Lake_Youngs-Outlet_Dam_Seismic_Slope_Stability_Evaluation_1981.pdf

200 ,2801

J 19/

ConverseWarciDaVlsDlxon

SEl\llW:. WAfER Dcl'Mllvic.!'H

CENTRAL FILES

SEISMIC SWPE STABILITY EVAlDATICN for

Lake Youngs OUtlet Dam King COlmty, Washington

Consultant Agreement W79-33 th(, ~~~ 4vlI/'w;;

P/~q5t' rt'r~rn t;

:;5;4 ~ Ulalh lJp-r

Anchorage, AK Seattle, WA San Francisco, CA Pasadena, CA Anaheim , CA las Vegas, NV Denver , CO Cincinnati , OH Caldwell . NJ w·~':2~~ William S. Bliton Senior Geologist

Dean E. Ryden Principal Engineer

Professional Certification

SEISMIC SIDPE STABILITY EVAUJATICN

for lAKE YOUNGS OUTIET DAM

KING COUNTY, WASHINGI'ON

This report has been prepared by the staff of Converse Ward Davis Dixen, Inc. under the professional supervision of the engineers and geologists whJse seals and signatures appear herecn.

The findings, recarmendatiens, specifications or professional opiniens are presented, within the limits prescribed by the client, after being prepared in a=rdance with generally ac cepted professional engineerin:J and geologic principles and practice. There is no other warranty, either express or :im plied.

Converse Ward Davis Dixon. Inc.

TABLE OF CDNI'ENTS

SUM1ARY

1. INI'RODUC!'ION

2. DAM AND RESERVOIR HISTORY

2.1 Sequence of Events

2.2 Design

2.3 Construction

2.4 Past Performance

3. SITE COODITIONS

3.1 Geologic Setting

3.2 Surface Conditions

3.3 Subsurface Conditions

4. ENGINEERING S'IUDIES

4.1 General Approach

4.2 Embankment Section and Soil Prc:perties

4.3 Method of Analysis

4.3.1 General

4.3.2 static Case

4.3.3 Seismic Case

4.4 Seismic Coefficient

Page No.

i

4.5 Results of Slope Stability Analyses 19

4.5.1 Pseudostatic Analysis of Circular Slip Surfaces 19

4.5.2 Evaluation of Pseudostatic Analyses 20

5. CDNCWSIONS AND REJXt.1MENDATIONS 21

Converse Ward Davis Dixon, Inc.

Table of Contents (Cont' d)

DRAWINGS

1. Project Location Map

2. Locations of Lake YOlIDgs Darns .

3. Site Plan with Exploration Locations

4. Embankment Section A-A'

5. Embankment Section B-B'

6. Results of Slc:pe Stability Analyses

APPENDIX A Field Explorations

APPENDIX B Construction PhJtographs

APPENDIX C Visual Inspection Checklist

Converse Ward Davis Dixon, Inc, -i-

The D..1tlet Dam at Lake Youngs was constructed in 1921 across Little Soos

Creek which was the outlet stream for the natural lake. The dam has a maximum height of 28 feet and consists of rolled fill with a concrete lining. Construction specifications and pootographs, and our explora tory borings, irrlicate the enbankment was well ccmpacted on a dense foundation free of organics and crmpressible material. In the 1950' s, the crest of the dam was raised fran elevation 493 to 496 with the addi tion of fill and a vertical concrete wall extension of the liner. This recent fill is rrore granular and less dense than the fill placed in

1921. At the same time the crest was raised, additional fill was placed to flatten the downstream slopes. Because the end of the outlet pipe and valves were not extended downstream, protection of these facilities resulted in a localized steep area on the downstream embankment slope.

A section through the embankment at the outlet pipe was selected to represent the least favorable stability conditions in the dam (maximum fill, steepest downstream slopes). Pseudostatic seismic slope stability analyses were =nducted for several possible failure surfaces using em pirical soil strength values and circular slip surfaces. For the maxi mum reasonable earthquake selected for evaluation (Richter magnitude 7.5 at 60 kilaneters directly beneath the site), the 10Nest calculated factor of safety was approximately 0.9 and applies to a shallCYvl slip surface extending approximately four feet belCYvl the surface of the ex terior slope surface. The calculated factor of safety for this sarre slip surface was approximately 1.05 under rrore probable earthquake locrl ings (lOO-year event). Factors of safety for a deeper slip surface extending to the middle of the crest, yet entirely within the recent loose fill, were calculated to be approximately 1.0 and 1.15 under the

Class I and Class II earthquakes selected for this evaluation, respec tively.

Based on the results of these calculations, it is our opinion that the stability of dONnstream slopes on the OUtlet Dam are rrarginal during a

Class II earthquake and are inadequate during Class I earthquake load ings. Altoough the positions of the significant slip circles suggest

-u-that the loose recent fill could slump without affecting the denser em bankment constructed in 1921, such a conclusion cannot be verified because the analyses herein do not include secondary damage effects from earthquakes such as pore pressure build-up, progressive slope failures, damage to liner, and piping. Loss of a major portion of the recent downstream fill could subject the remaining original embankment to water levels higher than originally intended since the lake elevation is now

493 rather than 488.

Because the downstream hazard potential for this dam has been classified as high, it is rea:mrended that the downstream slope, especially in the vicinity of the outlet pipe, be improved to increase seismic stability.

Improvement would consist of removal of loose, recent fill and replace ment with controlled, cc:mpacted fill. The downstream end of the outlet pipe would need to be extended beyond the toe of the new embankment slope and upstream controls should be added to the outlet pipe.

Design of the rrodification to the dam will require additional geotech nical investigation to define the lateral extent of loose fill, to de termine the most efficient revised slope configuration, and to identify and characterize new fill materials. It appears that suitable soils for remedial construction can be obtained . from borrow areas on the Lake

Youngs reservation.

A program of regular inspections and piezaneter readings is reccnmended to monitor the performance of this dam in the future.

1. INTRODUCTICN

This rep:>rt presents the results of our seismic slope stability for darn is located at the south the existing geotechnical investigation of

Lake YOun;js Outlet Darn. The end of Lake Youngs in southwest Kirg County as shown on the Project Location Map, Drawing 1.

