Final Geotech Report Theodore Roosevelt National Park 221-230_07-01-22.pdf

Final Geotechnical Report

ND FTNP THRO 11(10)

South Unit Scenic Loop Slide Repair

Theodore Roosevelt National Park, North Dakota

NPS PMIS No: 291791 Drawing No.: 387-177128

Yeh Project No.: 221-230

July 1, 2022

Prepared for:

HDR Engineering, Inc.

1670 Broadway, Suite 3400

Denver, Colorado 80202 Attn: Mr. Nick Zoller, P.E.

Federal Highway Administration

Central Federal Lands Highway Division 12300 West Dakota Avenue Lakewood, Colorado 80228

Prepared by:

Yeh and Associates, Inc.

2000 Clay Street, Suite 200 Denver, Colorado 80211

Phone: 303-781-9590

Final Geotechnical Report

ND FTNP THRO 11(10)

South Unit Scenic Loop Slide Repair

Theodore Roosevelt National Park, North Dakota

NPS PMIS No: 291791 Drawing No.: 387-177128

Yeh Project No.: 221-230

July 1, 2022

Prepared by: Reviewed by:

Todd Schlittenhart, P.E. T. Samuel Holder, P.E.

Principal Engineer Senior Project Manger

Independent Reviewer:

Brett Arpin, P.E.

Project Manager

7/21/2022 i

Table of Contents

1 PURPOSE AND SCOPE OF STUDY

2 PREVIOUS INVESTIGATION

3 PROPOSED CONSTRUCTION

4 SITE CONDITIONS AND GEOLOGIC SETTING

4.1 SITE CONDITIONS

4.2 GEOLOGIC SETTING

5 SUBSURFACE INVESTIGATION

5.1 FIELD EXPLORATION

5.1.1 DRILLING

5.1.2 SEISMIC INVESTIGATION

5.1.3 GROUND PENETRATING RADAR

5.2 SUBSURFACE CONDITIONS

5.3 LABORATORY TESTING

5.4 GROUNDWATER

5.5 GEOLOGIC HAZARDS AND GEOTECHNICAL CONCERNS

5.5.1 DISPERSIVE SOILS

5.5.2 UNSTABLE SLOPES

5.5.3 EROSION

5.5.4 SEISMICITY

5.5.5 SULFATE ATTACK AND CORROSION

6 CONSTRUCTION RECOMMENDATIONS

6.1 DEMOLITION

6.2 SITE GRADING CONSIDERATIONS

6.3 FILL MATERIALS AND PLACEMENT

6.3.1 SHRINK/SWELL FACTORS

6.4 EXCAVATION AND TRENCH CONSTRUCTION

6.5 DRAINAGE CONSIDERATIONS

7 SLOPE STABILITY

7.1 GLOBAL STABILITY ANALYSIS BACKGROUND

7.2 EXISTING SLOPE STABILITY

7.3 SLOPE STABILIZATION

7.3.1 SITE F

7.3.2 SITE G1/G2

7.3.3 SITE C

ii

7.3.4 SITE B1/B.5

7.3.5 SITE A

7.3.6 SITE 242

7.3.7 SITE J

8 RETAINING WALLS

8.1 DRILLED SHAFT RECOMMENDATIONS

8.1.1 DRILLED SHAFT AXIAL RESISTANCE

8.2 LATERAL EARTH PRESSURES

8.3 LATERAL RESISTANCE

8.4 FUTURE CONSIDERATIONS

9 LIMITATIONS

10 REFERENCES

List of Tables

TABLE 1. SITE DESIGNATION

TABLE 2. SEISMIC DESIGN PARAMETERS

TABLE 3. SHRINK/SWELL FACTORS

TABLE 4. INPUT PARAMETERS FOR SLOPE STABILITY ANALYSIS

TABLE 5. LPILE PARAMETERS

TABLE 6. DENSITY AND CONSISTENCY OF SOILS BASED ON PENETRATION RESISTANCE VALUES (N-VALUES)

