Moose_Creek_Mix_Design_Report_FINAL.pdf

Golder, Golder Associates and the GA globe design are trademarks of Golder Associates Corporation

MIX DESIGN REPORT

WEIR GEOTECHNICAL INVESTIGATION AND MIX

DESIGN FOR THE MOOSE CREEK DAM SAFETY

MODIFICATION STUDY

CHENA RIVER LAKES FLOOD CONTROL PROJECT,

NORTH POLE, ALASKA

Contract W911KB-17-D-0002

Task Order No. W911KB17F0034

Submitted To: US Army Corps of Engineers, Alaska District ATTN: CEPOA-EN-G (Robert Weakland, PE)

P.O. BOX 6898

JBER, AK 99506-0898

Submitted By: Golder Associates Inc.

2121 Abbott Road, Suite 100 Anchorage, Alaska, USA 99507

Distribution:

2 Copies + 1 CD – US Army Corps of Engineers, Alaska District 2 Copies – Golder Associates Inc.

February 27, 2018 1781138

M

IX

D E

S

IG

N R

E P

O R

T

February 2018 i 1781138

MCD Mix Design Report

Table of Contents

1.0 INTRODUCTION

1.1 Project Description

1.2 Purpose and Scope

2.0 FIELD EXPLORATION

2.1 USACE Soil Sampling Program

2.2 Relief Well Water Sampling Program

3.0 SITE CONDITIONS

3.1 Regional Geology

3.2 Regional Seismicity

3.3 Surface Conditions

3.4 Inferred Subsurface Conditions

4.0 LABORATORY TESTING

4.1 Soil and Soil-Cement Testing

4.1.1 Soil Testing

4.1.2 Soil-Cement Testing

4.2 Relief Well Water Testing

5.0 MIX DESIGN BACKGROUND AND DISCUSSION

5.1 Barrier Wall Mix Design Goals

5.2 Mix Design Background

5.3 Cement and Bentonite Discussion

5.4 Mixing Water Source

5.5 Typical Engineering Parameters of the Soil-Cement

5.6 Discussion of Subsurface Variability

6.0 MIX DESIGN BENCH-SCALE TESTING

6.1 Mix Design Procedure

6.1.1 Soil Preparation

6.1.2 Soil-Cement Mixing

6.2 Bench-Scale Program Outline

6.2.1 Composite Summary

6.2.2 Soil-Cement Mix Summary

6.2.3 Trial Mixes

6.2.4 Test Mixes

7.0 MIX DESIGN RESULTS

7.1 Unconfined Compressive Strength (UCS)

7.2 Tensile Results

7.3 Permeability

7.4 Mixing Water

February 2018 ii 1781138

7.5 Design Strength Variation

7.5.1 Laboratory and Construction UCS Variation

7.5.2 Admixture

7.6 Typical Pre-Construction Quality Control

7.7 Baseline for Contractor Bids

8.0 LIMITATIONS AND USE OF REPORT

9.0 CLOSING

10.0 REFERENCES

List of Tables Table 2.1 Borehole Summary Table 6.1 Composite Summary Table 6.2 Summary of Trial Mixes Table 6.3 Summary of Test Mixes Table 7.1 Mix 8 through 24 Summary, 7- and 28-Day Average UCS Results Table 7.2 Average Splitting Tensile Strength Summary – 28-Day Strength Table 7.3 Various Mix Design Parameters for the Minimum and Maximum Test Mixes

List of Graphs Graph 7.1 Mix 8 through 24, 7-Day Average UCS Results Graph 7.2 Mix 8 through 24, 28-Day Average UCS Results Graph 7.3 Average Splitting Tensile Strength vs Binder Factor

List of Appendices Appendix A Photo Log Appendix B Sieve Results Appendix C UCS – Test Mixes 7-Day Results Appendix D UCS – Test Mixes 28-Day Results Appendix E UCS – Trial Mixes Appendix F Splitting Tensile Strength – 28-Day Results Appendix G Permeability Testing Appendix H Relief Well Water Testing Appendix I Product Information Appendix J Mix Design Calculations Appendix K Resumes of Laboratory Oversight Personnel and DoD ELAP Certification

February 2018 1 1781138

1.0 INTRODUCTION

1.1 Project Description

Golder Associates Inc. (Golder) has been contracted by the U.S. Army Corps of Engineers (USACE), Alaska District to perform the services outlined in the revised Statement of Work (SOW) for the “Weir

Geotechnical Investigation and Mix Design, Moose Creek Dam Safety Modification Study, Chena River

Lakes Flood Control Project, at North Pole, Alaska,” dated October 17, 2017 (Contract No. W911KB-17-D-

002, Task Order W911KB17F0034 MOD2). The work entails executing a subsurface exploration program and providing a geotechnical data report for proposed seepage weirs and a mix design report for a bentonite-portland cement mix design for a proposed mix-in-place barrier wall at the Moose Creek Dam in

North Pole, Alaska. The geotechnical report for the proposed seepage weirs is presented under a separate cover and not discussed in this report.

1.2 Purpose and Scope

This report presents a summary of Golder’s activities related to developing a mix design for the proposed barrier wall at Moose Creek Dam. The USACE collected the subsurface soil samples for the mix design during the geotechnical exploration program along the embankment at Moose Creek Dam and the samples were transported to Golder’s Anchorage, Alaska laboratory for the mix design efforts.

The work was accomplished in the following six tasks outlined in the SOW:

Task A – Project Management and Plan Development

Task B – Field Exploration, including drilling and sampling (presented in a separate report)

Task C – Geotechnical Laboratory Testing, including index testing of soil samples (presented in a separate report)

Task D – Geotechnical Data Reporting (presented in a separate report)

Task E – Bench-Scale Testing of mix designs covering a range of strengths

Task F – Mix Design Reporting of the results related to the bench-scale testing

February 2018 2 1781138

2.0 FIELD EXPLORATION

2.1 USACE Soil Sampling Program

The geotechnical exploration for the barrier wall and mix design was managed and performed by the

USACE. The samples for the barrier wall mix design bench-scale testing were obtained from boreholes advanced along the top of the Moose Creek Dam embankment. The USACE program for the barrier wall mix design consisted of advancing 22 geotechnical boreholes with depths ranging from approximately 90 feet to 130 feet. The boreholes were advanced using a sonic drill rig, which performed continuous sampling through the embankment and underlying foundation soils. Select portions of samples were bagged by

USACE personnel and tested as part of the barrier wall geotechnical investigation. Results from these tests are presented separately from this report. The remainder of the samples were collected by USACE personnel in five-gallon buckets and shipped to Golder’s Anchorage laboratory for barrier wall mix design bench-scale testing. Table 2.1 summarizes the number of bulk samples received by Golder for the mix design bench-scale testing. Note that some buckets contained numerous small samples contained in separate plastic bags.