The purpose of this investigation was to explore surface and subsurface soil and gramdwater =rrlitions and review available re=rds of reser voir construction and past performance, and based on this informatioo, =rrluct pseudostatic analyses to evaluate slc:pe stability of the reser voir embankments under earthquake loadings. The srope of our work in-cluded the follawin:J: (1) a review of available plans, specifications and inspection re=rds related to construction and performance of the darn; (2) discussions with City of Seattle Engineering and Water Depart ment personnel regarding reservoir operation and dam performance; (3 ) field explorations =nsisting of subsurface b::Jrirgs and surface re=n naissance; (4) static and pseudostatic seismic slope stability analyses;

(5) develcpnent of geotechnical ergineering =nclusions based on our firrlings; and (6) preparation of this report.

The horizontal ground motioos associated with two earthquakes selected for evaluation of the city reservoirs were described in our October 15, 1980 rep:>rt entitled "Rep:>rt on Earthquake Motioos for Reservoir Seismic

Stability Analyses." Available materials reviewed during this evalua tion were:

1. Prop:>sal, Specifications and Contract - SWan Lake Darns, Cedar River Water Supply, City of Seattle Engineering Department, 1920.

2. Three design drawirgs for SWan Lake Darns (with as-built notations) dated October 29, 1920.

Drawing 1 - Vicinity Map Drawin:J 2 - Plans and Profiles Drawing 3 - Details

3. Twelve photographs of Outlet Darn construction provided by City of Seattle Engineering Department photograph lab::Jra tory.

4. Seattle Water Departrrent History 1854:1954, by Mary McWilliams, City of Seattle, 1955.

-2-

2. DAM AND RESERVOIR HISTORY

2.1 SmuEOCE OF EVENl'S

Prior to being used as a water storage reservoir, Swan Lake was a natur al lake covering approximately 584 acres with an average water surface elevation of 470 feet (City of Seattle datum). Maximum water depth in the natural lake was on the order of 80 feet. Little SaJs Creek was the outlet at the south end of the lake. Honey Creek flCM1ed into the east end of the lake.

As authorized by Board of Public Works ordinance nl.lllber 40634 on October

15, 1920, two earth dams were to be =nstructed across the natural drainages at the south and east ends of the lake to raise the lake level to elevation 488 and impound water for municipal use. The Outlet Dam, which is the subject of this report, was to be =nstructed across Little

Soos Creek in the southeast quarter of the southeast quarter of Section

11, 'Ibwnship 22 North, Range 6 East (Drawing 2).

A contract for CXlt1struction of both dams was awarded to the canpany of

Collucio and Erickson on January 3, 1921. Construction of the Outlet

Dam was campleted on September 22, 1921- for a total =st of $76,626.38.

The Inlet Dam across Honey Creek was =nstructed later in 1921 and is addressed in a separate report.

Swan Lake was renamed Lake Youngs in 1924.

Perimeter dikes were constructed during 1933-34 and 1958-62 to raise the level of the lake. A section of these dikes was placed inside the smaller, western segments of the Outlet Darn so that these previously existing dam embankments were abarrloned (Drawin:! 3). In the 1950' s, the major portion of the Outlet Dam was raised three feet with the addition of an inboard =ncretel:ulkhead and associated fill (Drawin:! 4). This addition and the raising of perimeter dikes in 1962 allow the normal lake level to be at elevation 493.

-3-

2.2 DESIGN

The Outlet Dam as originally designed in 1920 included three disconnect ed earth 6llba:nkments with a total crest length of 1280 feet as sh::1wn on

Drawing 3. Maximum heights of enibankments, fran crest to original ground surface along the centerline, were to be 25 feet for the major segment, eight feet for the middle segment, and ten feet for the west ern segment. The original design sectioo, sh::1wn on Drawing 4, irrli cates a crest elevation of 493, crest width of twelve feet, a three-foot by six-foot cutoff trench into hardpan, 2H: IV (horizontal to vertical) downstream slopes, ani 3H: 1 V upstream slopes.

The upstream slope of the enibankments was to be lined with twelve-foot square concrete panels, five inches thick. A 30-inch diameter, cast iron outlet pipe encased in concrete was to be installed beneath the major enibankment in the vicinity of the old stream channel.

2.3 CONSTRUCTICN

Design considerations are reflected by prOll'isions in the construction contract specifications dated 1920. Those sections providing some in sight to geotechnical details are quoted belCM.

Clearing and Grubbing

"The entire areas covered by the enibankments forrni.n:l the dams and of the 00= pits fran which earth for the construction of the dams is se=ed shall be cleared ani grubbed. Clearing and grubbing darn sites shall consist of re:roving and burning all trees, stumps, roots, logs, brush and other debris to a depth of two feet belCM the surface of the stripped areas upon which the dams are to be built. Clearing ani grubbing oorrCM pits shall consist of re:roving and burning all trees, stumps, roots, logs, brush ani other debris fran the surface of the

00= pits." (Page 29)

-4-

Earthwork

a) "Surface Stripping. Surface soil am all earth which con tains vegetable or other perishable matter and all soft or objectionable material shall be reroved fran the dam sites and fran borrow pits and deposited in spoil banks where ordered by the City Engineer. Soil raroved fran dam sites shall, in general, be deposited within the area tc be flooded by the waters of the lake. No dam shall be built on an unstripped area and no surface soil shall be used in the constructon of the embankments." (Page 29)

b) "CUt Off Trench. A cut off trench shall be excavated on the centerline of the dams; its dimensions are sh:Jwn on the detail plans. The cut off trench shall be excavated to a depth of about three feet in the solid hardpan and shall be refilled with lmfervious material which shall be solidly rolled... Special care shall be taken tc ccrnpact the earth on the edges of the cut off trench am tc se=e close contact with the hardpan sides and tc prevent the percolation of water through or along the infilled trench

••• " (page 30)

c) "Rolled Enbankments. Borrow pits shall be opened adjacent tc am south of the Darn at Outlet... The dams shall be constructed of the material krx:Jwn locally as "hardpan" and no earth containing vegetable or other perishable material shall be used. Earth fran the cut off trench sh:Jwn on the plans may be used in the rolled embankment if of proper quality. All rocks and stones which are large eoough tc prevent the proper spreading or canpacting of the earth shall be raroved. The earth shall be spread in layers of not rrore than six inches thickness am each layer shall be thoroughly ccrnpacted by rolling with a roller which shall be equal in effectiveness to a standard twelve-ton steam road roller having a grooved front roll. The roller shall be passed not less than four times over the entire area of

-5-each six-inch layer of earth arrl rolling shall be con tinued until the layer is thoroughly ccrnpacted. No earth shall be dumped in or up::m mud arrl all soft arrl muddy earth shall be reroved." (Page 30)

Concrete Slope Lining

"The water slopes of the dams which are rot covered by rip rap shall be protected by a concrete lining. It shall be five inches thick and shall be laid in slabs which shall be in general twelve feet square, to the shape of the slope."

