Loading...
HomeMy WebLinkAboutKing Cove Geotechnical Investigation 1984Alaska Energy Authority LIBURY COPY A GEOTECHNICAL INVESTIGATION OF THE PROPOSED KING COVE HYDROELECTRIC WEIR SITE ON DELTA CREEK, ALASKA Alaska Department of Nat~ral Resources Division of Geological & Geophysical Surveys (Ori i~al Report) DEPARTMENT OF NATURAL RESOURCES DIVISION OF GEOLOGICAL & GEOPHYSICAL SURVEYS 17 October 1984 Brent Petrie Alaska Power Authority 334 W. 5th Avenue Anchorage, AK 99501 Dear Mr. Petrie: t(rc_ 007 Or, 'A ,, ::::.j BILL SHEFFIELD, GOVERNOR 0 POUCH 7-028 ANCHORAGE, ALASKA 99510 PHONE: (907) 276-2653 794 UNIVERSITY AVENUE, BASEMENT FAIRBANKS, ALASKA 99701 PHONE: (907) 474-7147 * P.O. Box 772116 Eagle River, Alaska 99577 Phone: ( 907) 688-3555 Attached herewith please find the final report entitled "A Geotechnical Investi- gation of the Proposed King Cove Hydroelectric Weir Site on Delta Creek, Alaska" by J.W. Reeder, K.J. Krause, and R.D. Allely. This report summarizes data and conclusions of an investigation executed by the Engineering Geology Section, Alaska Division of Geological and Geophysical Surveys, under a Reimbursable Services Agreement with Alaska Power Authority (08-73-4-490-389). I trust that you will find that the report provides the data specified in our original agreement including: 1. Seismic refraction survey, three lines. 2. Subsurface exploration. 3. iO-component peizometer array installation. 4. Boulder distribution assessment. 5. Stream discharge variance study. 6. Indentification of material resources. 7. Ground-water flow net modeling. In addition, the project team has been able to make some evaluations of down- stream transport of suspended and bed load sediments of the creek, is providing a description of the regional geology and potential hazards to the site, and offers suggestions of an alternative weir site. It was a pleasure executing this work for Alaska Power Authority and we hope that our final report will be helpful in accomplishing design and construction of the facility. If we can be of any further assistance on this or other APA projects, please feel free to contact me. Sincerely yours, ~-;; ~ ~..J.o--U k . ~r aLL_ ____ ·- ot. Randall G. Updike Chief, Engineering Geology Section RGU/jlw cc: D. Denig-Chakroff R. Loeffler A GEOTECHNICAL INVESTIGATION OF THE PROPOSED KING COVE HYDROELECTRIC WEIR SITE ON DELTA CREEK, ALASKA Final Report to the State of Alaska Power Authority Department of Commerce and Economic Development 334 West 5th Avenue Anchorage, Alaska 99501 under RSA 08-73-4-490-389 by J.W. Reeder, K.J. Krause, and R.D. Allely Alaska Division of Geological and Geophysical Surveys September 1984 This document has not received official DGGS review and publication status, and should not be quoted as such. State of Alaska Department of Natural Resources Division of Geological and Geophysical Surveys P.O. Box 772116 Eagle River, Alaska 99577 TABLE OF CONTENTS I. Introduction A. Purpose of Investigation B. Location of Study Area C. Physiography D. Acknowledgements E. Scope of Work and Methods of Investigation II. Geologic Overview III. Unconsolidated (Surficial) Deposits IV. Seismic Refraction Surveys V. Hydrology A. Surface Water and Sediment Transport B. Ground Water VI. Potential Rock Sources VII. Conclusions and Recommendations References Appendix Page # 1 1 1 3 4 4 8 17 22 27 27 39 57 58 61 63 Figure 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Plate 1. Plate 2. Plate 3. List of Figures Location map of study area Photo of Mount Dutton and study area Photo of King Cove and glacial valley Photo of Delta Creek valley near proposed weir site Photo of coulees near proposed weir site Photo of Clear Water Tributary and bank slump Map showing Quaternary landslides and photo lineaments Photo of Delta Creek alluvium exposed in CAT trench Particle size distribution curve for alluvium, till, and suspended sediment Photo of CAT trenching alluvium Photo of John Reeder sampling till Seismic refraction profile for 550-foot line Seismic refraction profile for 1100-foot line Seismic refraction profile for 55-foot line 1982 discharge hdyrograph for Russell Creek Photo of Delta Creek high water flow Photo of Delta Creek before storm Photo of Delta Creek after storm Graph showing affect of depth to mean velocity and discharge of bed material Photo of painted red stripe across Delta Creek Close up photo of painted red stripe Slug test for piezometer no. 10 Slug tests for piezometers 1,2,3,4,5,6,7, and 9 Curves for seepage volumes beneath weir structures Flow net for 7.5 foot deep sheet-pile weir Flow net for 15 foot wide concete-apron weir Graph showing maximum exit gradients for weirs Photo of alternate DGGS site on Clear Water Tributary Plates located in back pocket Aerial photo enlargement of Delta Creek Study area Page # 2 9 11 12 13 15 16 18 19 20 21 24 25 26 28 30 32 33 36 40 41 43 44 50 52 54 56 59 Map showing seismic refraction lines and piezometer locations Map showing ground-water gradient in relation to piezometer locations INTRODUCTION Purpose of Investigation In August of 1984, Alaska Power Authority (APA) approached the Engineering Geology Section of the Division of Geological and Geophysical Surveys (DGGS -Alaska Department of Natural Resources) concerning collection of additional geotechnical data that were needed for the proposed King Cove hydroelectric facility on Delta Creek. APA was ready to finalize the design criteria for the King Cove hydroelectric facility; however, several geotechnical issues needed to be addressed before they felt this could be properly done. The two primary geotechnical issues of APA concern were: depth-to-bedrock beneath the proposed diversion weir site, and the expected ground-water seepage flow through the unconsolidated deposits under the proposed weir. Because the above geotechnical issues were within the capabilities of DGGS, the Engineering Geology Section (EGS) was contracted to do the geotechnical work. In addition, EGS proposed to study and classify the unconsolidated deposits, measure stream discharge at several sites along Delta Creek, and identify bedrock quarry sources that could be used for construction. In the process of doing the geotechnical investigations, EGS identified two other concerns that are also addressed in this report. These additional concerns are (1) local landslide and potential fault hazards, and (2) sediment problems due to suspended sediment and bedload transport. Location of the Study Area The study area is located approximately six miles north of King Cove (fig. 1). King Cove is a fishing community located on the Pacific Ocean side of the Alaska Peninsula approximately 19 miles southeast of Cold Bay (abandoned Fort Randall), Alaska. 182° 30' - Deer 8 30' Scale 1·=4 miles Figure 1. Location map of the Delta Creek study area. -2- Physiography The proposed hydroelectric site is located in the glaciated and eroded Alaska-Aleutian Range physiographic province at the southwestern tip of the Alaska Peninsula. The proposed site is located on Delta Creek, which drains the southwest flank of Mount Dutton, a 4,884 foot high, glaciated Quaternary volcano (fig. 1). The Mount Emmons-Pavlof volcano region is located 10 miles northeast of Mount Dutton. Frosty Peak is located 20 miles west of Mount Dutton. Mount Dutton, Mount Emmons, Pavlof volcano, and Frosty Peak are all Quaternary stratovolcanoes. Both Pavlof and Pavlof Sister of the Mount Emmons-Pavlof volcano region have been active within historic times (Coats, 1950). Over 40 inches of rainfall occur in the Delta Creek area per year. At least fifty inches of average snowfall usually contribute to this precipitation (Selkregg, 1974-77). Storms are a frequent occurrence in the Alaska-Aleutian Range province. They can occur any time of the year and are usually due to the passage of east-moving Aleutian lows. The climate is maritime with an approximate mean annual temperature of 38°F. Mean annual maximum and minimum temperatures vary less than 10°F (Waldron, 1981). The dominant winds, which also bring the wettest weather, are from the south-southeast. Winds from the Bering Sea usually bring colder and dryer weather, which is more frequent during February and March. August is usually the foggiest month; although it is not unusual for fog to occur throughout the year. In general, the Aleutians are notorious for unpredictable weather, and 11 bad 11 weather can occur anytime of the year. -3- The rich marine life of the Alaska Peninsula and Aleutian Islands has supported mankind for at least 8,000 years (Black, 1976). King Cove was founded in 1911 as a Pacific American Fisheries Cannery (Biery, 1966). Fishing is still the main economic backbone for the community. On land, the sea bird and other animal life is plentiful. Large brown bears are the most ominous of the animal life, and are quite numerous in the province. Trees are sparse in the province; however, tundra, alder and willow grow on lower slopes and usually blanket lowland areas. Grasses, mosses, flowering plants, and berry bushes are the dominant tundra vegetation. Acknowledgements This project was made possible through a Reimbursable Services Agreement between APA and DGGS. We would like to thank Brent Petrie, David Denig-Chakroff, Robert Loeffler, and Remy Williams of APA for their input