Table 2.1: Borehole Summary

Temporary Borehole Number

Permanent Borehole Number Number of Bulk Samples

TB-01 AP-2943 10

TB-04 AP-2946 9

TB-06 AP-2948 10

TB-08 AP-2950 10

TB-11 AP-2953 11

TB-13 AP-2955 10

TB-15 AP-2957 20

TB-18 AP-2960 16

TB-20 AP-2962-P 15

TB-23 AP-2965 14

TB-25 AP-2967 13

TB-28 AP-2970 13

TB-30 AP-2972 18

TB-33 AP-2975 15

TB-35 AP-2977 18

TB-37 AP-2979 19

TB-39 AP-2981 18

TB-42 AP-2984 18

TB-44 AP-2986 18

TB-46 AP-2988 19

TB-49 AP-2991 23

TB-51 AP-2993 10

February 2018 3 1781138

2.2 Relief Well Water Sampling Program

Golder sampled water from nine relief wells at Moose Creek Dam for use in the mix design. The relief well locations were selected by the USACE, however, water could not be accessed at Relief Well 441+50 due to a locked lid, and therefore was collected from Relief Well 441+75. Water was sampled from the following relief well locations:

Relief Well 270+35

Relief Well 290+25

Relief Well 310+25

Relief Well 330+35

Relief Well 360+10

Relief Well 378+07

Relief Well 418+00

Relief Well 441+75

Relief Well 478+00

Water was pumped using a small submersible pump capable of pumping approximately eight gallons per minute. Water from each relief well was pumped for approximately five minutes prior to collection or until a visible color change occurred in the water. Approximately 10 gallons of water was sampled from each well.

Water was collected in five-gallon buckets and shipped to Anchorage for additional testing.

February 2018 4 1781138

3.0 SITE CONDITIONS

3.1 Regional Geology

The Moose Creek Dam project site is within the Tanana Basin (Ridgway et al. 2007), located southwest of the Yukon-Tanana uplands. Surface sediments within the project area are mapped as Quaternary flood-plain alluvium (Qa) and swamp deposits (Qs). The alluvium facies is described as containing well-stratified layers and lenses of unconsolidated soil that includes pebbles, cobbles, and boulders, ranging in size from

¼-inch to three feet in diameter, of resistant rocks from neighboring uplands. Swamp deposits consist of peat and silt layers more than five feet thick in areas of poor drainage and where ponded water is present year-round. Small swamps and areas of seasonal flooding can be expected in both of the mapped units

(Péwé et al. 1966).

3.2 Regional Seismicity

The project site is located within the Fairbanks Seismic Zone (FSZ). The FSZ is part of a series of northeast-trending seismic zones located in interior Alaska that are situated south of the Tintina fault zone and north of the Denali fault system (Koehler et al. 2012). Deformation of the North American continental plate resulting from the collision of the Yakutat microplate along the southern Alaska coast is responsible for seismic activity in the seismic zones. As the Yakutat microplate converges with the North American continental plate, the collision drives deformation across Alaska, including right-lateral slip on the Denali fault, deformation of the Northern Foothills Fold and Thrust Belt, and seismicity and faulting in the interior

Alaska seismic zones.

The FSZ is a zone approximately 67 miles long in the northeast-southwest direction at its maximum extent

(Koehler et al. 2012). The FSZ represents a northeast-trending band of historic seismicity. On June 21, 1967, a series of three earthquakes occurred within the FSZ with epicenters located about 15 miles northwest of the project area (Ruppert et al. 2009; Stover and Coffman 1993). These earthquakes ranged in magnitude from surface-wave magnitude (Ms) 5.5 to 5.9. The earthquakes resulted in building damage and ground cracks in roads in the Fairbanks area. These earthquakes were followed by a Richter magnitude

(ML) 5.6 earthquake on June 23, 1967, with the epicenter located about eight miles west-northwest of the project area.

In 1937, the Ms 7.3 Salcha earthquake, the largest recorded earthquake in the interior Alaska seismic zones, occurred in the central portion of the Salcha Seismic Zone (SSZ) (Ruppert et al., 2008; Stover and Coffman, 1993). The SSZ is located to the southeast of the FSZ. Focal mechanism data for the 1937 earthquake suggests left-lateral motion of a steeply-dipping fault (Page et al. 1991). Based on the epicenter location provided by Ruppert et al. (2008) for the 1937 earthquake, the earthquake epicenter is located approximately 13 miles south-southeast of the project area, although the uncertainty in latitude and longitude coordinates for the epicenter may be as large as tens of miles. For interior Alaska and Fairbanks, February 2018 5 1781138 the 1937 earthquake produced cracks and mudboils in roads, a landslide across the highway at Salcha

Bluff, and broken windows and loss of merchandise in Fairbanks (Stover and Coffman, 1993). No documented evidence of surface fault rupture was found for the 1937 earthquake (Ruppert et al. 2008).

3.3 Surface Conditions

The project site consists of a 7.5 mile long dam located in North Pole, Alaska. The dam consists of an earth embankment and a concrete control structure which the Chena River flows through. The crest of the earth embankment is estimated to be 30 to 40 feet higher than the surrounding ground in the area. A road open to the public is located on the west side of the embankment to the south side of the control works.

Geotechnical boreholes for the barrier wall conducted by the USACE were advanced on top of the embankment.

3.4 Inferred Subsurface Conditions

Drilling and sampling was conducted by USACE personnel. Golder has received draft boreholes logs dated

October 20, 2017, and from these logs, the subsurface conditions were inferred. In general, the subsurface conditions consisted of sand and gravel embankment soil overlying similar sand and gravel in-situ soil. For this mix design, we have considered that soil samples collected from 0 to 40 feet are ‘embankment’ soils.