Outlet Pipes except where rrodified to conform

(Page 31)

"A thirty-inch cast iron pipe shall be laid under the Dam at

Outlet... The pipes shall be laid in trenches excavated in the solid hardpan foundations of the dams arrl shall be entire ly encased in concrete which shall be tamped into the trench so as to ccmpletely enclose the pipe arrl cover it not less than six inches." (Page 32)

Filling Drainage Tunnel

"A tunnel about one hundred arrl seventy feet long has been driven into the hill at the east end of the Dam at Outlet.

The Contractor shall rerrove all timbering fran the portal cut leading to this tunnel and shall slope the sides thereof until all broken or loose material has been reroved, he shall also rerove all planks fran the flcor of the tunnel. The turmel shall be filled with earth sluiced into it through Test pit

Number 122, which is slown on the plans. Free drainage shall be allONed at the lower end of the tunnel through the portal cut during the process of sluicing and it shall be filled ccm pletely full of the sluiced earth. Test pit Number 118 shall also be filled with sluiced earth. Test Pit Number 115 shall be filled with sarrl arrl gravel to serve as a drain for the water in the hill aJ::ove the dam. Earth sluiced into the tun nel shall be se=ed fran the hardpan deposits located above

-6-am east of Test Pit Number 122. A screen, or grizzly, having meshes of such size that all pebbles and clods of earth more than one-half inch in diameter will be rejected shall be placed over the entrai"lCe to the test pit and all sluiced rraterial Which enters the tunnel shall be passed through it."

(Page 33)

Filling Test pits

"The existing test pits on the dam sites shall be filled with earth before the darns are built. The pits shall be pumped out am the timber lagging renoved, they shall then be filled with earth, of the same quality as that used in the construction of the darns, Which shall be thoroughly tamped in layers of not more than four inches thickness." (Page 33)

Inspector's records were not available to indicate the extent to Which provisions in the specifications were carried out. H:Mever, a series of protographs taken in 1921 provides a unique opp::>rtunity to observe the construction practices and equipnent. Copies of these protographs are included as Apperrlix B. Interpretation of the =nstruction protographs and as-built notes on the design drawings indicates the =nstruction procedures describOO belo.v.

Site =rrlitions prior to construction are illustratOO by protographs (a) and (b) taken on February 18, 1921. The area appears to have been log goo and a small tunnel am test pits had been excavated in the proposed left (east) abutment area. The puipcse of this tunnel is rnt clear, altrough the written description on pootograph (b) refers to it as a

"drainage tunnel." Test pits were reportedly dug by laborers in the abutment areas to verify the presence of "hardpan" for a seepage cutoff and for borrow material. This tunnel and the test pits were apparently filled as indicated previously in the specifications.

Photograph (c) srows stripping operations with rorse-drawn plo.vs. As built notations on the design drawings indicate the depth of stripping was one to seven feet belo.v the original ground surface. In protograph

Converse Ward Davis Dixon,lnc.

-7-

(d), the project area appears to be stripped quite well • . The dam folm dation area was reportedly kept dry during construction by drawing the lake down arrl diverting inf10N aramd the lake.

Photographs (d) arrl (e) are of the excavation of the cutoff trench. The trench appears to be larger than required on the plans arrl this is con firmed by as-built notes Which irrlicate a trench width of twelve feet and a depth of three to five feet below the stripped surface.

The contract documents were arranged so that the bidders provided lmit prices for both rolled arrl hydraulic fill. The successful bidder, Co1-

1ucio and Erickson, submitted identical prices for both types of fill.

Fortlmate1y fran the standpoint of seismic stability, the embankment was placed as rolled fill as srown on photographs (f) through (k). These plotos indicate that bo=ON material was excavated fran a pit in the left abutment by a steam shovel and the soil was brought to the fill by wagons. Fill was spread by lorse-drawn screeds arrl was canpacted by a steel drum roller. Based on the apparent thickness of lifts and lim ited penetration of Wheel tracks, the fill srown in these plotographs appears to be well canpacted.

Photographs (m) and (n) shJw the downstream and upstream slopes nearing oamp1etion in July arrl August, 1921.

Based on =rrli tions irrlicated by these photographs arrl pra.risions in the specifications, =nstruction appears to have been adequate even by rrojern standards, with thorough raroval of organic soils arrl systanatic spreading and canpaction of fill materials. It is also favorable that the =nstruction period was during the dry season (April through August) since the "hardpan" materials used in construction are moisture sensi tive arrl do not oompact readily When wet.

If the tlmne1 arrl test pits near the left abutment were excavated in very dense glacial soils and were then backfilled with sluiced till as specified, seepage through this area is not expected to be excessive because the backfilled excavations would be localized areas of wet, less

-8-dense soil surrounded by dense, relatively irnpmneable natural soils.

Settlement of the embankment over the backfilled tunnel has rot been rerorted. The remaining thickness of intact glacial till between the base of the embankment and the rcof of the tunnel was apparently at least nine feet and all timbers and debris were reportedly removed fram the tunnel prior to filling completely with sluiced fill.

Evidence exists that =arse, water-bearing soils were encountered in the left abutment. First, the tunnel which existed prior to darn construc tion was refe=ed to as a "drainage" tunnel on the pootographs and in the specifications. Second, an as-buH t notation on sheet 2 of the design drawings indicates a "rock pocket" at station 2+59 with three inch and four-inch pipes extending fram the rock pocket to a point down stream of the embankment. Third, test pit nurrber 115 was specified to be "filled with sand and gravel to serve as a drain for the water in the hill located a1:ove the darn."