List of Figures

FIGURE 1. PROJECT LOCATION

FIGURE 2. EROSION AT STATION 283+00

FIGURE 3. CULVERT OUTLET NEAR STATION 283+00

FIGURE 4. SURFACE FEATURE WITH POTENTIAL FOR PONDING BEHIND ROADWAY CURB

FIGURE 5. CULVERT OUTLET NEAR STATION 219+00

List of Appendices

APPENDIX A – BORING LOCATION PLANS

APPENDIX B – GEOLOGIC MAPS

APPENDIX C – BORING LOGS

APPENDIX D – LABORATORY TEST RESULTS

APPENDIX E – COLLIER REPORT (DECEMBER 2021)

APPENDIX F – SLOPE STABILITY ANALYSIS

APPENDIX G – AS-BUILT DRAWINGS (SITE J)

Final Geotechnical Report YA Project No. 221-230 Theodore Roosevelt National Park (THRO) July 1, 2022

1 PURPOSE AND SCOPE OF STUDY

This report presents the results of our geotechnical investigation, analysis, and recommendations for the South Unit Scenic Loop Slide Repair project in Theodore Roosevelt

National Park (THRO) near Medora, North Dakota. Figure 1 shows the Park location in western

North Dakota. This study was performed in general accordance with our contract effective June

17, 2021, and associated contract modifications. Our work consisted of field exploration, laboratory testing, engineering analyses, and preparation of this report.

Figure 1. Project Location

This report includes our recommendations for the geotechnical aspects of project design and construction. The conclusions and recommendations stated in this report are based on the conditions found at the location of our exploratory borings at the time our investigation was performed. Our findings, conclusions, and recommendations should not be extrapolated to other areas or used for other projects without our prior review. Furthermore, they should not be used if the site has been altered without Yeh and Associates’ (Yeh) prior review to determine if these recommendations remain valid.

The purpose of this investigation was to evaluate the geotechnical characteristics of the subsurface soils for pavement and retaining wall design. Pavement design recommendations are provided under a separate cover. The scope of the study included the following tasks:

• A subsurface investigation for the roadway and proposed structures where accessible with track- and truck-mounted drill rigs.

• Geophysical subsurface investigation using seismic and 3D ground penetrating radar

(GPR) methods.

• Laboratory testing of the soils encountered during field exploration to evaluate relevant physical and engineering properties of the soil.

• Preparation of this report and the Pavement Design Report.

2 PREVIOUS INVESTIGATION

Shannon and Wilson (S&W) completed a field exploration program that included drilling exploratory borings in February and May 2020 focused on Sites A and C, with additional borings drilled throughout the corridor including at Sites B, C, and F to provide additional data. Site designations are discussed in Section 3. The purpose of the study was to assess landslide hazards at specific locations along South Unit Scenic Loop Road. The results of their field investigation and analysis are presented in the following referenced reports:

- Unstable Slope Assessment Report, October 2021

- Geotechnical Design Report, October 2021

S&W’s field investigation included drilling nine exploratory borings between Sites F, C, B.5/B1, and A and nine pavement borings throughout the project. S&W installed inclinometers in seven of the borings and groundwater monitoring devices at nine borings at the landslide sites.

3 PROPOSED CONSTRUCTION

We understand the South Unit Scenic Loop Slide Repair project includes reconstruction and minor widening, resurfacing, restoration, and rehabilitation of an approximate 6-mile stretch of the existing South Unit Scenic Loop Road. The desired pavement treatment is to perform full-depth reclamation of the existing pavement and aggregate base, overlayed with an asphalt wearing surface, without raising the road profile. In addition to the pavement reconstruction, the project will also include the construction of seven (7) retaining walls along the roadway to mitigate existing and potential future movement of the roadway due to unstable slopes. The retaining walls are planned as soldier pile and lagging walls with ground anchors. The project will also include changes such as signing and striping, drainage improvements, and additional safety-related features. The retaining walls are planned at the sites as designated in Table 1.