Soil samples collected below the embankment have been classified as ‘foundation’ soils. The fines content

(percentage passing the US #200 sieve) of the embankment soil were generally less than 12 percent. The foundation soils below 40 feet have lower fines contents, typically less than five percent. The depth to bedrock below a majority of the dam is unknown to Golder, however, we understand shallow bedrock was observed in the northern extent of the project in Boreholes TB-49 (AP-2991) and TB-51 (AP-2993). USACE encountered shallow weathered bedrock at approximately 11 feet in Borehole TB-51 (AP-2993) underlain by hard bedrock at approximately 34 feet. Weathered bedrock was encountered at approximately 90 feet in Borehole TB-49 (AP-2991) which extended to borehole termination.

February 2018 6 1781138

4.0 LABORATORY TESTING

This section presents the three types of laboratory testing performed for the mix design effort; soil testing, soil-cement testing, and relief well water testing. Photos of several aspects of the testing process and bench-scale laboratory testing are presented in Appendix A. Additional details of the testing procedures and methodology are presented in subsequent sections of this report.

4.1 Soil and Soil-Cement Testing

The bulk soil samples for mix design received by Golder were inventoried and each bulk sample was split into two representative portions. One half of the bulk sample was set aside for use in the soil-cement mix design and the remaining half of the sample was tested in Golder’s USACE-validated geotechnical laboratory in Anchorage, which was validated through November 2017. Several test methods related to the mixing, casting, and testing of the mix design specimens were performed outside of USACE validation, but were overseen by John Thornley, PE, who has specific experience in those test methods in compliance with the ASTM procedures. John Thornley is the laboratory technical director, who oversaw all testing performed by Golder, and has experience with portland cement concrete field and laboratory testing and advanced testing of soil and concrete, including triaxial and permeability testing. John’s abbreviated resume is included in Appendix K.

4.1.1 Soil Testing

Laboratory testing consisted of soil classification to Unified Soil Classification System (USCS) standards by grain size analysis (ASTM C136), moisture content (ASTM D2216) and specific gravity (ASTM D854).

Results of laboratory testing are presented in Appendix B. The results have been summarized by the material used in each mix design composite sample. Samples that were tested and not used in any soil composite for mix design have been summarized in “Unmixed Samples”.

The soil testing was performed in order to:

Provide a soil classification and gradation for the bench-scale laboratory testing

Measure other index properties, which provide a basis for estimating engineering parameters used in geotechnical analysis and design

4.1.2 Soil-Cement Testing

As part of the mix design effort, mixtures of the split bulk soil samples, bentonite, and varying amounts of portland cement were combined and 6–inch diameter by 12-inch height cylinders were cast for the bench-scale strength and permeability testing. The cylinders were generally cured for 7 and 28 days before testing.

In all, 21 mixtures were combined, representing varying portland cement content for composite mixtures of foundation and embankment soils. The soil-cement testing consisted of the following tests:

Unconfined Compressive Strength (ASTM D2166)

February 2018 7 1781138

Splitting Tensile Strength in general accordance with ASTM D3967 and C496

Permeability (ASTM D5084)

The results of the 7-day and 28-day unconfined compressive strength (UCS) testing are presented in

Appendix C and D, respectively for Mixes 8 through 24. Additional unconfined compressive strength testing for the initial experimental batches, Mixes 1 through 7, are presented in Appendix E. The results of the splitting tensile testing is presented in Appendix F. The results of the permeability testing are presented in

Appendix G.

4.2 Relief Well Water Testing

The relief well water samples were evaluated following the standard for Mixing Water Used in Hydraulic

Cement (ASTM C1602) and the following tests were performed on select samples in accordance with the following tests and specifications:

Analytical Water Testing by Ion Chromatography

Chloride using EPA 300.0 method

Sulfate using EPA 300.0 method

Sodium using EPA 6020A method

Potassium using EPA 6020A method

Time of Setting of Concrete Mixtures by Penetration Resistance (ASTM C403)

Compressive Strength of Hydraulic Cement Mortars Using 2-inch Cubes (ASTM C109)

Representative water samples were collected from each five-gallon bucket and transported to SGS’s

Anchorage laboratory for the analytical water testing. The EPA 300.0 method extraction and analysis is equivalent to the EPA 9056A method, with the difference that the 9056A method requires a matrix spike and matrix spike duplicate in each analysis batch. SGS is certified by the Department of Defense

Environmental Laboratory Accreditation Program (DoD ELAP) to perform the EPA 9056A and EPA 6020A test methods, and the certificate is provided in Appendix K. The Relief well water samples were provided to DOWL’s USACE-validated geotechnical laboratory for the time of setting test. The DOWL tests were performed outside of USACE validation, but were overseen by Maria Kampsen, PE, who has significant laboratory oversight experience. Maria Kampsen’s resume is included in Appendix K. DOWL prepared soil-cement samples for the time of setting test using a typical fine concrete aggregate, Type I cement, and relief well water. A control specimen was prepared using distilled water to compare the effects of relief well water on time of setting. Golder’s Anchorage laboratory prepared mortar samples for two water samples using a sample from the relief well and distilled water and silica sand. The compressive strength of the mortar samples was tested seven days after mixing. The results of the analytical water, time of set, and mortar compressive strength testing are presented in Appendix H. Additionally, five mixes were prepared using relief well water. The remaining mix design trials were performed with potable water.

February 2018 8 1781138

5.0 MIX DESIGN BACKGROUND AND DISCUSSION

The bench-scale laboratory testing program was completed using soil samples collected and provided by the USACE and commercially available bentonite and portland cement. This section presents our understanding of the mix design program in Section 5.1, and summarizes typical mix design basis in Section

5.2. Sections 5.3 to 5.5 present a discussion of the cement and bentonite and mix water, and the typical properties of soil-cement. Section 5.6 provides a discussion of the variability in subsurface conditions between the foundation and embankment soils.

5.1 Barrier Wall Mix Design Goals

The intent of the barrier wall bench-scale laboratory testing program was to evaluate the amount of cement binder and bentonite required to meet a range of unconfined compressive strengths from 150 pounds per square inch (psi) to 1000 psi. The USACE indicated that the unconfined compressive strength was to be tested at 7 days and 28 days after batching the mixes. The SOW indicated that the unconfined compressive strength was to be evaluated with a minimum of five mixes by varying the water to cement ratio.