The impact of this water and =arse soil on the present Outlet Darn is not clear because of the limited information available regarding these features. It is possible that containment of water flows with irnpenne able fill could result in undesirable hydrostatic pressures within the embankment or abutment. It is also possible that this rock pocket was an isolated concentration of =arse, water-bearing material within a till matrix that flowed tenporarHy and was not recharged. Witlout more detailed study, it has been assumed for purposes of this slope stability evaluation that this feature does rot represent a threat to the inte grity of the darn enbankrnent since no observations of seepage or insta bility in this areas were reported and the higher ground to the east of the left abutmmt was removed as borrow material.

As-built notes on the drawings indicate the outlet pipe was placed beneath the embankment at station 4+55.9 with the botton of pipe at ele vation 465.0 at the intake and at elevation 464.7 in the outlet valve chamber.

-9-

Plans, specifications, and inspector's notes were not available for the interior dike which replaced the two western segments of the original

Outlet Dam, or for the recent =ncrete wall extension and additional fill on the Outlet Dam.

2.4 PAST PERFOR'1ANCE

Although specific inspection reports were not available, the Outlet Dam is reported by Water Department personnel to have operated satisfactor ily since construction. The larger earthquakes in the Puget Sound area

(on April 15, 1949 and April 29, 1965) were observed to crack perimeter dikes and the new inlet dam (Cascades Dam) as in::J.icated in our Report on

Dike Re=nnaissance - Lake Youngs Reservoir, dated September 30, 1980.

lbwever, no reports of damage to the Outlet Dam were in::J.icated for these same seismic events. This performance :inplies suitable Outlet Dam en bankment stability under past static and seismic loadings.

-10-

3. SITE CXlNDITIOOS

3.1 GEOIroIC SEITING

The Lake Youngs Reservoir site is locatErl on the Covin:Jton Drift Plain topographic unit of southwestern King County. The drift plain is an uplarrl surface within the Puget lDwlarrl which is underlain by a sequence of glacially-related derosits. The Covington Drift Plain rises several hundrErl feet above the Duwamish, CErlar, and Green River Valleys which round the area on the west, north and south sides, respectively.

The Puget lDwland has been subjectErl to at least four separate periods of continental glaciation during the Pleistocene Epoch. Durin:J these periods of glaciation and accompanying non-glacial periods, a sequence of unconsolidatErl rraterials was derositErl.

The ground surface of the drift plain is m::x'ierately drumlinizErl, exhi biting an urrlulating and locally irregular surface characterizErl by northwest-southeast tremin:J swales and hills. The trerrl of these fea tures irrlicates the direction of movanent of the ice during the last glaciation.

In the Lake Youngs area, the sErliments are anticipatErl to overlie bErl rock of the Puget Group, a series of Eocene age sedimentary rocks. The thickness of the overlying sediments is highly variable, rangin:J fran a minimum of al:x:>ut 50 feet approxirrately on~arter mile northeast of the north em of the lake (at the site of the presently non-operatin:J New

Lake Youngs Coal Mine) to a maximum of several hurrlred feet in other areas.

Res ults of available geologic rrapping and a limited subsurface investi gation corrlucted during this study indicate that near-surface soils beneath the Lake Youngs Reservoir were derositErl during the last glacial period, krxJwn as the Vashon Stade of the Fraser Glaciaticn, which occurrErl approxirrately 15,000 to 13,500 years ago. These near-surface soils consist of glacial rraterial derosited as a ground moraine. The gramd rroraine consists of a thin, relatively continuous rrantle of

Converse Ward Oavis Dixon, Inc.

-11-

Vash:m glacial till anj in sane areas grades to stratified drift. Gla cial till is a relatively impervious, unsorted mixture of silt, sanj, clay anj gravel which was deposited directly beneath the glacier anj is camonly referred to as "hardpan." Stratified drift is sorted and layered sanj and gravel deposited by meltwater streams, probablyeminat ing fran an advancing glacier. The glacial till and urrlerlying sedi

!rents have been =npressed to a dense state by the weight of overriding glacial ice.

Generally, outwash and/or lacustrine (interglacial lake deposits) soils occupy the lON areas ar=d the lake. Specifically, these areas are at the outlet structure at the north end of the lake, the cascades Dam and abarrioned inlet dam on lbney Creek along the east side, and at the

Outlet Dam on Little Soos Creek at the south end of the lake. These deposits were encountered in explorations =npleted during this investi gation at the cascades Dam, abandoned inlet dam, and the Outlet Dam.

These soils are also expJsed in existing borrON pits near the dams.

Based on explorations at the abandoned inlet dam and Outlet Dam, the outwash/lacustrine deposits are at least in part overlain by glacial tilL The dense to very dense nature of the outwash/lacustrine soils strongly suggest that they were overridden by glacial ice.

Since the retreat of the ice, the major geolcgic process has been ero sion of the upland surfaces and deposition in the lON-lying areas. Lake

Youngs, the largest of several lakes which are present on the uplanj surface, occupies a natural topcgraphic lON.

3.2 SURFACE CONDITIONS

On June 23 and December 15, 1980, visual inspections of surface oorni tions at the Outlet Dam were =npleted by members of the CWDD senior reconnaissance team. The ccmpleted inspection checklist is included in

Apperrlix C.

The observed dam configuration was essentially unchanged fran the rrodi fied section described earlier. A localized steep area on the d0wn stream slope was noted above the outlet pipe aligrnnent. Evidently, when

-12-the dam was raised in the 1950's with the addition of the fill to the crest and the downstream slope, the downstream end of the outlet pipe was not extended so that fill could not be added to this area to flatten downstream slopes. Therefore, above the outlet pipe, the do.vnstream slope is locally at abo.lt 1.6 to 2H:IV. The remaining do.vnstream slopes are at alxmt 2.5 to 3H:IV.

No signs of instability were noted in the OUtlet Dam slopes. No seepage was observed in the do.vnstream slopes. Seepage at the outlet pipe was evidenced by a valve well full of water and slight seeps around the concrete structure at the tenninus of the pipe. The water observed was clear and no slope erosion was observed at the seepage points. Discus sions with Water Department personnel revealed that some of this seepage may be fran a leaking four-inch bypass valve.