Table 1. Site Designation

Site Designation Wall ID Project Stationing Site F MP-26.17 153+15 to 154+90

Site G1/G2 MP-25.86 168+50 to 173+26 Site C MP-24.87 220+98 to 225+00

Site B.5/B1 MP-24.64 231+50 to 238+95 Site A MP-24.38 247+00 to 249+75

Site 242 MP-24.24 255+30 to 257+56 Site J MP-23.01 319+88 to 322+13

4 SITE CONDITIONS AND GEOLOGIC SETTING

4.1 Site Conditions

The South Unit Scenic Loop Road is located northeast of Medora, North Dakota within

Theodore Roosevelt National Park (the Park). The Scenic Loop Road is a two-lane roadway that goes through badland terrain within the park. A Yeh representative visited the project site in June and September 2021 to perform field reconnaissance of the 6-mile stretch of roadway to evaluate areas of interest prior to implementing our field exploration program.

Two sites along the roadway, Sites A and C, have experienced significant roadway damage due to landslide movement, forcing closure of the roadway to visitors. Our observations at these sites and the other sites described in Table 1 included evidence of slope movement including tension cracks, settlement, toe-bulges near the bottom of the slope, and erosion of dispersive clay soils creating tunnels commonly known as sinkholes.

Refer to the referenced S&W reports (2021) for additional discussion of the roadway and landslide history. We investigated the site geology by reviewing published geologic maps and reports and by the analysis of our test borings.

According to the High Plains Regional Climate Center data for the monthly temperature and precipitation normal from 1948-2010 for THRO, the average annual maximum and minimum temperatures were 57° and 30° Fahrenheit, respectively, and the average annual precipitation was 14.97 inches. Data from the Central Federal Lands Highway Division (CFLHD)

Geotechnical Memorandum (March 2012) indicates the average annual precipitation from 2002 to 2011 in North Dakota was 19.15 inches.

4.2 Geologic Setting

Most of the geology along the South Loop Road is mapped (Bierk, 2001) as the Paleocene age

Sentinel Butte Formation (Tsb), Sentinel Butte Formation HT Butte clinker (Tsbc), and Bullion

Creek Formation (Tbc). A portion of the geologic map within the project limits is presented in

Appendix B.

The Bullion Creek Formation (Tbc) is mapped along much of the western portion of the South

Loop Road. The bright yellowish-brown siltstone, claystone, mudstone, and sandstone of this formation were deposited in river, lake, and swamp conditions. They are poorly lithified and can be a combined thickness of 600 feet. Thin carbonaceous shales and thick lignite beds are common.

Much of the central and eastern lengths of the South Unit Scenic Loop Road are mapped as

Sentinel Butte Formation (Tsb). This formation consists of siltstone, mudstone, claystone, sandstone, and lignite. These beds were deposited in river, lake, and swamp conditions, are generally gray to brown, and together can be greater than 750 feet thick. Some of the sandstones are well cemented and these layers form resistant ledges. Most of the other layers are generally weakly cemented. Sandstone and mudstone concretions, siderite nodules, and petrified wood are common. Several marker beds of the Sentinel Butte Formation are present in the South Unit of the Park, including the HT Butte clinker (mapped separately, described below), the basal sandstone, a petrified wood horizon, and yellow beds.

The HT Butte clinker (Tsbc) and HT Butte lignite forms the basal layer of the Sentinel Butte

Formation and marks the contact with the underlying Bullion Creek Formation. This clinker formed from the burning of the lignite. The clinker encompasses a variety of rock types and varies in color from pink to black but is most commonly red. The clinker texture varies from that in which the original structure is preserved to resembling glassy slag, and in some instances, it resembles volcanic scoria. As the fires burn, hot gases sometimes escaped to the surface through “chimneys” in the overlying layers. The sediments within these chimneys can be highly brecciated and altered. In other cases, the overlying sediments slumped into voids caused by the burning lignite. Clinker is harder than the original unaltered sediments and therefore can form erosion resistant ridges and caprocks. Other unnamed clinker beds of the Sentinel Butte

Formation (Tsbck) are found southeast of the South Unit Scenic Loop Road.

One segment of roadway is mapped as Landslide deposits (Qls) of Holocene and Pleistocene age with “small mass movements” (curved arrow symbol on map) marked nearby. Landslides are “widespread” and range from a variety of small failures such as slumps, debris flows, and rock falls to square-mile sized landslide complexes. Landslides are distinguished by hummocky topography, arcuate scarps, and disturbed bedding. Beds of the Sentinel Butte Formation form steep, barren slopes and therefore numerous landslides are found in this formation. Small landslides and creep are common, the most prominent of which are shown on the map with a symbol instead of outlined as a map unit. Creep is common on north facing slopes of the

Bullion Creek Formation.