5.2 Mix Design Background

The basis for our mix design is the Federal Highway Administration Design Manual: Deep Mixing for

Embankment and Foundation Support (FHWA 2013). While the manual describes wet and dry soil mixing, the anticipated method for barrier wall construction will utilize the wet method. Wet soil mixing entails the mechanical mixing of the in-situ soil with a slurry primarily consisting of binder and water. Binder is any chemically reactive agent, such as cement, that provides additional strength to in-situ conditions after mixing. In this mix design, cement was used as a binder agent. With typical Cutter Soil Mixer (CSM) equipment, bentonite is used to assist in penetration of the mixing tool on the down stroke, and then the cement binder is delivered at the bottom of the element on the upstroke of the tool.

During construction, the contractor controls the cement slurry mix by two key parameters, the water to binder ratio (w:b) and the volume ratio (VR). The water-to-binder ratio is defined as the weight of water in the cement slurry to the weight of the binder, and the volume ratio is the volume of the cement slurry divided by the volume of the soil. For a specific w:b ratio, the VR is controlled to achieve other target parameters during mixing, such as the binder factor or binder content. The binder factor is the weight of binder divided by the volume of treated soil, and the binder content is the weight of the binder divided by the weight of the slurry (FHWA 2013). Since bentonite is used as the CSM mixing tool is advanced, the amount of bentonite does not typically vary.

The range of specified unconfined compressive strengths provided in the SOW falls on the higher end of typical strengths for soil-cement samples and is outside the typical range discussed in the FHWA design manual. Golder’s cutter soil mixer construction personnel, located in Vancouver, British Columbia, identified that the binder factor typically ranges from about 16 to 31 pounds per cubic foot (lbs/ft3) (432 to 837 pounds

February 2018 9 1781138 per cubic yard (lbs/yd3)). Additionally, we understand the bentonite factor is approximately 1.8 pounds of bentonite per cubic foot of treated material (49 lbs/yd3). We have based our mix design bench-scale testing following the typical binder quantities and the anticipated variation in the material.

5.3 Cement and Bentonite Discussion

In the mix design bench-scale testing, locally available Type I/II cement was utilized and three different types of bentonite were evaluated. The initial trial batches, Mixes 1 through 7, were prepared with two different commercially available bentonite types that were intended for well drilling. After additional research, Barakade SP bentonite was obtained, which is typically used in barrier wall construction. This bentonite was utilized in the remaining bench-scale batches, Mixes 8 through 24. Due to the likely strength effects from the well drilling bentonite, the results of Mixes 1 through 7 are considered experimental and are presented separately. The product information for the different products used are presented in Appendix

I.

5.4 Mixing Water Source

The influence of the mixing water was evaluated following ASTM C1602. Additionally, five mixes were prepared using water from Relief Wells 270+35 and 290+25. The remainder of the mixes were prepared using potable water from the tap at Golder’s Anchorage laboratory.

5.5 Typical Engineering Parameters of the Soil-Cement

Some of the engineering properties of the soil-cement used in the soil mixing design include the compressive strength, tensile strength, permeability, elastic modulus, and unit weight. These properties can be highly variable and are primarily controlled by the contractor and the quality control of the work. The strength of the treated soil is dependent on the binder materials, in-situ soil, and the use of additives in the soil-cement mixture.

We have assumed soil-cement engineering properties from the FHWA deep mixing manual (FHWA 2013) and a review of literature. As a starting point, the FHWA manual provides an empirical relationship of the compressive strength of the treated soil and total amount of water and the amount of binder used in the mix. There is significant scatter in the published data, but the general trend is that the 28-day UCS decreases as the total water to cement ratio increases. The elastic (secant) modulus of the soil-cement is also determined from the UCS testing. It has been found that the ratio of the elastic modulus to the UCS can range from 75 to 1,000 but is generally around 300. The unit weight change of treated soils is generally negligible (FHWA 2013).

5.6 Discussion of Subsurface Variability

Based on our review of the draft borehole logs, laboratory testing, and our understanding of the geology, there appears to be little horizontal variability in the makeup of the subsurface soil along the alignment of

February 2018 10 1781138 the dam. It is important to note that Golder did not evaluate the consistency or density of the soil, as that data is not measured during sonic drilling. The main variability in the subsurface conditions occurred vertically between the embankment soil and the foundation soil. In general, both the embankment and foundation soil were a mixture of sand and gravel with variable fines content. However, the foundation soil generally had a lower fines content. The mix design bench-scale testing program was developed to evaluate the variability between the embankment soils and foundation soils, and the resulting effects of soil-cement strength. We understand the barrier wall may be installed in zones of weathered bedrock near Boreholes

TB-49 (AP-2991) and TB-51 (AP-2993). However, for our bench scale testing, it was not feasible to include weathered bedrock in our laboratory testing or mix design. Since the extent of the bedrock is limited, it is assumed that with proper mixing of weathered bedrock, portions of the barrier wall would display similar strength to the granular testing material.

February 2018 11 1781138

6.0 MIX DESIGN BENCH-SCALE TESTING

This section presents the mix design bench-scale testing procedure and outline of the blending of selected mixes. Photos illustrating several steps of the bench-scale mix design procedure are presented in Appendix

A.

6.1 Mix Design Procedure

6.1.1 Soil Preparation

As buckets of soil arrived at the Golder lab, they were inventoried and organized by borehole. Each bucket was then split into two samples using a large mechanical splitter. The material generally had a nominal aggregate size of three inches, and no over-size aggregate was screened out during soil composite preparation. If there was free water in the bucket or if it was determined that the sample was too wet for splitting, the entire sample was air dried prior to splitting. After splitting, half of the sample was saved for the mix design bench-scale testing, and the other half was saved for index testing.

Once samples were selected for mix design bench-scale testing, they were blended together generally following the guidelines in AASHTO T248, “Standard Method of Test for Reducing Samples of Aggregate to Testing Size”. The blending consisted of dumping buckets of the selected soil, typically four buckets at a time, onto a heavy plastic tarp, and then mixing the material a minimum of four times by pulling each corner of the tarp horizontally over the sample towards the opposite corner. After the soil was sufficiently mixed, it was quartered into four approximately equal piles. Piles opposite from each other were collected into the same bucket, with care taken to collect the fines that remained on the tarp. Buckets were organized into two piles after quartering. Once all of the selected buckets had been processed, the process was repeated with soil from one of the bucket piles until the necessary amount of soil was obtained for mix design. The soil in the unused mix buckets was discarded.