The downstream slopes of the OUtlet Dam embankment were =vered with grass. The presence of several small (three-inch diameter) tree stumps indicated that trees had recently been removed.

The crest of the embankment was gravel surfaced and is maintained as an access roadway.

Only the upper portions of the upstream slopes could be observed because the lake was at the nonnal operating level. The visible sections of the concrete lining and the three-foot bulkhead wall appeared to be uncrack

ed. The east and west errls of the upstream slopes, beyond the concrete lined center section, were protected with rip rap of 10- to IS-inch dia meter rounded cobbles and boulders. Sane minor erosion of this rip rap was noted frem wave action.

'TWo siphon pipelines carried water frem the lake, over the crest of the

OUtlet Dam, into the downstream channel to maintain the required flow in

Little Soos Creek.

3.3 SUBSURFACE <XlNDITlOOS

Subsurface soil and groundwater corrlitions in the maximum embankment area were explored by a series of four borings to depths ranging fran 28

-13-to 48 feet at the approximate locations sh:>wn on Drawing 3. Piezaneters were installed in all borings to monitor grourriwater level fluctuations.

A description of the field explorations, piezaneter installations, grourriwater level measurenents, and sll!l1Tla1:Y logs of the borings are pr~ sented in Apperrlix A.

The soils encountered in our borings consisted of fill over in-place glacial deposits. The thickness, content, and density of the material observed verify our interpretation of construction described previously.

A cross section drawing of subsurface materials interpreted fran the borings am 1920 construction plans is sh:>wn on Drawing 4.

Surficial soils in our three borings on the crest of the dam consisted of gravelly sand to samy gravel fill in a dry to moist and medium dense to dense corrlition. The thickness of this material was about four feet in LY-l am LY-2, and was six feet in LY-4. It is interpreted t-lJat this fill was placed in the 1950's during the raising of the dam am is associated with the bulkhead wall constructioo. Similar granular fill was observed on the downstream slcpes of the ~t.

A secorrl type of fill, consisting of medium dense to very dense silty sam, was encountered beneath the upper granular fill in the three borings through the dam enbankment. It is interpreted that this fill was placed in 1921 during the original Outlet Dam constructioo. The soil has the appearance of glacial till, or "hardpm" as required in the construction specifications and is generally more dense than the upper fill. Thicknesses of this older fill were observed to be 16, 27 am 25-

1/2 feet in borings LY-l, LY-2, and LY-4, respectively. The elevation of the base of the fill (top of natural soil) at these same locations was approximately 481, 468, and 470 feet Which is within one foot of the elevation of ground after stripping irrlicated by as-built notes on sheet

2 of the 1920 design drawings.

In-place glacial till was encountered beneath the fill in all three borings through the dam ~t and these borings were terminated in this deposit at depths of 29, 43 and 48 feet below the crest. The till was classified as very dense, silty sarrl with some gravel. No organics

-14-or loose material were observed at the fill/natural till =ntact, Which verifies the interpretation of construction photo:Jraphs that stripping was thorough.

A fourth boring, LY-3, was drilled beyorrl the downstream toe of the dam on a fill berm to the west and al::xJve the outlet valve. Eight feet of sandy gravel, sand, and silty sand fill were en=untered from the sur face in this boring. This was not high quality fill because it was loose and =ntained organic fragments. It is interpreted that this material is associated with the 1950's modifications and was added to flatten downstream dam anbankment slcpes and provide an access road to the outlet valve house.

In-place glacial till was observed beneath the fill in LY-3. The till was penetrated at a depth of 26 feet (approximately elevation 453) and was observed to be underlain by lacustrine derosits of hard, laminated, clayey silt. Boring LY-3 was terminated in this fine-grained soil at a depth of 32.7 feet (elevation 446).

Groundwater levels were measured in all of the piezometers as tabulated in Apperrlix A.

The piezometer readings indicate that the groundwater level is within the dense glacial foundation soils. Although some rroist to wet sanples of the fill were =llected, it is =ncluded from the piezometer indica tions that these are local zones within the fill and that a phreatic surface does not exist within the anbankment. This =nclusion is =n sistent with the excellent =rrlition of the upstream slope lining and the absence of seepage on the downstream slope observed during our in spection.

-15-

4. EN3INEERING S'IUDIES

4.1 GENERAL APPRO'\.Cll

In accordance with the scope of this contract, seismic analyses employed for this evaluation were limited to pseudostatic slope stability cal culations based on empirical approximations of soil strengths. '!he results of these analyses are surmarized as factors of safety against slcpe failure both prior to an::1 during an earthquake for three general slip surface configurations.

A critical con:lition for interior anbankment slopes ~uld be rapid draw d= during an earthquake. However, dra,.u= of lake Youngs is slow an::1 the lake is maintained at a constant level by the plant operators.

Therefore, this unlikely canbination of events was not considered appro priate for evaluation of the lake Youngs Outlet Dam. A consideration of interior slcpe stability under normal operating corrlitions indicated relatively high factors of safety. Since similar considerations of ex terior slopes under the same corrlitions resulted in significantly lower factors of safety, the seismic case evaluated for this investigation was that of exterior slcpes under normal operating corrlitions.

'!he embankment section above the outlet pipe was selected for analysis.

Only circular slip surfaces were examined. In our opinion, the profile an::1 soil properties associated with the embankment section eval=ted suggested that the possibility of a wedge-type failure mechanism or failure along other types of non-circular slip surfaces was unlikely.

'!he analyses focused on three general slip surface COnfigurations as follows:

Shallow

Deep

- A surficial, near planar slip surface.

- The soil mass within the slip surface includes a major portion of the exterior ernbanknent slope.

Deep Massive - A slip surface Which intersects the maximun stored water level.

-16-

Each of the three categories were initially analyzed under static con ditions. Seismic stability was then evaluated by applying a range of pseudostatic forces to the most significant slip surfaces for each cate gory determined fram the static analysis.

Descriptions of our stability analyses and a surrrnary of the results are presented in the following sections.