into the investigation and the assistance given in logistical planning. We would also like to thank the City of King Cove and the King Cove Native Corporation for their interest and logistical support of the field work. Special thanks goes to Chester Wilson, our CAT operator, who transported our equipment between the weir site and King Cove using the city's CAT. Chester's operating ability also enabled us to install the piezometers which was considered a doubtful task in our proposal. Special thanks also goes to Jenny Weir, our typist-editor, for assisting with the report preparation. Scope of work and methods of investigation Geotechnical field investigations were conducted between August 20-26, 1984. Travel time required to go to King Cove and return to Anchorage was -4- three days. A storm during the week of field work caused Delta Creek to rise and overflow its banks for approximately 24 hours, precluding any field work during that period. During the three remaining days, the proposed objectives described below were accomplished. APA was concerned about two main geotechnical issues involving the King Cove hydroelectric project. As mentioned earlier, these issues were depth-to-bedrock beneath the proposed weir site, and groundwater seepage flow within the unconsolidated sediments beneath the proposed weir. APA also expressed concern about the erosion and sedimentation stability of the Delta Creek channel. The Engineering Geology Section proposed the following scope of work. The depth-to-bedrock information would be acquired by employing standard seismic refraction survey methods. A 12-channel signal-enhancement seismograph and 55-foot, 550-foot, and 1,100-foot geophone lines were used. Explosives designed for use in seismic exploration were used as energy sources for the two longer lines. The depth-to-bedrock and thickness of the glacial till and alluvium overlying the bedrock were determined. It was also proposed that an excavation pit be dug and accurately logged and sampled. It was not possible to get a backhoe to the site, so the proposed pit investigation was done with the CAT. The CAT was able to trench into the alluvium to the top of the glacial till. A 150-pound representative alluvium sample and a 150-pound representative glacial till sample were collected from the trenches. The samples were mechanically analyzed for -5- particle size distributions by the Department of Transportation and Public Facilities (DOTPF) lab. An array of 10 well-point piezometers at pre-selected locations were installed below the water table. The CAT trenched to the water table, then the piezometer screens were driven below the water table with the aid of a steel fabricated driver-hammer tool. Piezometer elevations and locations were surveyed after all were installed. The gradient for the water table was determined from the water levels in the piezometer. Slug tests were successfully used for determining hydraulic conductivity values on 9 of the piezometers. Ground-water flow nets were constructed for the unconsolidated deposits underneath the proposed weir structures in order to determine seepage losses. Boulder distributions were photographically documented as trenches were dug for piezometer installations. Stream discharge measurements were also proposed and were successfully completed at the powerhouse site and proposed weir site. Surface-water loss was determined from the discharge measurements for the portion of Delta Creek between the weir site and powerhouse site. In addition to the discharge measurements, a suspended sediment sample was collected from the stream at the weir site. The ratio of suspended sediment to discharge was then calculated. Above the weir site, a 1-foot-wide channel cross-section strip of alluvium was painted bright red. Survey stakes with flagging were also installed along the painted strip. The painted strip of alluvium was photographed so that approximate erosion andjor deposition of bed-load can be measured in the future. -6- Investigation for competent bedrock that could be used for backfill ballast revealed that the best source would be from a rock quarry between King Cove and the airport. -7- GEOLOGIC OVERVIEW The region around Mount Dutton consists dominantly of well-bedded tuffaceous sandstones and conglomerates. These were originally mapped as the Belkofski Tuff which is prevalent in the Belkofski Bay area (Kennedy and Waldron, 1955). The Belkofski Tuff is thought to be late Eocene or early Miocene in age (Waldron, 1961). On the southern flank of Mount Dutton at the 1,500 foot elevation (fig. 2), the Belkofski Tuff strikes N toNE and dips about 12 degrees south. Approximately 2-3 miles south of the weir site, the Belkofski Tuff strikes N to NE and dips about 14 degrees south. Belkofski Tuff beds exposed along Delta Creek also dip to the south. North of King Cove, the Belkofski Tuff is intruded by a quartz diorite porphry. The contact between the quartz diorite and Belkofski Tuff has never been mapped, so the size of the intrusive is not known. Kennedy and Waldron (1955) recognized this intrusive and another intrusive on the east flank of Mount Dutton. The Belkofski Tuff appears to be highly altered in the Delta Creek canyon above the weir site. The alteration could be a result of nearby intrusive activity. No plutonic (intrusive) clast were seen in the alluvium along Delta Creek, suggesting that the intrusive is not yet exposed on the southern flank of Mount Dutton. Waldron (1961) assigned a middle-upper Tertiary age to these intrusives. West of the weir site, the Belkofski Tuff has been unconformably capped by massive andesitic lava flows. These flows form the ridge shown on the right in figure 2. The lava flows have not been dated but they are probably late Pliocene or Quaternary in age. The lava flows may have come from Mount Dutton or the Emmons Lake-Pavlof volcano region to the northeast. -8- • • ;:::::. Commun Figure 2. Looking north from the airport towards Mount Dutton and the Delta Creek study area. Note the dark basaltic flows on the upper flanks of Mount Dutton that dip away from the summit area of this Quaternary volcano. -9- The upper flanks of Mount Dutton consist of thinly layered basaltic lava flows. These flows dip away from the summit and are exposed on the north and northeastern flanks in the summit region. In figure 2, the volcanic flow units can be seen dipping as steeply as 33 degrees on the upper eastern flank of the mountain. The Mount Dutton region has never been mapped by the U.S. Geological Survey (Kennedy and Waldron, 1955; and Waldron, 1961). Coats (1950) claims that Mount Dutton is not a volcano. Based on our observations, it appears that the basaltic lava flows originated from Quaternary Mount Dutton eruptions. Most of the unconsolidated deposits in the region consist of glacial drift. The valley between King Cove and the airport is a classic U-shaped glacially-carved valley (fig. 3). King Cove itself is situated on a gravel spit that formed across a fiord. Dense, well-compacted glacial till underlies the alluvium in Delta Creek and forms the benches bordering Delta Creek near the proposed weir site. The bench on the south side of Delta Creek contains scabland-type topography which resulted from meandering glacial outwash streams. The glacial till and outwash are deposits remaining from middle to late Pleistocene glaciations occurring on Mount Dutton. Numerous abandoned stream channels (coulees) are present in the vicinity of the proposed weir site (plate 1). Delta Creek previously occupied these coulees (figs. 4 and 5) prior to entrenching itself to its present position. Delta Creek, in the past, flowed where the Clear Water Tributary now exists. Soil solifluction is an active process in this region. Solifluction is more pronounced on vegetation-free slopes, and is visible in the photographs as patterned ground. Talus fans occupy lower slopes west of the weir site. -10- • • • Figure 3. Looking south-southwest towards King Cove which is located on a gravel spit. Note the U-shaped, glacially-carved valley in the foreground • -11- • • • Figure 4. site • Looking north-northwest up Delta Creek from near the proposed weir Note Quaternary landslide and coulees. -12- • • Figure 5. Close-up view of the coulees located above proposed weir site . • -13- A small slump has occurred after 1980 on the east bank of the Clear Water Tributary just above its confluence with Delta Creek (fig. 6 and plate 1). Two large landslides west of Delta Creek are shown on plate 1 and figure 7. The landslide below the weir site and on the slope above the powerhouse site appears to be stable. This landslide did not move below the 450 foot contour (fig. 7 and plate 1). The second landslide is located upstream above the proposed weir site (fig. 4). Delta Creek is cutting into the toe of this slide, thus it may not be stable. If this landslide failed again it could temporarily dam the Delta Creek channel. No other landslides or small bank-slumps have been recognized in the study area. Two prominent air photo lineaments were recognized crossing Delta Creek (figure 7 and plate 1). One is in the vicinity of the proposed powerhouse, the other is near the proposed weir site. These lineaments trend N 41°W and N 45°W respectively. The direction of maximum horizontal compression in this region is N 40°W. This compression is from the Pacific Plate underthrusting the Aleutian Arc (Nakamura et al, 1977). Southwest of this region, on Unalaska Island, numerous small Holocene faults also strike in the direction of maximum horizontal compression (Reeder, 1984). The Delta Creek lineaments are traceable across both bedrock and Quaternary glacial deposits. Therefore, these lineaments are thought to be Holocene faults. The lineaments trend across the two large landslides and may have been responsible for triggering these failures. Since these lineaments may represent small Holocene faults, they probably would not generate large earthquakes. -14- • • Figure 6. View of recent bank slump east of Clear Water Tributary. This view is just upstream from the Delta Creek confluence. Piezometer no. 1 is shown in foreground. -15- -16- ~ Quaternary landslide \ Air Photo Lineament/Possible \' Quaternary Fault \. .