6.1.2 Soil-Cement Mixing

Mixing was conducted using a gasoline-powered concrete mixer capable of mixing around 300 pounds of material at once. The procedure to complete the mixing process generally follows ASTM D4832 and is outlined in the following section. However, due to the large amount of coarse aggregate in the mix, the cylinders were prepared following ASTM C31.

Bentonite and water were hydrated for at least one hour and were then mixed for approximately 15 to 20 minutes in a five-gallon bucket until a slurry paste with no major solids was obtained. Mixing was performed until the bentonite slurry reached a uniform consistency. A hand drill and paddle mixer was used to mix the bentonite and water. Once the bentonite slurry was ready, a marsh funnel reading was obtained following ASTM D6910.

Cement and water were mixed in a five-gallon bucket until a slurry paste with no major lumps or dry pockets was obtained. Typically, this took less than 15 minutes. A hand drill

February 2018 12 1781138 with paddle mixer was used to mix the cement and water. Once the cement slurry was ready, cement content was measured using a mud balance, following ASTM D4380.

Soil was added to the concrete mixer and mixed for approximately five minutes. The bentonite slurry was then added and mixed for five minutes, followed by addition of the cement slurry.

The material was mixed for up to 20 minutes, or until no dry pockets were observed in the mixture.

After completion of the mixing, the soil-cement was transferred to a wheelbarrow and hand mixed for one minute. Eight 12-inch high by 6-inch diameter cylinders of the mixture were then cast. The mixture was added to the cylinders in three equal lifts and the material was tamped using a metal rod and hand tamping the outside of the cylinders at each lift to remove air voids. Any 3-inch aggregate was hand plucked from the mix when the cylinders were cast.

Cast cylinders were stored in the laboratory for approximately one day prior to being placed in a hydration bath.

The cylinders were left in the molds and placed inside of water lime baths to hydrate for seven days.

After seven days of curing, the samples were stripped out of the molds and placed back into the water lime bath.

6.2 Bench-Scale Program Outline

6.2.1 Composite Summary

A total of nine soil composite blends were developed for the bench-scale testing. The composite blends are summarized in Table 6.1 along with the soil source and composite blend USCS classifications. Little variability was observed in the subsurface conditions spatially, therefore, boreholes for each soil composite were generally selected to represent each third of the proposed barrier wall alignment from South to North.

Table 6.1: Composite Summary Composite Number

Soil Unit1 Borehole Origin Composite

USCS

Classification

Composite 1 Embankment

TB-01 (AP-2943), TB-04 (AP-2946), TB-06 (AP-

2948) GP-GM

Composite 2 Foundation

TB-01 (AP-2943), TB-04 (AP-2946), TB-06 (AP-

2948) GP

Composite 3 Embankment

TB-44 (AP-2986), TB-46 (AP-2988), TB-49 (AP-

2991), TB-51 (AP-2993) GP-GM

Composite 4 Embankment

TB-20 (AP-2962-P), TB-23 (AP-2965), TB-25 (AP-

2967) GP-GM

Composite 5 Embankment

TB-08 (AP-2950), TB-11 (AP-2953), TB-13 (AP-

2955), TB-15 (AP-2957), TB-18 (AP-1960) GP-GM

Composite 6 Foundation

TB-08 (AP-2950), TB-11 (AP-2953), TB-13 (AP-

2955), TB-15 (AP-2957), TB-18 (AP-1960) SP

Composite 7 Foundation

TB-30 (AP-2972), TB-33 (AP-2975), TB-35 (AP-

2977), TB-37 (AP-2979), TB-39 (AP-2981), TB-42

(AP-2984) GW

February 2018 13 1781138

Composite Number

Soil Unit1 Borehole Origin Composite

USCS

Classification

Composite 8 Embankment

TB-30 (AP-2972), TB-33 (AP-2975), TB-35 (AP-

2977), TB-37 (AP-2979), TB-39 (AP-2981), TB-42

(AP-2984) GP-GM

Composite 9 Foundation

TB-20 (AP-2962-P), TB-23 (AP-2965), TB-25 (AP-

2967), TB-28 (AP-2970) GW

Notes: 1. Embankment refers to samples from 0 to 40 feet and Foundation refers to samples depths greater than 40 feet.

6.2.2 Soil-Cement Mix Summary

In the bench-scale testing, a total of 24 mixes were created during the mix design process. The mixes used a constant water to binder ratio and the volume ratio was varied by changing the binder factor to vary the strength of the mix.

6.2.3 Trial Mixes

Seven trial mixes were initially batched to provide general indications of the range of anticipated strength.

These mixes were prepared using readily available bentonite as described in Section 5.3. Table 6.2 summarizes the trial mixes and the material source.

Table 6.2: Summary of Trial Mixes Mix No.

Soil Unit Material Source Binder Factor (lbs/ft3)

Bentonite Type

1 Embankment Composite 1 31 Perma-Plug Bentonite Chips 2 Embankment Composite 1 16 Perma-Plug Bentonite Chips 3 Foundation Composite 2 31 Perma-Plug Bentonite Chips 4 Foundation Composite 2 16 Quik Grout Bentonite Powder 5 Foundation Composite 2 16 Perma-Plug Bentonite Chips 6 Embankment Composite 3 31 Perma-Plug Bentonite Chips 7 Embankment Composite 3 22 Perma-Plug Bentonite Chips

Notes: 1. Water to Binder Ratio = 0.6

2. Mixes 1 -6: Bentonite Factor = 1.8 lb/ft3. Mix 7: Bentonite Factor = 2.5 lb/ft3

3. Water to Bentonite Ratio = 10

6.2.4 Test Mixes

Based on the results of the seven trial mixes, 17 bench-scale test batches were prepared to evaluate the influence of the binder factor on strength and determine the impacts of any soil variability. Additionally, the influence of mixing water was evaluated using a consistent soil type and binder factor and water from the relief wells. Table 6.3 summarizes the test mixes, material source, binder factor and water source.

February 2018 14 1781138

Table 6.3: Summary of Test Mixes Mix No.