4.2 EMBI\NKMENI' SEl::TION AND SOIL PROPERTIES

'!he Oltlet Dam enbankment can be described as a hancgeneous rolled fill on a strong, iropenneable foundation with concrete lined interior slopes.

Based on the 1921 as-built notations and recent field measurements, the anbankrnent is at maximum height CNer the outlet pipe and also has the steepest da.vnstream slcpes at this fOint as described in Section 3.2.

Therefore, an embankment section at the outlet pipe was chosen for anal ysis.

Soil perarneters for slcpe stability analyses were estimated primarily fram anpirical relationships between the Standard Penetration resis tance, grain size distribution, relative density, and shear strength.

Effective static strength pararreters '"ere used and were assumed to renain constant for the seismic analysis. Drawing 5 ShONS the embank ment cross section and soil properties used in the static and seismic stability analyses.

4.3 METHOD OF ANALYSIS

4.3.1 General

The rrodified Bishop method of analysis was used to evaluate slope sta bility. This method is considered a limit equilibrium method wherein the following approach is used:

1) A circular fOtential sliding surface is assumed and the soil nass contained within the slip surface is divided into a number of vertical slices.

-17-

2) The shearing resistance required to equilibrate the slip is calculated by assuning each slice is in a state of limiting equilibrium and that the failure criterion of the soil is satisfied everyWhere simultaneously along the proposed sur face. In evaluating the slice equilibrium =rrlitions, the rrodified Bishop rrethod assurres no shear is developed be tv.Jeen slices and thus does not satisfy force equilibrium

=rrlitions in the horizontal direction. It has been fourrl, however , that this method gives nearly the same results as the most rigorous methods Which satisfy all =rrlitions of equilibrium, with extreme variations of about eight per cent. Approximate static equilibrium of the potential slide mass is determined by the surmation of forces and m:::xrents for all slices.

3) '!he calculated shearing resistance required for equilibriun is canpared with the available shear strength of the soil.

The factor of safety against slope instability is expressed as the total available soil shear strengths divided by the shear resistance required for equilibrium.

4) The significant slip surface is identified by searching for the circular surface with the 10Nest factor of safety.

4.3.2 static Case

Ccrrputer solutions were utilized to evaluate the static stability of the three failure =rrlitions described earlier. '!he ccrnputer program BISHOP employs the rrodified Bishop method of analysis and was used to evaluate circular slip surfaces. '!his program =ntains subroutines Which provide a variety of options including: a search for the most critical slip surface, canputations of pore water pressures within the embankment, a limitation of slice depths below the slope surface and a pseudostatic approximation of earthquake effects. For this study, the slice depths evaluated were at least four feet below the slope surface.

Converse Ward Oavis Dixon, Inc.

-18-

4. 3. 3 Seismic case

Pseudostatic analyses consist of representing earthquake-irrluced iner tial forces in enbankments by an additional horizontal static inertial force acting through the center of mass. Embankment stability under earthquake loads is approxi.rre.ted by applying a single pseudostatic force to the soil mass being considered and the factor of safety is then cal culated by conventional static methods (Le., rrodified Bishop rrethod).

'!he magnitude of the pseudostatic force is approxi.rre.tely related to anticipated earthquake groUIU accelerations by means of a seismic coef ficient. A discussion of the seismic coefficients used for this study is included in the following section.

The BISHOP cCIlputer program contains a pseudostatic force option and was used to perform pseudostatic stability analyses on the IOC>re significant configurations of the three general slip surfaces considered in the sta tic stability analyses.

4.4 SEISMIC COEFFICIENT

In pseudostatic slcpe stability analyses, the horizontal seismic force is nonnally assumed equal to the mass of soil above the sliding surface times the effective rigid bcdy horizontal acceleration or, as IOC>st fre quently expressed, a "seismic coefficient", k, multiplied by the weight of the potential sliding soil mass:

W Fh=-~=kW g

The seismic coefficient should be regarded as an overall average effec tive multiplier for estimating the resultant pseudostatic lateral force.

Pseudostatic analyses are often used to approxi.rre.te the seismic stabil ity of reservoir anbankrnents.

Based on our experience and judgement, the seismic acceleration for this study was assumed as half the peak horizontal ground acceleration for tlose earthquakes previously selected for evaluation of the reser voirs. The maximum seismic coefficient for evaluation was 0.17 since

-19-the maxim..nn reasonable earthquake, as designated in bur earthquake notion report dated October 15, 1980, was a magnitude 7.5 event which was estimated to cause a peak horizontal ground acceleration of 0.35 g.

To gain a canparative indication of ernbanJcrnent behavior under seismic shaking less severe than the maxim..nn reasonable event , a Class II 100-year earthquake was suggested for analysis . The seismic coefficient for this earthquake was estimated to be approximately 0.1. It was assumed that these seismic coefficients were average values acting over the total height of the enIDanlarent.

4.5 RESULTS OF SLOPE STABILITY ANALYSES

The results of static and pseudostatic slope stability analyses are surnnarized in Table I and sh:Jwn on Drawing 6. Since static stability analyses for interior slopes indicated factors of safety greater than three for normal operating conditions, seismic stability was evaluated for exterior slopes only.

4 . 5.1 Pseudostatic Analysis of Circular Slip Surfaces

Calculated factors of safety for three circular slip surfaces wiL~ seis mic coefficients ranging from 0.05 g to 0. 20 g are sh:Jwn on Drawing 6 .

Table I includes specific factors of safety for Class I (maxim..nn reason able) and Class II (maxinurn probable) earthquakes .