~.. Scale 1·=1 mile Figure 7. Map showing locations of Quaternary landslides and photo- lineaments (possible faults) in relation to study areas as outlined. UNCONSOLIDATED (SURFICIAL) DEPOSITS Stream alluvium overlies glacial till along Delta Creek in the vicinity of the proposed weir site. The glacial benches bordering Delta Creek near the weir site consist of till with a thin mantle of overlying colluvium. Figure 8 shows a CAT trench wall in the alluvium near piezometer no. 8 (plate 2). The alluvium was approximately 6-8 feet thick near piezometer no. 8. As shown in figure 8, the alluvium is poorly sorted and poorly stratified, and consists of boulders, gravel, sand, and silt. A 150-pound alluvium sample was collected above the water table in a CAT trench near piezometer no. 6 (plate 2). Material analyses were performed on the sample and the cumulative probability plot for the particle size distributions is shown in figure 9. The material analyses revealed that the alluvium consisted mainly of boulders and gravel with about 17 percent of the sample consisting of sand and 3 percent consisting of silt. Boulders weighing more than 2,000 pounds are scattered throughout the alluvium (fig. 10). A 150-pound glacial till sample was collected near piezometer no. 8 (plate 2, fig. 11). Till exposed above the water table in the CAT trench was dense and well-compacted and was difficult to excavate with the CAT. A blow from the pointed end of a rock hammer would only penetrate the till about 1 inch. When the till became water saturated, it would flow easily. The material analyses for the till is also shown in figure 9. The cumulative probability plot for the different particle size distributions indicates that the till consists of 53 percent sand and 9 percent silt. Glacially rafted boulders weighing upwards of 20,000 pounds tend to be most prevalent along the glacial bench east of Delta Creek (fig. 10). -17- • • • Figure 8. Three-foot vertical section of Delta Creek alluvium exposed in CAT trench near piezometer no. 8 (plate 2}. Alluvium is poorly sorted and consists of boulders, gravel, sand, and some silt . -18- Boulders/ ENGINEERING GEOLOGY LABORATORY Particle Size Distribution Curve A.B.T.M. Claaalflcatlon Cobbles Gravel I ~and I Silt Clay 100 , \ 90 80 -70 s; Q • \ \ I ----------~---~ ---. ---~-·-----.. ---- \ \ ..... \-,__ ------------,____ ~~~-SE dim4 I· ----- \ ~~ i\ 1--ill 'a -+-f---·------· -1---- \ '-. .. ...... ( laci ~ 60 --'---~ ~+---------· -------·----- "" .D .. 50 • c ~ ~ -~ ----1--------f- II. -40 c • u .. • 30 A. 20 \ vS lreat ILJ~ ri.m_ I\ -------~------ r--~- r---... ~ ~ ----------· . l-------\. 1--- !\ \ --.::_~ -----. \. ---~---- h "' " 10 I·-~-~----f.--. 1--!------t-..... t'~ ' ~ 1"-r-, Pooo 100 10 0.1 0.0 1 Particle Size Diameter In Mllllmeten Figure 9. Particle size distribution curves for stream alluvium, glacial till, and suspended sediments. Stream alluvium and glacial till samples were collected above the water table. Stream alluvium was sampled near piezometer no. 6 and till was sampled near piezometer no. 8 {plate 2). Suspended sediment sample was collected at Upper Delta Creek discharge site {plate 1). -19- --~- --. -- 0.00 1 • • • Figure 10. Picture of King Cove's CAT at work trenching alluvium for a piezometer installation. Note large boulders in the alluvium and the large glacially-rafted boulders on the glacial bench in the background . -20- • • • Figure 11. John Reeder collecting a 150-pound till sample in the CAT trench near piezometer no. 8 (plate 2) • -21- SEISMIC REFRACTION SURVEYS Three seismic refraction surveys were conducted in the vicinity of the proposed weir site to determine depth-to-bedrock and thicknesses of overlying glacial till and stream alluvium. A Geometries 12-channel signal enhancement seismograph was used for the surveys. An average of 5 pounds of high velocity Kinepak explosives per shot and instantaneous blasting caps were used for energy sources. As shown in Plate 2, locations for the profiles are: a 550-foot geophone line (50-foot spacing) trending N-S along Delta Creek and Clear Water Tributary {1090 ft long, Au -AD profile, fig. 12); an 1100-foot geophone line (100-ft spacing) trending N-S on the glacial bench east of Delta Creek and the Clear Water Tributary (1280 ft long, Bu -BD profile, fig. 13); and a 55-foot geophone line (5-ft spacing) trending N-S across the proposed weir site {65 ft long, Cu -c0 profile, fig. 14). The seismic refraction method consists of measuring the initial arrival times of compressional seismic waves at points along a line by means of geophones. The elastic compression waves refract at various geologic boundaries according to Snell's Law. Based on this law, the velocities of and depths to interfaces between layers of differing density can be calculated from the initial arrival time with respect to distance from the energy source (shot point). The specific analysis technique used in this study is ca 11 ed the "time intercept method" as described by Dobrin (1960). The seismic profiles shown in figures 12, 13, and 14 represent corresponding arrival times versus distance from the shot point. The velocities found for glacial till vary between 5,820 and 7,520 ft/sec. Bedrock velocities were found to vary between 10,000 and 13,000 -22- ft/sec. This is the approximate velocity range for a metasedimentary rock (Dobrin, 1960), which is the assumed bedrock type beneath the site. The depth-to-bedrock along the 1090-foot profile (fig. 12) is approximately 150 feet on the south end and 190 feet on the north end. The depth-to-bedrock along the 1280-foot profile (fig. 13) is approximately 320 feet on the south end and 225 feet on the north end. Beneath the weir site, the bedrock is approximately 170 feet deep. Along the 1280-foot seismic profile, Bu -B0 , low surface material velocities of approximately 1800 ftjsec are inferred for a shallow layer from 15 to 40 ft deep. This may represent a loosely compacted glacial outwash which overlies the till. Glacial outwash features are evident on the photos for this area (plate 1). The alluvium beneath the weir site has an average velocity of 4,200 ft/sec (fig. 14). Based on seismic profile Cu -c0 the alluvium varies between 10 and 12 feet in thickness. Near piezometer no. 8 the alluvium was found to be between 6 and 8 feet thick in a CAT trench which bottomed in glacial till. -23- (/) 0 80 -z -......... 0 -......... -(,) -...... -....... UJ -......... ~ 60 -........ -........ ...J -...... ...J :E 0 UJ Slope•~ Slope•~ 2 I f A I i & I l ; I • 0 I lh i'l I I~ GEOPHONES Till ~ = 613.9 f/s ---------------------- Bedrock V2• 12,612 f/1 NO 400 800 800 L I N E D I S TAN C E in F E E T Figure 12. Seismic refraction profile Au - A0 for 550-foot line trending N-S along Delta Creek and the Clear Water Tributary (plate 2). Upper part of figure represents first arrival seismic travel times and lower part of figure represents velocity model. A geophone spacing of 50 feet was used. -24- SHOT PT. 1000 s Slope• 11,776 ' ' 2 ' .c80 ' / ' LLI / :1 / I / Slope. 7S32 / • 0 * __ L.I -.LI __ .J.I _..LI-...LI--'-.......JIL---L.I -.L.' -.L' --':-1-~1-* IH()Tptl 2 3 4 IS II 7 •• 8 10 II 12Stm.I"T. GEOPHONEI <> ~~----------------------1-------------· Zc:tLIJ 0LLILI.I400 ~CJ)LL. Till V. = 6883 c( Ll.l .c 300 >>- LI.IO..J - - - - - - - -..JCD~ 200 - - - - - - LI.I 4 LLI Bedrock v2 • IQ,582 - - - - ..J IOOOL---~+---~400~--~~~----~~~--~1000~--~~~~~ LINE DISTANCE in FEET Figure 13. Seismic refraction profile Bu -B0 for 1,100-foot line trending N-S on glacial bench above Delta Creek and the Clear Water Tributary (see plate 2). Upper part of figure represents first arrival seismic travel times and lower part of figure represents velocity model. A geophone spacing of 100 feet was used. -25- t:. _L Slope=-6381 / / ::! / 6 / c: / ·-/ 0 w / 0 ~4 /, I 1-0 Slope•~ Slope·3~t:. 