original concrete were coated with

an acrylic latex selected through an extensive laboratory

screening process.The work was carried out in the

sum-mer of 1990

PERFORMANCE

In order to test the effectiveness of the repairs, during

the following year it was decided to operate the sluice

gates mostly in the worst range A year later, the

re-paired and coated surfaces began to show signs of pitting

The performance of the repair still did not appear

satis-factory It became obvious that besides repairing the

eroded areas other initiatives were needed to alleviate

recurrence of the problem

DISCUSSION

Based on the observations of the effect of gate

open-ing on cavitation, it was decided to limit gate operation

to that outside of the destructive range Gate operating

orders were rewritten to require “passing over” the rough

zones as quickly as possible without any sustained

operation in those zones

REFERENCES

B.C Hydro, Hydroelectric Engineering Division,

“Hugh Keenleyside Dam, Cavitation Damage on

Spill-way,” Report No H1922, Vancouver, B.C., Mar 1987.

B.C Hydro, Hydroelectric Engineering Division,

“Keenleyside Dam, Comprehensive Inspection and

Re-view 1986,” Report No H1894, Vancouver, B.C., May

1987

B.C Hydro, Hydroelectric Engineering Division,

“Hugh Keenleyside Dam, Cavitation Damage on Spillway,

Field Investigation of Cavitation Noise and Proposed

Gate Operating Schedules,” Report No 2305, Vancouver,

B.C., June 1992

CONTACT/OWNER

British Columbia Hydro Structural Department

HED6911 Southpoint Drive

Bumaby, British Columbia, Canada V3N4X8

CHAPTER 3-ABRASION-EROSION

CASE HISTORIES ESPINOSA IRRIGATION DIVERSION DAM

EspBnola, New Mexico, on the Santa Cruz River

Fig 2.8-Keenleyside Dam Cavitation erosion of concrete invert and adjacent damage to steel liner Maximum depth approximately 9 in.

BACKGROUND

The diversion dam is a reinforced concrete structurethat is capable of diverting up to 13 f& in the EspinosaDitch for irrigation purposes A 50-ft-long reinforced rec-tangular concrete channel, sediment trap, and sluice gatestructures were constructed between the headgate andthe ditch heading A sidewall weir notch is provided inthe rectangular ditch lining to allow emergency discharge

of flood flows back to the river A 24-in.-round sluicegate at the right side of the dam was placed at the slabinvert elevation, to sluice sand and cobbles through thedam and to prevent these materials from entering the ir-rigation ditch head gate The dam is tied back into theriverbanks on either side with small earthen dikes thatprotect the

diver-ACI COMMITTEE REPORT

Fig 3.1-Espinosa Irrigation Diversion Dam Erosion damage to the floor blocks

Fig 3.2-Espinosa Irrigation Diversion Dam Steel plate protection added to floor blocks and endsill

floor blocks (Fig 3.1) due to impact and abrasion by the

bedload The bedload consists of gravels and boulders

ranging up to 24 in in diameter The concrete in the

apron in the impact area was abraded to a depth of 6 in

Except for very low flows and flows diverted for

irri-gation, the bedload is carried over the weir

SOLUTION

Repairs were made by extensive structural tions These modifications included the following (Fig.3.2): 1) removing and replacing the top layer of rein-forcement in the apron; 2) removing and replacing thetop 6 in of concrete; 3) protecting the apron with a ?&n

modifica-REPAIR OF EROSION DAMAGED HYDRAULIC STRUCTURES 210.1R-15

steel plate; and 4) replacing the 24-in-round sluice gate

with a 36-in square sluice gate

PERFORMANCE

The structure has been operating satisfactorily since

rehabilitation in 1982

DISCUSSION

Five alternatives were evaluated for the placement of

the diversion dam back into service The ones not

selected as the solution are as follows:

1 Install a reinforced concrete lining inside the walls

and apron of the existing structure

2 Protect the apron with a 1 / 2-in steel plate

3 Remove the entire apron of the structure and

re-place it with one that is adequately reinforced Add

the liner inside the structure

4 Remove the entire structure and replace it with a

new one

REFERENCES

U.S Department of Agriculture, “Espinosa Diversion

Dam, Report of Investigation of Structural Failure,” Soil

Conservation Service, Albuquerque, NM, Nov 1980

U.S Department of Agriculture, “Espinosa Diversion

Dam, Design Engineer’s Report," USDA, Soil

Conserva-tion Service, Albuquerque, NM, Sept 1982

CONTACT/OWNER

State Conservation Engineer

U.S Department of Agriculture Soil Conservation

Kinzua Dam became operational in 1965 The stilling

basin consists of a horizontal apron, 160 ft long and 204

ft wide It contains nine 7-ft-high by l0-ft-wide baffles,

located 56 ft upstream from the end sill The

vertical-faced end sill is 10 ft high and 6 ft wide The basin slab

was constructed of concrete with a 28-day compressive

strength of 3000 psi

The outlet works consists of two high-level and six

low-level sluices A maximum conservation flow of about

3600 ft3/s is supplied by the high-level sluices The

low-level sluices with flared exists containing tetrahedral

deflectors are located 26 ft above the stilling basin slab

Bank-full capacity, 25,000 ft3/s, can be discharged through

these sluices at reservoir elevation 1325 The maximum

24,800-ft3/s record discharge was discharged through the

sluices in 1972 The maximum velocity at the sluice exit

was 88 ft/s

PROBLEM

Because of the proximity of a pumped-storage plant on the left abutment and problems from spray,especially during the winter months, the right side sluiceswere used most of the time Use of these sluices causededdy currents that carried debris into the stilling basin.The end sill was below streambed level and contributed

power-to the deposition of debris in the basin

Divers reported erosion damage to the basin floor asearly as 1969 Also, piles of rock, gravel, and other debris

in the basin were reported About 50 cubic yards ofgravel and rock, ranging up to 8 in in diameter, wereremoved from the basin in 1972 Abrasion-erosion dam-age reached a depth of 3.5 ft in some areas before initialrepairs were made in 1973 and 1974

These repairs were made with steel fiber-reinforcedconcrete Approximately 1400 cubic yards of fiber con-crete was required to overlay the basin floor From thetoe of the dam to a point near the baffles, the overlaywas placed to an elevation 1 ft higher than the originalfloor

In April 1975, divers reported several areas of sion-erosion damage in the fiber concrete Maximumdepths ranged from 5 to 17 in Approximately 45 cubicyards of debris were removed from the stilling basin.Additional erosion was reported in May 1975, andanother 60 cubic yards of debris were removed from thebasin At this point, symmetrical operation of the lowersluices was initiated to minimize eddy currents down-stream of the dam After this change, the amount ofdebris removed each year from the basin was drasticallyreduced and the rate of abrasion declined However,nearly 10 years after the repair, the erosion damage hadprogressed to the same degree that existed prior to therepair

Construction of a debris trap immediately downstream

of the stilling basin end sill was also included in therepair contract Hydraulic model studies showed thatsuch a trap would be beneficial in preventing downstreamdebris from entering the stilling basin The trap was 25 ftlong with a 10-ft-high end sill that spanned the entirewidth of the basin

210.1R-16 ACI COMMlTTEE REPORT

0B -CONVENTIONAL CONCRETE, SYLVANIA LIMESTONE AGGREGATE, 5710 PSI (39 MPa)

PENN-0C -CONVENTIONAL CONCRETE, LOS GELES AGGREGATE, 7470 PSI (52 MPa).

AN-0D -SILICA-FUME CONCRETE, LOS GELES AGGREGATE, 11,500 PSI (79 MPa)

AN-0E -SILICA-FUME CONCRETE, VANIA LIMESTONE AGGREGATE, 13,850

In August 1984, after periods of discharge through the

upper and lower sluices, abrasion-erosion along some

cracks and joints was reported by divers The maximum

depth of erosion was about % in The divers also

dis-covered two pieces of steel plating that had been

em-bedded in the concrete around the intake of one of the

lower sluices Because of concern about further damage

to the intake, the use of this sluice in discharging flows

was discontinued This nonsymmetrical operation of the

structure resulted in the development of eddy currents

Sluice repairs were completed in late 1984, and metrical operation of the structure was resumed A diverinspection in May 1985 indicated that the condition ofthe stilling basin was essentially unchanged from the pre-ceding inspection A diver inspection approximately 31 / 2

sym-yr after the repair indicated that the maximum depth of

erosion, located along joints and cracks, was about 1 in

REFERENCES

The next inspection, in late August 1984, found

ap-proximately 100 cubic yards of debris in the basin In

September 1984, a total of about 500 cubic yards of

debris was removed from the basin, the debris trap, and

the area immediately downstream of the trap The rock

debris in the basin ranged from sand sized particles to

over 12 in in diameter Despite these adverse conditions,

the silica-fume concrete continued to exhibit excellent

resistance to abrasion Erosion along some joints

ap-peared to be wider but remained approximately 1 / 2-in

Fenwick, W.B., “Kinzua Dam, Allegheny River, sylvania and New York; Hydraulic Model Investigation,”

Penn-Technical Report HL-89-17, U.S Army Engineer

Water-ways Experiment Station, Vicksburg, MS, Aug 1989.Holland, T.C., “Abrasion-Erosion Evaluation of Con-crete Mixtures for Stilling Basin Repairs, Kinzua Dam,

Pennsylvania,” Miscellaneous Paper SL-83-16, U.S Army

Engineer Waterways Experiment Station, Vicksburg, MS,Sept 1983

Holland, T.C., “Abrasion-Erosion Evaluation of crete Mixtures for Stilling Basin Repairs, Kinzua Dam,