Soil Unit Material Source Binder Factor (lbs/ft3)

Water Source

8 Embankment Composite 5 16 Potable Water 9 Embankment Composite 5 10 Potable Water 10 Embankment Composite 5 20 Potable Water 11 Embankment Composite 6 31 Potable Water 12 Foundation Composite 6 20 Potable Water 13 Foundation Composite 6 40 Potable Water 14 Foundation Composite 7 25 Potable Water 15 Foundation Composite 7 25 Relief Well Water: STA 270+30 16 Foundation Composite 7 35 Potable Water 17 Foundation Composite 7 35 Relief Well Water: STA 290+25 18 Foundation Composite 8 10 Potable Water 19 Embankment Composite 8 13 Potable Water 20 Embankment Composite 8 18 Potable Water 21 Embankment Composite 4 25 Potable Water 22 Foundation Composite 9 10 Relief Well Water: STA 418+00 23 Foundation Composite 9 20 Relief Well Water: STA 330+35

24 Foundation Composite 9 30 Relief Well Water: STA 378+07 Notes: 1. Water to Binder Ratio = 0.6

2. Bentonite Factor = 1.8 lbs/ft3.

3. Water to Bentonite Ratio = 10

4. All mixes prepared with Barakade SP Bentonite

February 2018 15 1781138

7.0 MIX DESIGN RESULTS

7.1 Unconfined Compressive Strength (UCS)

The results of the UCS testing (ASTM D2166) for test Mixes 8 through 24 are presented in Appendix C and

D for the 7-day and 28-day breaks, respectively. A summary of the results is also presented in Table 7.1, and Graphs 7.1 and 7.2. The results indicate there is a linear relationship between the binder factor and the UCS strength within the range of testing. The 7-day average UCS strengths range from 124 psi to 1,107 psi for binder factors ranging from 10 to 40 lb/ft3, and the 28-day average UCS strengths range from 165 psi to 1,586 psi. In general, the results indicate the treated foundation soils may be slightly higher in compressive strength than similar treated embankment material. However, this variation is less pronounced at 28 days. In addition, the variation between the composite samples was small.

Table 7.1: Mix 8 through 24 Summary, 7- and 28-Day Average UCS Results

Mix ID Soil Unit Mix Water Binder Factor (lb/ft3)

Average UCS 7-Day (psi)

Average UCS 28-Day (psi)

8 Embankment Potable Water 16 245 372

9 Embankment Potable Water 10 124 165

10 Embankment Potable Water 20 416 616

11 Embankment Potable Water 31 649 1115

12 Foundation Potable Water 20 450 664

13 Foundation Potable Water 40 1107 1586

14 Foundation Potable Water 25 662 861

15 Foundation Relief Well Water: STA 270+30 25 710 934

16 Foundation Potable Water 35 1011 1334

17 Foundation Relief Well Water: STA 290+25 35 1067 1511

18 Foundation Potable Water 10 159 248

19 Embankment Potable Water 13 218 313

20 Embankment Potable Water 18 348 520

21 Embankment Potable Water 25 626 840

22 Foundation Relief Well Water: STA 418+00 10 171 258

23 Foundation Relief Well Water: STA 330+35 20 398 582

24 Foundation Relief Well Water: STA 378+07 30 936 1199

February 2018 16 1781138

Graph 7.1: Mix 8 through 24, 7-Day Average UCS Results

Graph 7.2: Mix 8 through 24, 28-Day Average UCS Results y = 27.899x - 151.72 R² = 0.9553 y = 34.597x - 197.8 R² = 0.9726

0 5 10 15 20 25 30 35 40 45

U n co n fi n ed C o m p re ss iv e St re n gt h p si

Binder Factor (lb/ft3)

Average 7 Day UCS Break - Embankment Soils Average 7 Day UCS Break - Foundation Soils y = 46.156x - 308.78 R² = 0.9965 y = 46.744x - 250.94 R² = 0.9817

0 5 10 15 20 25 30 35 40 45

U n co n fi n ed C o m p re ss iv e St re n gt h p si

Binder Factor (lb/ft3)

Average 28 Day UCS Break - Embankment Soils Average 28 Day USC Break - Foundation Soils

February 2018 17 1781138

7.2 Tensile Results

Tensile tests were run as requested in the USACE SOW in general accordance with ASTM D3967 and

C496. Generally, the tensile strength of barrier walls is particularly important to understand when evaluating seismic performance. The results of the splitting tensile testing at 28 days for Mixes 6 through 21 are presented in Appendix F, and summarized in Table 7.2 below. There is some scatter in the results, but the trends are consistent with UCS testing. The average splitting tensile strength ranges from 21 to 160 psi.

Table 7.2: Average Splitting Tensile Strength Summary – 28-Day Strength

Mix ID Soil Unit Mix Water Binder Factor (lb/ft3)

Average Splitting Tensile Strength (psi)

6 Embankment Potable Water 31 144

7 Embankment Potable Water 22 51

8 Embankment Potable Water 16 51

9 Embankment Potable Water 10 21

10 Embankment Potable Water 20 82

11 Embankment Potable Water 31 119

12 Foundation Potable Water 20 86

13 Foundation Potable Water 40 134

14 Foundation Potable Water 25 95

15 Foundation Relief Well Water:

STA 270+30

25 118

16 Foundation Potable Water 35 145

17 Foundation Relief Well Water:

STA 290+25

35 160

18 Foundation Potable Water 10 34

19 Embankment Potable Water 13 46

20 Embankment Potable Water 18 70

21 Embankment Potable Water 25 84

February 2018 18 1781138

Graph 7.3: Average Splitting Tensile Strength vs Binder Factor

7.3 Permeability

Permeability testing of select cast cylinders after at least 28 days of curing was performed in a large triaxial cell following ASTM D5084. Tested samples were back-saturated prior to permeability testing. The results of the permeability testing are presented in Appendix G. The permeability results range from 1.3x10-8 centimeters per second to 1.48x10-7 centimeters per second which is consistent with typical portland cement concrete.