TABIE I

calculated Factors of Safety for Exterior Embankment Slope

Factors of Safetx

Shape of Potential Static Class II Class I Slip Slip Conditions Earthquake Earthquake

Surface Surface 0 . 0 g 0.10 g 0.17 g

Circular Shallow 1.3 1.05 0.9

Deep 1.5 1.15 1.0

Deep Massive 2.0 1.5 1.3

-20-

4.5.2 Evaluation of Pseudostatic Analyses

The slip surface with the la.vest calculated factor of safety is repre sented by a shallow circular surface in the downstream slope as shown on

Drawin:::r 6. Based on the results of these pseudostatic analyses tabu lated above, it is our opinion that the stability of the Outlet Dam embankment is marginal durin:j ani after the 100-year earthquake selected for evaluation (Richter magnitude 6.0 with peak ground a=eleration of

0.2 g). Satisfactory performance of the Outlet D3m has been reported by the Water Department after the 1949 Olympia and 1965 PUget Sound earth quakes. The peak ground acceleration at Lake Youngs durin:j the 1949 earthquake is calculated to be less than the lOO-year a=eleration se lected for this evaluation. The actual peak acceleration felt at the subject site during the 1965 earthquake is estimated to be similar to that fran the Class II earthquake for this evaluation. Therfore, past experience indicates that the Outlet Dam has perfonned satisfactorily durin:j a Class II earthquake even though the calculated factor of safety is marginal.

Based on the results of our slope stability ccmputations, the integrity of the steepest dONnstrearn slope of the Outlet D3m appears to be inade quate during and after the Class I (lO,OOO-year) earthquake selected fur evaluation.

Both of the potential slip surfaces described as "shallCM" ani "deep" are through the loose fill added in the 1950's and do rot extend into the more dense fill placed in 1921. Although this may suggest that the loose fill oould slump without affecting the original enbankment, this conclusion cannot be verified because the type of analysis ccmpleted does not include secondary effects such as pore pressure buildup, pro gressive slope failures, development of a phreatic surface (as a result of liner damage), and pipin:j. LDss of most of the recent fill, as im plied by the shape of the "deep" slip surface, could result in the renaining enbankment being subjected to greater loads than th:Jse con sidered in the original design since the normal lake level is nCM eleva tion 493 rather than 488.

• .. 1 • _ ... .' • < • <

RESERVOIR LOCATIONS ...... , ...

CD BEACON RESERVOIRS

CD MAPLE LEAF RESERVOIR

CD MAGNOLIA RESERVOIR

0 GREEN LAKE RESERVOIR

CD WEST SEATTLE RESERVOIR

0 LAKE YOUNGS INLET DAW

CD LAKE YOUNGS OUTLET DAW

0 LINCOLN RESERVOIR

0 LAKE FOF(EST PARK RESERYOIR

® BITTER LAKE RESERVOIR

® MYRTLE FrESERVOIR

® VOLUNTEER PARK RESERVOIR

' .. '11

REFERENCE: KING COUNTY STREET BASE-METRO TRANSIT,

DATED JANUARY 1973.

'-L' +i,-"'l,L;..;.J;.._-,~",,- ' ~~K

2 o 2

SCALE IN MILES

PROJECT LOCATION MAP

Sc.'. Noted Pro;'C1 No, LAKE YOUNGS OUTLET DAM

King County. Washington OI"FEB 1981 79-5220-09 for City of Seattle Engineering Department Prepared by BH Drawing No.

Checked by DER 1 ~by DER

@ ConverseWardDavasDlxon a-eehnic., Conoultanto

A//of: ~;'...:oJ) '10:.. ,'.' ~.'.'~ ..' ~, • , . _ -., • • • -, -

I 'I

I il ;t

E c , o

" o o z

• • o

I

LAKE YOUNGS

RESERVOIR

/'\ I

I I l-. __

\ () OUTLET JI

\w(~~uth) 1---

I 1 , DAM LOCATIONS

LA KE YOUNGS OUTLET DAM

King County, Ilashington

Shady Lake

( ___ .J

- __ .---.J

INLET

(Cascades)

DAM

Projecl No.

79-5220-09 for City of Seattle Eng i neering Department

@ ConverseWard DavIs Dixon Geotechnical Consultants

Drawing No.

Abandoned Outlet Dam Embankments

Existing Perimeter Dike and Access Road

LAKE YOUNGS

concrete lining unpaved access road to S.E. 224th St.

REFERENCE: 1) "Swan Lake Dams-Plans and Profiles" sheet 2 of 3-0ctober 29, 1920

2) "Lake Youngs Reservation" file 68-3, serial 2782, City of Seattle Water Dept., dated April 1967.

Ex ist ing Perim eter Dike and Access Road

B V-:--30" ct> outlet pipe

A <t of filled drainage tunnel

L TP-118 •

-----f---... TP-122 • LY-2

TP-115

original outlet Dam presently in use ( " South Dam" )

Over steepened area

LY-~

TP-115'

LEGEND

Indicates approximate location of boring by CWDD

Indicates approximate location of Test Pit backfilled in 1921 .

50 o 50 100 200

SCALE IN FEET

stream flow SITE PLAN WITH EXPLORATION LOCATIONS

LAKE YOUNGS OUTLET DAM

Scale

NOTED

Project No.

" 79-5220-09 r u ti ... NU. 11 33/1 0 MJ.if.l I VVUV IV I !-'UV .... dl l O l ' -=:..J.Y~""- V)' ,..-

Z

-<> '" r 0 ..... · :J::t ~::l", '" rrI n ...... n-< r+OCl ,<ceo

::l :z Or+G> -<>'< U1

U1 0 (1;)~C

""" --< r+ V> r r+ :y rrI --' -'. --I ro ::l

'" 0 rrI r+ :x> ::5 0 3:

'" ::l ~.

::l ro ro

::l o ro -0 r+ ffi ::l r+

'" I

'" N N Cl I

Cl

Z

A ~ I

...J elev. 496

I

...J

A' normal lake elev. 493 9 existing embankment configuration (approx.

avg. ) dense SILTY SAND w/gravel (F ill) l oose to dense, SANOY

1920 des i gn __ '--._~ GRAVEL & SI LTY SAND '--. '--. ( F i 11 ) --1----

}.--yt= yh2~ ~ cutoff trench (from 1920 as-bu i lt notes) very dense SILTY SAND w/gravel (Glacia l Ti ll )

(Lacustrine Depos its)

EMBANKMENT SECTION A- A'

Sca l e: 1"=20 ' hard, CLAYEY

SILT

, - .... " ... I"'" _ u ,,:.>1 I tI J-IjJjJI UV UU Ii.'f J)I) O I,Ci:HIOIi - ;' - - f -uy _'_'" __ _

01 3'

" z

-t, 7< r 0 ... · );::.