2 0* I I I I I I I I I I * SHOT PT. I 2 3 4 5 I 7 8 t 10 II 12 SHOT PT. GEOPHONES -------i Alluvium V1 = 4230 f/s --.... __ .......... _ ----Till V2 = 6683 f/s N 0 s Figure 14. Seismic refraction survey profile Cu-CD for 55-foot line trending N-S along Delta Creek at proposed weir site (see plate 2). Upper part of figure represents first arrival seismic travel times and lower part of figure represents velocity model. A 5-foot geophone spacing was used. -26- HYDROLOGY Surface Water and Sediment Transport The Delta Creek watershed is situated on the southern flanks of Mount Dutton. These flanks face the Pacific Ocean and receive substantial precipitation from passing Pacific storms, which are most common between summer and mid-winter. Dryer storms from the Bering Sea are most common from mid-winter to spring. Increased discharge will occur in Delta Creek during any storm that contains significant precipitation. Precipitation falling during the winter is usually stored in the snow pack at higher elevations until it melts in spring and summer. Because of the maritime environment for this region and its mean annual temperature of 38°F, heavy rainfall and runoff can occur even in winter. Figure 15 is a 1982 hydrograph for Russell Creek near Cold Bay (fig. 1). The hydrograph shows numerous high water flow periods which occur during and after storms. Russell Creek drains the eastern and northeastern slopes of Frosty Peak, a Quaternary volcano like Mount Dutton. Low water flows occurred for Russell Creek between February and the middle of May. Short periods of high water flows occurred throughout the rest of the year. High surface flows occurred during September and October. May through August also marked a period of continuous high surface flows; however, this is probably partly due to snow pack and glacial melt. Because Russell Creek and Delta Creek are in similar hydrologic environments, it is felt that their hydrographic discharges would be roughly similar. While the field work was underway for this investigation, a southeasterly storm occurred on the evening of August 23. The storm caused -27- 2000 1800 1100 1400· i ,. I • I , I .... UOOr· • II. 0 • 1000· • • .. • aoo; ,. c 100• j ! I 4001 j , 1181 • &!"a 20 :·! 10 • i== 301 •• • • · o+--O~c-t-.-r~N~o-v-.-r~o-.-c-."•J~.-n-.-,--F~e~,~.~tr-"'u~.-,.-.r-A~~-,-.~,r-M~a-,--,,~Ju_n_•-.r-~Ju~l-y-.--A~uo-.-.~a~.-p-t.,, R•eeel Creel! Near Cel~ •••, Ala•ll• Lat. 11 1 o• ao• LHt. 111 41' oa• Figure 15. 1982 discharge hydrograph for Russell Creek near Cold Bay (fig. 1). Suspended sediment data for four samples is shown in lower graph. -28- Delta Creek to rise and overflow its banks. The high water was impossible to cross on August 24, and therefore the weir site was inaccessible. The water in the channel rose almost 1 foot and flooded many of our previously installed piezometers (fig. 16). The water in Delta Creek, which was previously milky-white in color, changed to a muddy-brown color. Yet, the water in the Clear Water Tributary rose only 2 inches during the storm and remained crystal clear. Prior to the storm, discharge measurements were taken on August 23 at the powerhouse site and just upstream from the proposed weir site (plate 1, App. A-1 and A-2). After the waters in Clear Water Tributary had subsided from the storm, a discharge measurement was taken (App. A-3). A discharge measurement of 54.3 cfs± 8 percent was measured at the powerhouse site, and a discharge of 57.8 cfs± 10 percent was measured at the proposed weir site. Of the 57.8 cfs± 10 percent, 42.0 cfs was measured in Delta Creek above the Clear Water Tributary and 15.8 cfs was measured in the Clear Water Tributary. The discharges at the weir site and powerhouse site are similar. Some surface water loss due to ground water storage probably was occurring between the weir and the powerhouse sites during our visit, since several small tributaries were flowing into Delta Creek between the weir and powerhouse sites and the measured discharge upstream was found to be larger. It would be expected that some of the ground water would be released during periods of very low flow, such as in winter. The Clear Water Tributary has a drainage basin area of approximately 1.4 square miles. The Delta Creek drainage basin area above the Clear Water -29- • • -· Figure 16. Delta Creek high water flow the day after the August 23, 1984 storm. Creek overflowed its banks in many places. Weir crosses where gravel island occurs. -30- Tributary confluence is approximately 2.0 square miles. Collectively, this represents a total drainage area of 3.4 square miles. The Clear Water Tributary contains 41 percent of the total drainage area but only 27 percent of the flow. No glaciers exist in the Clear Water Tributary drainage basin. A glacier with an approximate 0.39 square mile area exists in the Delta Creek drainage basin. The higher discharge rate per drainage basin area for Delta Creek is probably due to glacial melt water. The gap between the discharge measurements for Delta Creek and the Clear Water Tributary would probably be narrower during winter when glacier and snow pack melting at higher elevations is negligible. After the high water from the storm had subsided, extensive storm related sand and silt deposits were observed along Delta Creek. Figure 17 is looking south before the storm and figure 18 is looking northeast at approximately the same location after the storm. Prior to the storm and while taking the stream discharge measurement at the powerhouse site, the stream discharge wading rod tended to sink continuously in loose sand, making the discharge measurement somewhat difficult to obtain. During the storm, large boulders, cobbles, and pebbles could be heard moving along the Delta Creek stream bottom. After the storm waters partially subsided, a 500 milliliter water sample was collected on August 25 at the upstream Delta Creek discharge site. The sample was collected one inch from the water surface at the approximate position of maximum surface water velocity (i.e., sta. 17.5 ft, App. A-2). This sample contained 0.7 gram of suspended sediment, of which 52 percent was sand (fig. 9). -31- • • Figure 17. Looking south from weir site just before August 23, 1984 storm • • -32- • • • Figure 18. Looking northwest from weir site two days after storm (August 25, 1984). Note sediment deposition differences between this and the previous photo . -33- Two different diversion weir structures have been proposed for this project. One weir would have a 4.5 foot water level (Dowl Engineers et al, 1982), and the other weir would have a 8.0 foot water level (APA design). The 4.5 foot diversion weir would create a reservoir volume of 9,100 cubic feet, and the 8.0 foot diversion weir would create a reservoir volume of 40,600 cubic feet. In order to make an estimate on how long it would take for the two reservoirs described above to fill with sediment, the following assumptions were made: (1) average suspended load for the stream is 0.7 grams per 0.5 liter; (2) 52 percent of the suspended load is sand; (3) the average discharge for Delta Creek above its confluence with the Clear Water Tributary is 42.0 cfs; (4) the Clear Water Tributary contains no suspended sediments; (5) all suspended sand will settle and be deposited behind any diversion weir structures while all other suspended sediments will pass over or through the structure; and (6) the dry bulk density for settled sand is 122 pounds/ft 3 (Hough, 1969). Based on the reservoir volumes and the assumptions described above, the 4.5 foot diversion weir reservoir would be filled with sand in 6.6 days, and the 8.0 foot diversion weir reservoir would be filled with sand in 29.2 days. In actuality, the upper foot or so would probably not fill with sand because of turbulent water flow, and the sand would most likely be transported over the top of the weir structure or through the penstock. Independent of the validity of the above assumptions, it should be emphasized that these calculations are based on only one suspended sediment sample which may or may not be the statistical norm for suspended sediment load in Delta Creek. The suspended sand will need to be addressed in the final weir design. -34- Most of the suspended sand in Delta Creek is probably being eroded from the Quaternary glacial tills which blanket the drainage basin. The particle size distribution plot for the glacial till that was collected near piezometer no. 8 showed 52 percent sand (fig. 9). During the Holocene, Delta Creek eroded and transported a large volume of sediment. Between the powerhouse site and the proposed weir site, Delta Creek flows through a canyon that is approximately 50 feet deep and 100-300 feet wide. The canyon was cut into glacial tills. These eroded glacial deposits, and other eroded glacial deposits from upstream, have been deposited on the large alluvial fan located between the airport and the powerhouse site (fig. 7). The airport gravel quarry is located on this alluvial fan. Einstein (1950), after extensive field work in the western United States and after extensive laboratory experimentation, developed an empirical derivation for total sediment transport in streams. Colby (1961) revised Einstein's procedure and empirically derived discharge of sediments at 60°F with respect to water depth and mean water velocity. In figure 19, Colby shows such a relationship for a streambed of well-sorted (0.3 mm median diameter) sand. Bed material discharge in figure 19 means bedload sediments plus suspended sediments being transported by flowing water. By using figure 19, an empirical calculation of total bed material discharge can be made for the upper Delta Creek discharge measurement site by determining the sediment transported for each section of velocity measurement (App. A-2) and then adding them together. The