Table 7.3: Summary of Permeability for Various Mixes

Mix ID Age at Testing

Soil Type Binder Factor (lb/ft3)

Permeability (cm/sec)

Mix Water

8 35 Embankment 16 3.05E-08 Potable Water

9 50 Embankment 10 1.17E-07 Potable Water

11 38 Embankment 31 3.29E-08 Potable Water

12 32 Foundation 20 6.36E-08 Potable Water y = 4.2944x - 14.028 R² = 0.9448 y = 3.8326x + 6.3916 R² = 0.8644

0 5 10 15 20 25 30 35 40 45

Sp lit ti n g Te n si le S tr en gt h p si

Binder Factor (lb/ft3)

Average 28 day Tensile Break - Embankment Soils Average 28 day Tensile Break - Foundation Soils

February 2018 19 1781138

Mix ID Age at Testing

Soil Type Binder Factor (lb/ft3)

Permeability (cm/sec)

Mix Water

13 39 Foundation 40 2.27E-08 Potable Water

14 61 Foundation 25 3.73E-08 Relief Well Water STA. 270+30

15 40 Foundation 25 3.65E-08 Potable Water

16 48 Foundation 35 4.42E-08 Relief Well Water STA. 290+25

17 62 Foundation 35 1.30E-08 Potable Water

18 56 Foundation 10 1.48E-07 Potable Water

7.4 Mixing Water

The results of the mix water testing are presented in Appendix H. Following the requirements for evaluating the use of non-potable water, as described in ASTM C1602, the relief well water may be used for the barrier wall with anticipated minor variations in the design strength. ASTM C1602 requires that the effects of non-potable water are limited to a 10 percent reduction of the UCS at 7 days, and a time of setting of one hour faster to one and one-half hours slower than the control made with potable water. The relief well water meets the following criteria outlined in ASTM C1602; the chemical limits for chloride, sulfides and metals, the time of setting. However, the 7-day UCS variation for one relief well exceeds the criteria in ASTM C1602.

The results of the mixing water characterization are summarized in Table 7.4. We performed a comparison of the UCS variation of potable water and relief well water by testing additional mixes with relief well water.

A summary of these results is presented in Appendix H. There appears to be some minor variation in strength between the potable and relief well water mixes which should be fully evaluated as part of the contractor’s mix design testing using the proposed water source prior to construction. However, as seen in

Appendix H, the variations are relatively minor.

Table 7.4: Summary of the Relief Well Time of Setting and UCS Variation

Water Source Time of Setting Variation1 (hr:min)

Mix Strength Deviation at 7 Days2

Relief Well 270+35 1:15 -7%

Relief Well 290+25 1:15 -6%

Relief Well 310+25 1:05

Relief Well 330+35 0:55 -12%

Relief Well 360+10 0:55

Relief Well 378+07 1:15 +11%

Relief Well 418+00 0:50 +8%

Relief Well 441+75 1:00

February 2018 20 1781138

Water Source Time of Setting Variation1 (hr:min)

Mix Strength Deviation at 7 Days2

Relief Well 478+00 0:55

ASTM C1602

Allowable Limits -1:00 / +1:30 -10%

Notes: 1) Variation in Time of Setting from samples prepared with relief well water compared to distilled water.

2) UCS Deviation between mixes prepared with relief well water and potable water at the same binder factor.

7.5 Design Strength Variation

There are additional sources of variation during construction that should be considered during design.

These variations should be evaluated with a contractor mix design program and test section as discussed in Section 7.6.

7.5.1 Laboratory and Construction UCS Variation

During bench-scale laboratory testing program, the actual mixing of the soil-cement may be more thorough than field mixing. This could lead to variation of UCS between field-mixed materials and laboratory-mixed samples of 20 to 100 percent based on results from EuroSoilStab and Coastal Development Institute of

Technology (FHWA 2013). In the United States, it is typically assumed that strength of field-mixed samples can consistently achieve 50 percent of the laboratory mixes (FHWA 2013). It is unknown how material type affects the strength variation. The actual difference is dependent on the contractor and their means and methods and should be evaluated. When developing the project specifications, it is important to develop a thorough field testing program with test panels to better estimate the strength relationship between field-mixed and laboratory-mixed samples.

7.5.2 Admixture

We understand that fly ash is not typically used in cement in Alaska due to the high cost, and fly ash has not been evaluated in our mix design. However, the influence of fly ash can change the compressive strength of the soil-cement and delays the hardening process in soil-cement (Egorova et al. 2017). Other contractor proposed admixtures should be evaluated in the contractor’s mix design testing.

7.6 Typical Pre-Construction Quality Control

There are several elements of quality control prior to construction for successful barrier wall construction including the following: contractor mix design and construction of test panels. The contractor should prepare the specific mix design with the proposed water source and any admixtures. FHWA (2013) has a thorough series of recommendations for developing the construction mix design. Additionally, a test panel program is important prior to the start of production in an effort to ensure that construction equipment and methods are appropriate to meet the design recommendations.

February 2018 21 1781138

7.7 Baseline for Contractor Bids

Table 7.5 summarizes the range of the various mix design parameters based on the tested minimum and maximum binder factors of 10 and 40 lb/ft3, respectively. Mix design proportions and example mix calculations are shown in Appendix J. A linear relationship between 28-day unconfined compressive strength and binder factor exists between the low strength and high strength mixes evaluated, based on the results presented in Graph 7.2.

Table 7.5: Various Mix Design Parameters for the Minimum and Maximum Test Mixes

Parameter Low Strength Mix

High Strength Mix

Binder Factor (α) 10 lb/ft3 40 lb/ft3

Water-to-Binder Ratio of the Slurry (w:b) 0.6 0.6

Total Water-to-Binder-Ratio (Wt:b)1 3.1 1.2

Volume Ratio (VR) 45 percent 89 percent percent Binder Content (αw) 8 percent 30 percent

28-day Average Unconfined Compressive Strength2 165 psi 1,586 psi

28-day Splitting Tensile Strength 21 psi 134 psi

Notes: 1) Based on an assumed dry unit weight of 125 lb/ft3 and moisture content of 5 percent for in-situ soils.

2) Based on a mix with potable water and foundation material.

February 2018 22 1781138

8.0 LIMITATIONS AND USE OF REPORT

This report has been prepared exclusively for the U.S. Army Corps of Engineers, Alaska District, for the proposed barrier wall project at Moose Creek Dam in North Pole, Alaska. If there are significant changes in the nature, design, or location of the facilities, we should be notified so that we may review the information within this report in light of the proposed changes and provide a written modification or verification of the changes.