-':::07<

<0 ,., n "'-n -< rt- 0 0 c C :::I :z ort-G> -t,'< en en 0 ro::::::c "''''---l rt- '" r rt- ::r ,., --' ...... --I ro :::I

<00 ,., rt- :» :::003:

:::0

.., ...., " '" ii' , !l

'" Z N ?

N 0 , sz e 1 ev ,_ 4_9_3 ____ _

Natural Glacial Till ~ ' =43°, y m =140 pcf

1921 Ti 11 Fill

~ '=36o, Ym=135 pcf cutoff trench

EMBANKMENT SECTION B-B'

Recent Granular Fill

~'=32°,y =125 pcf m

30" diameter outlet pipe invert elev. approx, 465

(Sl ope configuration and soil parameters used in slope stabil ity ana lyses) Scale: 1"=20'

8', Static factor of Safety = 3.6 , \} Nor ma L a k e L e v e l

I I

/1 / 1 I ' I / .

2 . 0 L·'

(f) 0 (f)

-' rc :J «

1 5 UJ

u.

'" 1 0 u.

c:

u.

G 5j o +-----"~-- ---'1 -----" ---' ---, o 0 . 05 0 . 10 015 020

PSEUDOST A TIC SEISMIC COEFFICIENT

50~

I-

490 w

UJ

u. ::;

Z ::J - .... r 4&0 z «

I- .... « tHO > <)

UJ

w

4 (,(1

RESULTS OF SLOPE STABILITY ANALYSES

fo r

LAK E YOUNG S OUTLET DAM

K ing Co unty, Was h ington

Ci t y of Seattl e Enginee ri ng Depar t ment

~ W Converse Ward DavIs Dixon Geotechnical Consultants

Pr('pared ~} .

.JM Drilw lng N o

DER 6

DER

APPENDIX A

FIELD EXPLORATIONS

Drilling and Sampling

Subsurface =n:1itions were explorerl by rreans of four test borings at the approx:imate locations sh:Jwn on Drawing 3. '!he borings rangerl fran 29 to

48 feet in depth and were drillerl with a truck-rrounterl drill rig using a

3-5/8 inch inside diameter, rollO'o>l-sten auger.

'!he explorations were lcx:aterl by our Project Engineer who =ntinuously lO3'gerl the explorations, examinerl arrl classifierl materials recovererl, evaluaterl drilling progress in:1icating variations in soil =n:1itions, and selecterl intervals for obtaining representative soil samples. '!he soils en=untererl are sh::Jwn graphically on the boring lO3's, Drawings A-l through A-6. Soils are classifierl in ac=rdance with the Unifierl Sys ten, which is describerl on the chart follO\>ling the boring lO3's.

Samples were obtainerl in a disturberl =n:1ition with the "Starrlard Pene tration Test"' using a 2-inch O.D. split spoon sampler. '!he sampler was driven with a 140-poun:1 hammer falling 30 inches and was generally driven 18 inches at each sample interval. '!he arrount of re<XNery and resistance to driving for each 6-inch interval are sh::Jwn on the boring lO3's. '!hose samples not canpletely used for testing will be storerl in our laboratory for a period of three rronths and are available for exami nation should it be desirerl. Due to the elapserl time since they were obtainerl, changes in rroisture =ntent and appearance of the samples have occurrerl.

Piezareters

Piezaneters were installerl in the borings canpleterl at the site so that groundwater level observations =uld be made follO\>ling drilling opera tions. '!he piezometers consist of a one-inch diameter, slotted plastic tip with a 3/4-inch PVC riser pipe to the surface. '!he piezcrneter tip was installerl at the desirerl depth and pea gravel was placerl as backfill

A-2 around the tip. 'lhe renainder o f the boring was backfilled with cut tings fran the test holes and a bentcnite seal was placed at the ground surface. Cast iron ITOnuments (with a "MON" cap designation) were placed at the surface of each piezometer location fOr long-term protection of the starrlpipe. All ITOnuments can be located at approximate pcsitions shown on Drawing 3.

Although piezameter measurerrents were collected throughout the period of this evaluation, it became apparent that the readings were not consis tent with conditions observed in the borings. Therefore, on February

13, 1981, all of the piezaneters were read and then filled with water tc check fOr plugging. On February 17, the stabilized water levels were recorded. 'lhe table below includes thJse measurements Which are con siderable reliable.

Date

2/13/81

2/17/81

GROONl);'lATER lEVEL MEASUREMENI'S

(elevation in feet)

Boring Numbers

Lake Level LY-l LY-2 LY-3

489.0 454.8 448.5

488.2 471.1 454.8 448.1

Converse Ward Davis Dixon, Inc, LY-4 plugged plugged

DA.TE DRILLED: June 23, 1980 SUMMARY: BORING NO. L Y -1 ELEVATION, 496.2 TNI' .U ..... . II' ..... LIl. ONL' .T TNf. LOC.TlOII 0' TN .. 1011'"0 ... O.T TNf. TUIE OF OIlILLlIlCO. ·

'U"UII".CIE COIiOlnOIi. 'U.' Ol"f.1I AT OT"I[II LOCA"ONS AII O "''''' CII "'IIGIE AT TNIS LOCATION

WITH T HI! ''''',AOf. 0' " "' f. . '"I OA'A '1I1.EIITI! O IS .... t""LlftC .. "O" 0 ' ACTU.L COIIOllIOII'

ENCOUIITEIIEO. ~.~

I.WU

SYMBOL MOISTURE CONSISTENCY---DESCRIPTION , ... 0

lA I~ GRAVELLY SAND brown, well graded . SW (Fill )

2A 7 4~ ----- ---- -- - ----

5- 3A 7 gray-brown, locally

1~ very moist

I

!~ 4A

- SILTY SAND SM

0- 5A 14 (Fill )

6A 21 ----------------

5- 2g trace of charcoal.

----------------7A gray

8X 1~

· 14 0- 9A 30

- 51 bro~n , highl y weath-~ ...•

-lOA 50/4" SILTY SAND ere .

S~l g-ra~ ~ -we 11 -gr-aded,--- (Glacia l wit gravel. -5- Ti 11 ) unweathered .

-l1A 61/6"

70 ·Im Ig~}:, increasing sand.

Bottom of boring at 29.0…