calculation comes to 244 tons/day. If an average depth and velocity of 1.0 feet and 3.33 feet/second respectively is assumed, then the Delta Creek channel at the upper discharge measurement site will need to be 12.73 feet wide. Based on figure 19, the discharge of sand for -35- J: I-1000 Cl :: u.. 0 I- 0 0 u.. 0:: LLJ a. 100 >- <l: Cl 0:: LLJ a. (f) 3 z 0 I-10 z --..J <l: 0:: LLJ I- <l: ::!' Cl 1.0 LLJ CD u.. 0 LLJ C) a: <l: J: u (/) -0.1 Cl 0.1 1.0 10 100 DEPTH, IN FEET Figure 19. Shows the affect of depth on the relationship between mean velocity and empirically determined discharge of total bed material (i.e., suspended and bed load sediment) for a well sorted 0.3 mm median diameter sand channel at 60°F (Colby, 1961). -36- this simplified channel would be 19.35 tons/day per foot width of channel, or 246.33 tons/day. Sand deposition of 84.89 tons/day was calculated using the sediment data derived from the water sample collected; namely, 0.7 grams per 0.5 liter of which 52 percent is sand. The 84.87 tons/day of sand calculation is far below the 244-246 tons/day empirical calculation. In actuality, the Delta Creek channel alluvium consists mostly of boulders and gravel with only 20 percent sand or finer sediments (fig. 9). Therefore, the actual sand transport in Delta Creek would be expected to be less than the empirically estimated sand transport. The calculations derived from the water-sample suspended load may not be unreasonable since the values were much lower than that estimated by the Colby-Einstein technique. The U.S. Geological Survey, in monitoring the Chulitna River near Talkeetna, Alaska, has recorded suspended load values as high as 0.715 grams per 0.5 liter for flowing water having a mean velocity of 8.2 feet/second and a 0.0014 (7.4 ft/mi) channel slope gradient (Knott and Lipscomb, 1983). Delta Creek at the weir site had a channel gradient of 0.055 (290 ft/mi) and an average flow velocity of 3.3 feet/second. At such a high gradient, large suspended sediment loads would be expected. For Russell Creek near Cold Bay, the suspended sediment transport was 0.39 tons/day in January, 7.6 tons/day in June, and 30 tons/day in September (Lamke et al, 1983). The gradient at the discharge site on Russell Creek is only 0.002 (10.5 ft/mi). One percent of the Russell Creek drainage basin contains glacial ice. Based on this information it would seem that the suspended sediment transport for Russell Creek would be much lower than that for Delta Creek. This assumption may be realistic since the total calculated -37- suspended-sediment transport for Delta Creek on August 25, 1984, was 158.6 tons/day. During the August 23 storm and resulting high water, we heard pebble, cobble, and boulder-size bedload material moving along the Delta Creek stream bottom, which suggests significant bedload transport. The large alluvial fan near the airport and the high stream gradient (0.005 slope or 26.4 ft/mi) also suggest significant bedload sedimentation. The general lack of brushy vegetation along the Delta Creek canyon bottom indicates that Delta Creek overflows its channel frequently during storms. Based on the above, bedload sediment transport should be investigated for Delta Creek. In contrast, the Clear Water Tributary canyon bottom contains a continuous moss cover which indicates the stream does not overflow its banks very often (fig. 6). Both the suspended sediment and bedload transport are considered to be fairly low for this tributary drainage. For example, even during the peak flow period from the August 23 storm, the water in the tributary was crystal clear. It is not unusual for stream bedload transport in the western United States to be an order of magnitude less than suspended load transport (Colby, 1961). This relationship appears to hold for streams examined in the Susitna Valley of Alaska also (Knott and Lipscomb, 1983). The calculated total suspended-sediment transport for Delta Creek is 158.6 tons/day. This again is based on the surface water sediment sample we collected. Average bedload -38- sediment transport for Delta Creek would therefore be about 10 percent of the 158.6 tons/day or 16 tons/day. We suspect this estimate is probably lower than what is actually being transported. To assist in future investigations of the bedload transport for Delta Creek, a one-foot wide strip of stream alluvium was painted bright red across the entire canyon floor bottom (plate 1, fig. 20 and 21). If this paint sticks to the alluvium, then any changes due to the channel scour or deposition can be noted. Scour chains have been used in the western United States for the purpose of determining the depth of channel scour (Leopold et al, 1964). Scour chains were not installed in the coarse alluvium of Delta Creek because of difficult excavation conditions. Suspended load sediment transport for most glacial streams in Alaska is 1-4 orders of magnitude higher during spring, summer, and fall than during the winter (Lamke et al, 1983). Stream bedload transport is usually at a minimum during the winter. Most winter precipitation is stored as snow or ice and glacial melt is low. In the winter, if channel bank ice is present, it helps protect the banks from being eroded. Recent studies by Knott and Lipscomb {pers. commun., 1984) for the Susitna River drainage indicate the bedload transport in the winter is much smaller than the suspended sediment transport. Because this region has a mild maritime climate large sediment transport could occur in glacial streams anytime during the winter. Ground Water Ground-water seepage beneath the proposed Delta Creek weir was one of APA's primary concerns. To investigate this problem, an array of 10 -39- • • • Figure 20. Shows the 1 foot wide red stripe painted across the Delta Creek canyon bottom above the proposed weir site on August 24, 1984 (see plate 1}. Future channel scour or deposition can be monitored using this stripe . -40- I ~ ..,_. I • • - Figure 21. Looking up Delta Creek at painted stripe and survey stakes, August 24, 1984. piezometers were installed (plate 3) in the vicinity of the proposed weir site. Hydraulic conductivity for the alluvium was determined by performing slug tests on each piezometer. These tests required measuring the rate of water infiltration for each piezometer after it was filled with water. The hydraulic conductivity values were used with the Darcy equation and appropriate flow nets to calculate ground-water seepage under the two proposed weir designs. The design and dimensions of the piezometers used are shown in the upper part of figure 22. Each piezometer consisted of a 47.8 em long sand screen (manufactured by Mark Controls Corp., trade number SBD 2460) threaded into a coupling that was attached to a 5 foot long steel pipe. The CAT was used to trench to the water table at the pre-selected piezometer sites (fig. 10). The piezometers were then driven into the alluvium and till using a 30 pound fabricated driver-hammer device. Each piezometer was driven into the sediments until the top of the sand screen was several inches below the water table. Fine sand and silt was then hand packed around the piezometer pipe before backfilling. Hand shoveling and the CAT were used to backfill around the installed piezometers. The trenches were backfilled as much as possible to their original state. The piezometer array was then surveyed for control (plate 2). Piezometer spacing was measured using a Brunton pocket transit and a 300-foot tape, and elevations between the top of the piezometer pipes were measured using a hand level and survey rod. Horizontal accuracy between piezometers is ±2 feet, and vertical accuracy is ±1 foot. The ground-water table was then measured at each piezometer. The ground-water table has been contoured from this data and is shown on plate 3. -42- h Ho 1.0 .9 .8 .7 . 6 1\ \ \ Delta Creek, King Cove, Alaska o Piezometer No. 10, To= 225 seconds L=478mm = 47.8cm=0.478m r= 175mm = 1.75cm = 0.0175 R= 22.5mm~ 2.25cm=0.022Sm '\ 1 ___ 1top of pipe, H20 at t=O ·\ . Figure 22. Slug test data for piezometer no. 10. Piezometer schematic is also shown. -43- Delta Creek, King Cove, Alaska 1.0 'l ~ c PiezOftllter No. I , To= 7. 5 seconds t:::. Piezometer No. 2 , To= 46.5MCOI"'dd [\: 0 Piezometer No. 3, To= 4.5 seconds i\. o Piezometer No.4, To= 12.5seconds ~ ~·"\ • Piezometer No.5, Tt~= 12.5 seconds .9 .8 .7 ' • Piezometer No.6, To= 17.0 stconds I~\ • Piezometer No.7, To= 69.0seconds '\. • Piezometer No.9, To= 30.0secondl ~ \\ '\ ~ L:478mm= 47.8cm=.478m r= 17.5mm = 1.75cm= .0175m R=22.!5mm= 2.25em=.022!5m ~ \ '\ 1'\ 1\ '\.. .. I \ \ \ '\ \ "" 1\ " \ 1\\ ~ ~ 1'\ ~\ ~\ ~ \' .d) 'I\. ~~ ~ ""' ~· ~ 1'\ \ t~ ~\ \\ "" ~ ~ \\ \ \ .6 .5 .4 .3 .2 \ \ • !