Variations are likely in subsurface conditions between explorations and also with time. Therefore, inspection and testing by a qualified geotechnical engineer should be included during construction to provide corrective recommendations adapted to the conditions revealed during the work. A contingency for unanticipated conditions should be included in the construction budget and schedule in the event corrective measures are necessary based on conditions revealed in the excavations.

This work program followed the standard of care expected of professionals undertaking similar work in

Alaska under similar conditions. No warranty expressed or implied is made.

February 2018 24 1781138

10.0 REFERENCES

Egorova, A.A, Rybak J., and Stefaniuk, D., 2017. IOP Conf. Ser.: Mater. Sci. Eng. 245 022019

FHWA (Federal Highway Administration) (2013). “Design Manual: Deep Mixing for Embankment and Foundation Support” FHWA-HRT-13-046, FHWA, McLean, VA, October.

Golder Associates Inc., 2017a, Final Geotechnical Data Report, Weir Geotechnical Investigation for the Moose Creek Dam Safety Modification Study, Chena River Lakes Flood Control Project, North Pole, Alaska, Contract W911-KB-17-D-0002, Task Order No. W911KB17F0034, Golder Project Number 1781138, September 29, 2017.

Golder Associates Inc., 2017, Weir Geotechnical Investigation and Mix Design for the Moose Creek Dam Safety Modification Study, Chena River Lakes Flood Control Project, North Pole, Alaska, Work Plan, Contract W911-KB-17-D-0002, Task Order No. W911KB17F0034, Golder Project Number 1781138, July 7, 2017.

Koehler, R.D., Farrell, Rebecca-Ellen, Burns, P.A.C., and Combellick, R.A., 2012, Quaternary faults and folds in Alaska: A digital database, 31 p., 1 sheet, 1:3,700,000.

Page, R.A., Biswas, N.N., Lahr, J.C., and Pulpan, H., 1991, Seismicity of Continental Alaska, in Slemmons, D. B., Engdahl, E. R., Zoback, M. D., and Blackwell, D. D., eds., Neotectonics of North America:

Boulder, Colorado, Geological Society of America, Decade Map Volume 1.

Péwé, T.L., Wahrhaftig, C., and Weber, F.R., 1966, Geologic map of the Fairbanks Quadrangle, Alaska:

U.S. Geological Survey Miscellaneous Geologic Investigations Map 455, 5 p., 1 sheet, scale 1:250,000.

Ruppert, N.A., Ridgway, K.D., Freymueller, J.T., Cross, R.C., and Hansen, R.A., 2008, Active Tectonics of Interior Alaska: Seismicity, GPS Geodesy, and Local Geomorphology, in: Freymueller, J.T., Haeussler, P.J., Wesson, R.L., and Ekström, G., eds., Active Tectonics and Seismic Potential of Alaska: American Geophysical Union Geophysical Monograph Series 179, p. 109-134.

Ridgway, K.D., Thomas, E.E., Layer, P.W., Lesh, M.E., White, J.M., and Smith, S.V., 2007, Neogene Transpressional Foreland Basin Development on the North Side of the Central Alaska Range, Usibelli Group and Nenana Gravel, Tanana basin: Geological Society of America Special Papers 431, p. 507- 547, doi:10.1130/2007.2431(20).

Stover, C.W. and Coffman, J.L., 1993, Seismicity of the United States, 1568-1989 (revised): U.S. Geological Survey Professional Paper 1527, 418 p.

APPENDIX A

PHOTO LOG

February 2018 1 1781138

Mix Design for the Moose Creek Dam Safety Modification Study

PHOTO 1

Splitting bulk samples for lab testing and mix design.

PHOTO 2

Blending composite samples for bench-scale testing.

Photo Log

February 2018 2 1781138

PHOTO 3

Preparing a test mix.

PHOTO 4

Sulfur capping of a cylinder for UCS testing.

February 2018 3 1781138

PHOTO 5

Performing a UCS test with strain measurement.

PHOTO 6

Performing a splitting tensile test.

February 2018 4 1781138

PHOTO 7

Performing a permeability test.

APPENDIX B

SIEVE RESULTS

SUMMARY OF THE NINE COMPOSITE BLENDS

0.0010.010.1110100

20 406 60

U.S. SIEVE NUMBERS

3 4 20061/2 103 50 100301

3/81.5 8 140

GRAIN SIZE IN MILLIMETERS

SANDGRAVEL

SILT OR CLAY

coarse finefinecoarse medium

COBBLES

U.S. SIEVE OPENING IN INCHES HYDROMETER

P E

R C

E N

T F

IN

E

R B

Y W

E

IG

H T

3/4

ASTM D422

ASTM D6913

Composite Samples, 1 Composite Samples, 2 Composite Samples, 3 Composite Samples, 4 Composite Samples, 5 Composite Samples, 6 Composite Samples, 7 Composite Samples, 8 Composite Samples, 9

Sample Location, Number poorly graded gravel with silt and sand (GP-GM) poorly graded gravel with sand (GP) poorly graded gravel with silt and sand (GP-GM) poorly graded gravel with silt and sand (GP-GM) poorly graded gravel with silt and sand (GP-GM) poorly graded sand with gravel (SP) well-graded gravel with sand (GW) poorly graded gravel with silt and sand (GP-GM) well-graded gravel with sand (GW)

Reference(s)

SUMMARY OF PARTICLE SIZE DISTRIBUTION RESULTS

USCS ClassificationDepth (ft)

Project:

Client: USACE

Moose Creek Dam Barrier Wall Mix Design Project No.: 1781138

North Pole, AlaskaLocation:

Reviewed By: Date:A. Daggett 10/9/2017

M O

O S

E C

R E

E K

B ar rie r W

A

LL

B U

LK

S

A M

P

LE

S .G

P J

LI

B R

A R

Y -A

N C

(1 0-

-1

7) .G

LB

A N

C L

A B

G R

A

IN

S

IZ

E

7]

A D ag ge tt

2/

/1

COMPOSITE 1 SUMMARY

Client:

Project:

Location: Reviewed By:

TO

P

B O

TT

O M

G R A V

EL

SA

N D

FI

N

ES

Composite 1 46.6 41.7 11.7 GP‐GM

TB‐01 C‐2 0 7.5 41.2 44.0 14.9 SM

TB‐01…