\ \ \ .I 0 10 20 30 -10 !50 EK> 10 EK) 90 KlO 110 120 I~ 140 s.condl Figure 23. Slug test data for piezometers 1, 2, 3, 4, 5, 6, 7, and 9. -44- Slug tests were performed by quickly filling each piezometer using a gasoline engine pump, and then measuring the water level drop with time. A battery operated water level indicator (Watermarker manufactured by Johnson Division of U.O.P., Inc.) and a stop watch were used. The water levels dropped rapidly for each test performed and required fast data recording. Data for the slug tests show an exponential decline in water level drop with respect to time. The straight lines on the logarithmic graphs shown in figures 22 and 23 represent the normalized pressure-head difference between the water table and the piezometer water level with respect to time. Hydraulic conductivities for the sediments surrounding each piezometer were calculated using the slug test data. The calculations assume that the sediments are homogeneous, isotropic, incompressible, and infinite in volume, and that the ground-water flow is laminar and incompressible. Given the above criteria, Darcian ground-water flow will occur where the rate of water flow out of the piezometer screen is proportional to hydraulic conductivity, K, of the soil and to the pressure head of the piezometer with respect to the water table. Assuming the above situation, Hvorslev (1951; and described by Cedergren, 1967) has shown if L/R ::::> 8, that: K = r2ln(L/R) ; 2LTo where K = hydraulic conductivity; r = inside radius of steel piezometer pipe; L = length of piezometer screen; R outside radius of screen; To = basic time lag which is equivalent to the time value when h/Ho = 0.37, and h = water level height in the piezometer above the water table at any given time and H0 = water level height in the piezometer above the water table after the piezometer has been filled (i.e., at time= zero for the test). The To values for each slug test as shown on figures 22 and 23 are as follows: To 10 -45- = 225 seconds; To 9 = 30 seconds; To 7 = 69 seconds; To 6 = 17 seconds; To 5 = 12.5 seconds; To 4 = 12.5 seconds; To 3 = 4.5 seconds; To 2 = 46.5 seconds; and To 1 = 7.5 seconds. Based on Hvorslev•s calculation, Equation (1), the following hydraulic conductivities for the sediments surrounding each piezometer are: K10 = 1.43 X 10-5 ft/sec = 1.23 ft/day; Kq = 1. 07 X 10-4 ft/sec = 9.25 ft/day; K7 = 4.65 X 10-5 ft/sec = 4.02 ft/day; K6 = 1.89 X 10-4 ft/sec = 16.32 ft/day; K5 = 2.57 X 10-4 ft/sec = 22.20 ft/day; K4 = 2.57 X 10-4 ft/sec = 22.20 ft/day; K3 = 7. 14 X 10-4 ft/sec = 61.67 ft/day; K2 = 6.91 X 10-5 ft/sec = 5.97 ft/day; and K1 = 4.28 X 10-4 ft/sec = 37.0 ft/day. The lowest hydraulic conductivity value, 1.43 X 10 5 ft/sec for piezometer no. 10 (K 10 ) is in the range of silty sand, and the highest value of 7.14 X 10-4 ft/sec for piezometer no. 3 is within the range of sandy gravel. The hydraulic conductivity values for various sediments are summarized by Freeze and Cherry (1979). The large variation in hydraulic conductivity values is probably due to the fact that the alluvium is not a homogeneous and isotropic sediment as assumed. Piezometer no. 8 was installed in glacial till (fig. 11). The first slug test on this piezometer was unsuccessful because, when the pipe was filled with water, the till sediments surrounding the screen failed and a 11 blowout 11 occurred. A 11 blowout 11 is when the water flows up the outside of a -46- piezometer pipe to the surface (i.e., the seal around the piezometer has been broken). In a "blowout" the pumped water can flow out so fast that the piezometer cannot be filled. When this occurred with piezometer no. 8, we moved it to a new location and tried another slug test. This time a fast-setting concrete designed for underwater application was packed around the piezometer in hopes of sealing it. After pumping water into the pipe, a second ''blowout" occurred within 4 seconds. The slug test for the glacial till piezometer was then abandoned. Two other "blowouts" occurred for two of the piezometers installed in the alluvium. These piezometers are shown in figure 23 and they are identified with the notation ''seal breaking". The "blowouts" fortunately occurred long enough into the test so that reliable conductivity values were obtainable. The occurrence of "blowouts" during our slug tests points to the possibility of ground-water piping underneath the weir structure especially if the weir is placed on glacial till. Remembering that the glacial till consists primarily of sand and silt, and that it is well-compacted, suggests that it would have a very low hydraulic conductivity. Freeze and Cherry (1979) describe the simple Hazen empirical approach for determining hydraulic conductivity based on grain size diameter. This relationship is: K = d10 ; where K = hydraulic conductivity {cm/s); and d10 = grain size diameter (mm) at which 10 percent by weight is finer. Using the Hazen approach and a d10 of 0.5 mm for the alluvium near piezometer no. 4 (fig. 9), the K for the alluvium will be equal to 0.25 em/sec {8.2 X 10-3 ft/sec or 708 ft/day). Using a d10 of 0.083 mm for the glacial till (fig. 9), the K for the glacial till will be equal to 6.9 X 10-3 em/sec (2.26 X 10-4 ft/sec or 19.5 ft/day). The hydraulic conductivity for the material -47- surrounding piezometer no. 4, as determined from the slug test, was 2.57 X 10-4 ft/sec, which is 32 times lower than the Hazen empirical estimate. The Hazen approach was derived in a laboratory test using well-sorted sands. Because the alluvium and the glacial till are not very well sorted, the previous Hazen empirical estimates for hydraulic conductivity are considered to be poor. Because we were unable to obtain the hydraulic conductivity of the glacial till with piezometer no. 8, an empirical estimate is presented below. In the previous Hazen approach calculation, the K for glacial till is 36 times lower than the K for alluvium. It is assumed that this ratio for hydraulic conductivity holds. The average hydraulic conductivity for the alluvium, using the Hvorslev calculation, is 2.31 X 10-4 ft/sec or 19.98 ft/day. Assuming this value represents the K value for the material actually collected near piezometer 4, then the glacial till would have an approximate K value that is 36 times smaller than the average K value for the alluvium (6.42 X 10-10 ft/sec or 0.55 ft/day). Naturally, this is only a rough estimate for the hydraulic conductivity of the glacial till. In actuality one should expect even a smaller value due to the dense compaction of the ti 11. The approximate saturated ground-water flow through the alluvium can be calculated using the well known Darcy equation: Q AM!. K· where Q AL ' total discharge across a vertical cross-section; A = area of vertical cross-section; ~~ = unconfined water table gradient perpendicular to the vertical cross-section; and K = hydraulic conductivity of material at the cross-section. The alluvium beneath the proposed weir site is approximately 10 feet thick. The water table is about 2 feet deep, thus 8 feet of alluvium is saturated with ground water. Beneath the weir site the alluvium is -48- approximately 100 feet wide, and it has an average K value of .00023 ft/sec which was previously calculated from the slug test. Using this information and the Darcy equation, the total ground-water flow beneath the weir site is about 0.015 cfs or 1,300 ft 3/day. Based on the theory of saturated ground-water flow, Polubarinova-Kochina (1939) has determined the two-dimensional confined flow per unit normal to the direction of flow underneath various weir structures (fig. 24a). If a sheet pile weir was built so that the sheet piles penetrated the 10 feet of alluvium, and if the surface water head on the reservoir side was 8 feet, then seepage estimates under the weir can be made using figure 24a. One such estimate for seepage assumes that the sheet pile foundation rests entirely in glacial till that is 200 feet thick. The estimated hydraulic conductivity value of 6.42 X 10-6 ft/sec for the till is considered to be representative throughout its entire thickness. Applying this information to figure 24a, the S/T will be 0.05, and the b/t will be 0. Then q/Kh will equal 1.2375, where q is the quantity of ground-water flow per unit normal to the direction of flow, K is the hydraulic conductivity of the till which is 6.42 X 10-6 ft/sec, and h, the total head behind the weir, is 8 feet. This assumes that the ground surface adjacent to the downstream side of the weir represents the water table. If the weir is 116 feet long, and assuming no piping occurs beneath the weir and that only two dimensional flow occurs, then the total seepage would be 7.0 X 10-2 cfs or 637 ft3jday. If we took the highest hydraulic conductivity value for the alluvium, K = 7.14 X 10-4 ft/sec, and substituted this K value (we are assuming 200 feet of alluvium and no till), then the total seepage flow would be 0.88 cfs or 7.57 X 10 4 ft3Jday. The actual seepage flow is probably between these two extreme cases. -49- q kll 1.SI I 1.4 1.3 \ 1.2 \ 1.1 \ 1.0 0.9 1\ \ 0.8 f---+----"f---+----l--+----l--+---+-+---l ql.tll r-... ~d 0. 7 ......_ .:(,'o-+-+--+-+---+-+---1 "~" 0~~~~~~~~~~~~~~-J 0 0.1 0. 2 0.3 0.4 0.5 0.6 o. 7 0.8 0.9 1.0 s;r --Position A f ---Position 8 or 8' I 0~-L~~-L~~-L~--L~ 0 0.2 0.4 0.6 0.8 s;r (a) (b) Figure 24. Seepage volumes beneath weir structures per unit normal to the direction of flow divided by hydraulic conductivity and total pressure head. (a) Polubarinova-Kochina, 1952; (b) Harr, 1962. -50- Another procedure often used for estimating seepage flow beneath dam structures is the construction of a flow net which consists of a set of equipotential lines and a corresponding set of orthogonal flow lines. A flow net for the sheet-pile weir structure is shown in figure 25. In constructing the flow net, the alluvium and till are considered to be homogeneous, isotropic, and water saturated. In figure 25, the foundation base for the sheet pile is 7.5 feet deep. It would probably be a good idea not to disturb the glacial till during construction. A total seepage flow can be calculated by combining the flow net with the Darcy equation. Assuming only two- dimensional flow, a weir length of 116 feet, an average K of 1.87 X 10-4 ft/sec for the alluvium, and a K of 6.42 X 10-6 ft/sec for the till, then total seepage flow, o116 , would be 4.04 X 10-2 cfs or 3,490 ft 3 /day. This is our best approximation of seepage flow for this design. If the same weir design were constructed on the Clear Water Tributary at the proposed DGGS site (plate 1), then the total seepage flow would be 4.4 X 10-2 cfs or 3,800 ft 3 /day. This calculation contains the same assumptions and similar conditions as used for the Delta Creek site, except that the K value for the alluvium is 4.28 X 10-4 ft/sec (value determined from piezometer no. 1) and the weir length is 60 feet. A 15-foot wide concrete apron weir as described in the Dowl Engineers report (1982) has also been proposed. The weir design has a water head of 4.5 feet, and a basal apron foundation depth of 2.5 feet (fig. 26). A calculation for seepage flow beneath this weir design can also be done using figure 24a. In this calculation it is assumed that the glacial till is impermeable. Seepage flow is not affected much by the position of the -51- 180' 7.2' 8.4' Figure 25. site. 8' 5X Enlargement 10' 1·=1 o· Shee .. le 10' high 7.5' deep Alluvium o.a:___ Glacial Till Bedrock Flow net for a 7.5 foot deep sheet-pile weir at proposed weir -52- Alluvium concrete apron legs (Harr, 1962). Figure 24b shows that a leg placed under the middle of the apron, or under one side results in little seepage difference. If the foundation contained a single 2.5 foot leg positioned under the middle of the apron, then S/T is 0.25, b/T is 0.75, and q/Kh is 0.4 (fig. 24a). In the above calculation, q is ground-water flow per unit normal to the direction of flow, K (2.31 X 10-4 ft/sec or 19.96 ft/day) is the average hydraulic conductivity calculated for the alluvium and h (4.5 feet) is the total head behind the weir. If no piping occurs, and if the flow beneath the weir is two-dimensional, then the total seepage flow under this weir, which is 93 feet long, would be 3.87 X 10-2 cfs or 3,340 ft 3 /day. A flow net was also constructed for this weir design as shown in figure 26. In calculating the seepage flow for this weir design, the same conditions and assumptions were used as applied to the sheetpile weir flow-net calculations. The alluvium and till are homogeneous, isotropic, water-saturated, and are assumed to contain only two-dimensional flow. The average K determined from piezometers 4, 5, 6, and 7, which are the closest to the weir, is 1.87 X 10-4 ft/sec. The K for the till is 6.42 X 10-6 ft/sec. Combining the Darcy equation and flow net shown in figure 26, the total seepage flow beneath a 93 foot long weir would be 2.39 X 10-2 cfs or 2,065 ft 3 /day. This is our best approximation of seepage flow for this design. If this same weir design were constructed on the Clear Water Tributary at the proposed DGGS site (plate 1), then the total seepage flow would be 2.26 X 10-2 cfs or 2,390 ft 3 /day. This calculation contains the same assumptions and similar conditions as used for the Delta Creek site, except -53- 5X Enlargement t.c.l' 4.1'·-,---r---r-f::...l.__---lo,llo..._1·_=_1_o_· -""""""T"-.---r----.--o·----- AIIuvl..,.. 10' Glacial Till __ 4.01' 110' a.eo' / ···~ 3.11' 2.7' 2.21' Figure 26. Flow net for a 15 foot wide concrete-apron weir at proposed weir site. -54- Glacial Till that the K value for alluvium is 4.28 X 10-4 ft/sec (value determined from piezometer no. 1) and the weir length is 50 feet. Ground-water piping is a major concern in seepage flow calculations. The 11 blowouts 11 with piezometer no. 8 in the till and the 11 blowouts 11 with piezometers no. 2 and 9 in the alluvium are examples of piping. Piping leads to substantial increases in ground-water flow. Harr (1962) numerically defines the reciprocal of the maximum exit gradient for seepage flow, IE, as a factor of safety estimate against the occurrence of piping. Estimates for IE can be determined from the theoretical work of Khosla et al (1954), whose results are shown in figures 27a and 27b. IE for a sheetpile weir with a foundation 10 feet deep and an 8 foot water head is 0.25465 (fig. 27a). This has a factor of safety of 3.93. IE for the concrete apron weir with a 2.5 foot deep leg and a 4.5 foot water head is 0.246. This has a factor of safety of 4.06. Khosla et al (1954) recommends a factor of safety of at least 4 for gravel, 5 for coarse sand, and 6 for fine sand. This suggests that a concrete apron weir with a 4.5 foot water head may actually be safer with respect to piping than a sheetpile weir, although piping could be a problem with either. -55- IEs 1 h = 7i7f A=1+~ 2 (a) 0.351----+--t-+--+-t----+--+-+--+---l 0.301\ 0. 251--'t-\-+--t-+--+-t----+--+-+--+----1 0.20 '""· 0.15 ~ ........... 0.10 t-+-t-+--=~--=±--+-+-+----+-----1 r--~-. (b) 0.05 I - 0 0 10 20 30 40 50 b!s I Figure 27. Maximum exit gradient for various weir configurations (Khosla, Bose, and Taylor, 1954). -56- POTENTIAL ROCK SOURCES The best source for competent, angular shot rock is from the rock quarry situated along the road between King Cove and the airport. The rock in the quarry would be durable and consists of slightly altered diorite. The alluvium in the vicinity of the weir site is the best source for coarse gravel. The alluvial fan near the airport is a gravel quarry source and contains large volumes of gravel. Gravel for the airport runway came from this source. Numerous glacially rafted andesite boulders, weighing upwards of 10 tons are present on the glaci?l bench east of the weir site and Delta Creek canyon. Coarse rock fall talus is present west of the weir site at the base of the mountain. Therefore, whatever the construction specifications require, several local material sites are available. -57- CONCLUSIONS AND RECOMMENDATIONS Sediment transport is probably the most important condition to be addressed in the final design phase for the Delta Creek weir site. Total sediment transport includes both suspended load and bedload. Approximately 50 percent of the suspended load is sand. Most of this sand will settle out in the reservoir or be transported through the penstock. Bedload consists of sand and gravel. This material also will be deposited in the reservoir. The total sediment transport will be substantial from early spring to late fall. Further study of the annual sedimentation cycle (at least one spring to late fall cycle) should be considered before a final design and final site selection is made, especially since our data calculations are based on only one suspended sediment sample. If sedimentation is a problem in the final weir-design phase, then the DGGS alternate weir site should be seriously considered (plate 1 and fig. 28). The only drawback to the Clear Water Tributary is the lower discharge. If this is a problem only during the winter, then possibly some of the main Delta Creek flow could be diverted into the Clear Water Tributary reservoir by way of the coulees. Seepage losses should be low and not a problem if piping does not occur. Seepage flow losses in the range of 2,000 to 4,000 ft3jday would probably occur. A concrete apron weir design would probably be less conducive to piping flow than the sheetpile design. A concrete apron weir might also be easier to build because the numerous large boulders in the alluvium would not have to be removed. Also, because the glacial till appears to promote piping when saturated and disturbed, we recommend that the weir be anchored and -58- • • Figure 28. Looking up Clear Water Tributary at alternate DGGS weir site. Note recent slump on right (east) bank . • -59- penetrate alluvium only. Depending upon weir design, further soil engineering tests of the till may be required especially if the weir abutments penetrate into the till. At the proposed weir site as based on the seismic refraction surveys, the stream alluvium is about 11 feet deep in the middle of the stream canyon and the till is about 170 feet deep. The air photo lineaments interpreted from the photography may be active Quaternary faults. There does not appear to be any significant offset associated with these lineament/fault structures, and our mentioning them is not intended to inhibit design progress. However, it would be advisable to trench these structures in order to accurately document their existence and nature. -60- REFERENCES Biery, G., 1966, King Cove, Alaska, The Buyer, Oct. 5 issue, p. 12. Black, R.F., 1976, Geology of Umnak Island eastern Aleutians as related to the Aleuts: Arctic and Alpine Research 8(1), p. 7-35. Cedergren, H.R., 1967, Seepage, Drainage, and Flow Nets: John Wiley and Sons, New York, 489 p. Coats, R.R., 1950, Volcanic activity in the Aleutian Arc: U.S. Geological Survey Bulletin 974-B, p. 35-49. Colby, B.R., 1961, Effect of depth of flow on discharge of bed material: U.S. Geological Survey Water-Supply Paper 1498-D, 12 p. Dobrin, M.B., 1960, Introduction to Geophysical Prospecting, 2nd ed.: McGraw-Hill Book Company, New York, 446 p. Dowl Engineers, Tudor Engineering Company, and Dryden and LaRue, 1982, Feasibility Study for King Cove Hydroelectric Project, Vol. B, Final Report to the State of Alaska Power Authority, Anchorage, Alaska. Einstein, H.A., 1950, The bedload function for sediment transportation in open channel flows: U.S. Department of Agriculture Technical Bulletin 1026, 70 p. Freeze, R.A., and Cherry, J.A., 1979, Groundwater: Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 604 p. Harr, M.E., 1962, Groundwater and Seepage: McGraw-Hill Book Company, Inc., New York, 315 p. Hough, B.K., 1969, Basic Soils Engineering, 2nd ed.: The Ronald Press Company, New York, 634 p. Hvorslev, M.J., 1951, Time lag and soil permeability in groundwater observations: U.S. Army Corps of Engineers Waterways Experimental Station Bulletin 36. Jacob, K.H., and Hauksson, E., 1983, A seismotectonic analysis of the seismic and volcanic hazards in the Pribilof Islands -eastern Aleutian Islands region of the Bering Sea: final report to U.S. National Oceanic and Atmospheric Administration under contract NOAA 03-5-022-70, Lamont-Doherty Geological Observatory of Columbia University, New York, 224 p. Kennedy, G.C., and Waldron, H.H., 1955, Geology of Pavlof Volcano and vicinity, Alaska: U.S. Geological Survey Bulletin 1028-A, p. 1-19. Khosla, R.B.A.N., Boss, N.K., and Taylor, E.M., 1954, Design of weirs for permeable foundations: Central Board of Irrigation, New Delhi, India. Knott, J.M., and Lipscomb, S.W., 1983, Sediment discharge data for selected -61- sites in the Susitna River Basin, Alaska, 1981-82: U.S. Geological Survey Open-File Report 83-870, 45 p. Lamke, R.D., Still, P.J., Bigelow, B. B., Seitz, H.R., and Vaill, J.E., 1983, Water resources data, Alaska water year 1982: U.S. Geological Survey Water-Data Report AK-82-1, 363 p. Leopold, L.B., Wolman, M.G., and Miller, J.P., 1964, Fluvial Processes In Geomorphology: W.H. Freeman and Company, San Francisco, California, 522 p. Nakamura, K., Jacob, K.H., and Davis, J.N., 1977, Volcanoes as possible indicators of tectonic orientation -Aleutians and Alaska: Pageoph 115, p. 87-112. Polubarinova-Kochina, P.Ya., 1939, On the continuity of the velocity hodograph in plane steady motion of ground water: DAN 24(4) {Russian). Reeder, J.W. 1984, An analysis of fault and volcanic dike orientations for the Makushin Volcano region of the Aleutian arc, Proceedings of the International Symposium on Recent Crustal Movements of the Pacific Region: held February 9-14, 1984 at Victoria University, Wellington, New Zealand, Royal Society of New Zealand Bulletin, in press. Selkregg, L.L., ed., 1974-77, Alaska Regional Profiles--Southwestern Region: University of Alaska Arctic Environmental Information and Data Center, 313 p. Waldron, H.H., 1961, Geologic reconnaissance of Frosty Peak volcano and vicinity, Alaska: U.S. Geological Survey Bulletin 1028-I, p. 677-708. -62- Appendix A Discharge measurement notes for Delta Creek and the Clear Water Tributary of Delta Creek -63- Appendix A-1 S t a t e o f A 1 a s ka Dept. of Natural Resources Div. of Geological & Geophysical Surveys jj,;.e_ Mcu. No. ------------· WATER RESOURCE INVESTIGATIONS Camp. by -------------- DISCHARGE MEASUREMENT NOTES C"-c.lr.td by------------- Sta. No. _________________ _ i'rlf"' &. ~' r.;,:J r!'cv~ (if llfP. t;-~~ '!>,...:Z:........_) ~~~~-~---~~~i--~-~--------~~:·;~---~_(_ ---~~rty·_-_·;~;-_--/i(?i_i!,_-_-_!~{j~~-t_;~·::?•!~-~:~:. Width ____________ Arra ____________ Vel. __ ____ G. H. __ _ ___ Di•ch. ;:_o--"--~"'-'-'-'--'""":..:. Mrthod ----------No. &eCJ. ____________ G. H. change ___________ in--------hn. Susp. ______ _ Method corf. ______ Hor angle coef. _-____ Susp. cod.----------Mrter No. ---------------- ---~--~---G~E_ READINGS Type of mettr __ ft:!_r !'__ /f_/l__r:-_~_r_~_.,._ _ Meter _________________ ft. above bottom of ,.-eight. Spin before meas. L'3.f.t_~.,_PC:_ after _/_'~-~ • ___Ij~~i ______ :_R~o:~l~~~~ Outside_ --------------------__________ ! __________ ---------- 1 Date rat.d . for rod, other. Meas. plots ____ !j, dill. from rating ____________ _ ~able, ICe, boat. upstr., downstr., silk bridge _ _ __ feel mile. above, below -----------------I--------- i ---------(-- gage. and _ Che,k-bar. found ------------. -I --- W<ighted MG. H. ___ :-=-:==-~---1_,--- C. H. toni"Ction ____ ! __ changed to . at Correct -,-- C.,..,...,t M G H. i 1 --------------- MeaLrrrnent rated excellent (2~'() ~•ud (S~~~~r (over 87c;). ksd on following con:!itions· Cross section _f.Prf._,_Y7.~,,. ~,r: t r.,_~~{"-J/-.-~!e:.~---/- 5-Q---;/ ... , ... ,--.... ,......-r-. -<,..-.-.., 7.,.,; ~--c Flow ______ _____ _ _______ __ _ __ _ __ Weathrr .. r~-J.;.?-_,... __ !f''!':A-..7 _______ _ Other_ -------------------_______ Air __________ °F@ _______ _ G"i• ---------------------------------______________________ Wat.r ______ "F@ _______ _ u . _ -----____ -----------____ Rrcord reJnoved . lnti>ke Aus!xd L ------------- Obaerver --------7-----------c·------------------ T ::;./·-.<. ~----------------------------.-C.----------------- Rvr.a.rlu· G. H. of zero flow _ . ft. s /68 .0 .10 .20 -~~~ Appendix A-1 .50 Rh·er at-- .60 .7Q .7S Appendix A-2 I' State of A I aska Dept. of Natura} Resources Div. of Geological & Geophysical Surveys WATER RESOURCE INVESTIGATIONS DISCHARGE MEASUREMENT NOTES Me .. No. -~-~--~_<;_:_. Comp. ~~------ Ch«ked by Sta. No. --------------------_fl!"..~(~---(';.Jd~-t-j;~(-~~;;;f-.-./:~}J;.'{~-p!f{/!ft!j~{;;ii;.f~~~ (' Date .Jl~j~--l.-f. ---·---• 19.~4-Party . ..lf<@..ed'.e-r--·-----····-···-············------·--···· Width ------------Ana ····-------· Vel. C. H. ........... Disch .................. .. Method .......... No. sees. G. H. change ............ in ........ hn. Sutp ....... . Method toef. .. .. Susp. coef. ------------''--------, Type of meter Date rated . for rod. other. Meter . . . .... ft above bottom of weight Spin before rneas. (_~Y~~-~--after lw t'~-,~c • Meas. plot> .. '"; diff. from rating Gable. ice, boat, upstr .. dowmti .. side bridge .... _ feet. mik above, below gage, and . Check-bar, found changed to at Correct . G. H. tPrr~tKJnww-~-w Com<~ M G. H I Level~ obtained Measurrr.oent rated excellent (2:(). good (5%). fair (8s;)), €£o~~Lil..%l .. "'":t.aaed on following con:li~ions: Cross section .J?~c:· _(J.':'./J.~-t;i;il:'§/· -~~:};, .. -:!:!.":.~-~~-'-~ Flow ..... !.£5?' .. -~~-!:.".::.. Weather .. Jf"'.;:'!.;t .. !: .. C-!J,_T _____________ __ Other------------------------------------------------Air ........... 0 f@ ....... . Gqe ---------------------______ ................ . Water . ... .. Cf@ ........ -------··---------------------Rtwrd removed Intake flu~hed Ob~oerver -------------------------------------.... _ ............................ .. R~ma.rlla ~---------- G. H. of zero flow .................. . S-768 --h. .> ~ ,.,,..,, • ..,., ec''""' ''"<""'"''" ........ .__,._,. .0 .10 .:G .SG Appendix A-2 .50 River at- .60 .70 .75 Appendix A-3 State of Alaska Dept. of Natural Resources Div. of Geological & Geophysical Surveys Meu No. _t!_~·_!:E__! WATER RESOURCE INVESTIGATIONS Comp. by _:;;_""!:f ___ _ DISCHARGE MEASUREMENT NOTES Chool..d by------------- Sta. No.___________ d. .. 0.~-;=_-;;;:.:{._ ?'~ r T ~ f~fRr pii~;ff;•---@_ __ _((t~l_f-t(t~~ -' ____ _ Date-------------·-__ , 19 ______ Pa.ty _________________ -------------------------------------- Width ____________ Area___________ Vel. ____ C. H ___ , ________ Disch. ______________ ---- Method__________ No. sw1. ____________ G. H. change------------in--------hrs. Susp. ------- ;_M_:e_th_o.:;.d_c..:.o.:.:ef_. ::..··--...:··::..· _:_:_..:.H..:.or___:_a~n~gl~e..:.c~oe~f. __ , . Sus~. cod. '-··-;;·;-:~ter ~/1-~-~-,.-t-·;,.·· _______ C~G~_R_t:_AI:>_IN_Gs_ ___ ~ Type of meter ________________________________ _ • I • I • ! ••••••• •··••• •·• • -------,.T ________ T _________ ~------------------- ::: •• •· •• •1•:•:: ·•·•1: ·••• •• i•• •· • •· • · · . --------: ____ ------I-.--------; __ ------- ! ! ;.ight.dM c:H. ___ -. -_ 1 1 ~ = G. H. corrl"'nion .. ____ )~~--~ i ----~~---- ! i CorrM:t MG. H ____ . _____ , Other c..,e ----------------------------------- Re<.ord r:111oved Date rated ______________________ for rod. other. Meter-----------------ft. above bottom of weight. Spin before meas. _/_"_'lf:ft:C after 4."":"_:·.~-- Meas. plots ____ '/(dilL from rating ____________ _ ~able, ice, boat, upstr.. downstr., side bridge ______________ feet. mile, above, below ~age. and Check-bar. found changed to _. _______ ~--. _ at Water ______ °F@ -------- u Intake Aushed L ___ . ----------- -----.. --------.7·---------------------------------- .0 .10 .H .341 .0 .tO • .?0' .30 Appendix A-3 .10 .50 River at--- . .iO .60 .. .60 .70 .75 .10 .75 ' . , ) ... -t ' -----' \ ---------------1 \ .,...'-_ ........ ~ ', '----.......... -1. I \ ' -+- -~7 s:_.\ -o-..... ~0 /1'-'---',. . _,;72:--~ -t y .> / SCA L E 1·=1 0 0 ' 200 ' ~~~~~~~~3~00 ' ---I_ I I ~ Contour ln terval =2' ~ ~~.// ~ Piezome t er S t ation ~ --(! / Geophone )-. "-. Se 1sm ic Li ne / , I ~~/"-. H ~,,---- a m mer Po i nt ~~~hot Po i nt ----/f' PLATE +2 STATE OF AIJ.sKA ~~ant of flabnl ..... DIVISIOR of Geof-I Q , P.O. Box 77 2116 liii•QqlllllbiiNI IIIIIJI . . \ / I I / ) I PLATE #3 I i \ -----~ -- \ STATE Of AlASKA llepatp IIIII ...... ,.. = Dim. il 8111 jull II,I;IICld lniJs P.O. Box n2116 Eaglt! River, Alaslla 99577