Two sections of the western portion of the South Unit Scenic Loop Road are mapped as Modern alluvial deposits (Qal) of Holocene age. These alluvial deposits generally consist of sand, silt and clay in river channels and floodplains. The deposits are generally within 35 feet or less above current stream channel elevations but may be up to 50 feet above. These deposits are less than 500 years old. Older alluvial deposits (Qoal) of Holocene age are mapped near, but not along, the investigated section of roadway. These deposits are similar material to Qal but are older than 500 years of age.

Two sections of the eastern portion of the South Unit Scenic Loop Road are mapped as Mantled

Pediments (Qmp) of Holocene age. These are very gently sloping bedrock surfaces at the base of steep, rapidly eroded slopes. Though pediments are erosional features, their cover or

“mantle” material is distinguished as a geological unit because it consists of “complex assemblages” in this region, including slopewash and reworked eolian material. Some mantled pediments are deeply incised and others are zones of active alluvial processes. Mantle thickness generally ranges from less than three feet to six feet but can be greater than six feet thick.

Other mapped geological units in the area include several alluvial units. Alluvium beneath

Pleistocene terraces 2, 3, and 4 (Qt2, Qt3, Qt4) generally consist of sand and gravel. They are most commonly exposed on plateau margins and poorly vegetated, south facing hillslopes.

Each of these four mapped units are differentiated by their elevation above the modern channel of the Little Missouri River, with Qt4 being the highest and oldest. Holocene engineered fill

(Qef) is mapped in areas disturbed by construction activities, including along Interstate 94.

5 SUBSURFACE INVESTIGATION

5.1 Field Exploration

5.1.1 Drilling

Yeh’s subcontractor, SK Geotechnical Corporation (SK Geotech) drilled thirty-four (34) borings;

twenty-five (25) were drilled with a truck-mounted Diedrich D-120 drill rig and nine (9) were drilled with a track-mounted Mobile B-57 drill rig, both equipped with an automatic (i.e.

hydraulic) hammer as noted on the boring logs. Sixteen (16) of the borings were focused on pavement investigation (YA-P-01 through YA-P-10 and YA-PG-01 through YA-PG-06). The rest of the borings explored subsurface conditions in areas of slope movement to aid in understanding subsurface conditions and design of mitigation methods. The borings were advanced with 3¼-inch inside-diameter (I.D.) hollow-stem augers between August 30 and

September 15, 2021. The subsurface conditions encountered in the borings were logged by representatives of Yeh. The general locations of borings were identified by roadway failure locations, areas of potential roadway movement or concern, and for pavement investigation with the understanding that ground penetrating radar (GPR) would also be performed to aid in pavement design. Yeh’s boring locations supplemented borings previously drilled by S&W. At the time of our investigation the roadway was closed to the general public. We drilled all borings on the pavement, with the exception of YA-C-01 which was drilled off the roadway to compare to S&W’s borings and in a location where the roadway alignment may be moved. The latitude and longitude of the boring locations were recorded in the field using a mobile phone app providing accuracy to within about 12 feet and later surveyed by Chad Sexton with the

Park.

Penetration resistance measurements were taken by driving Modified California (MC) samplers into subsurface materials in general accordance with ASTM D3550, “Standard Practice for Thick

Wall, Ring-Lined, Split Barrel, Drive Sampling of Soils.” Measurements were obtained at approximate continuous to 5-feet intervals. The MC sampler is a 2.5-inch outer-diameter (O.D.), 2.0-inch inner diameter (I.D.) split-barrel sampler containing 1.95-inch inner diameter brass liners. The MC sampler was driven 12 inches for each sample. The sampler was advanced by an automatic-hammer simulating the energy of 140-pounds falling 30 inches. The automatic hammers used to drive the MC sampler on the Diedrich D-120 and Mobile B-57 drill rigs had calibrated energy transfer ratios of 77 percent and 87 percent, respectively.

The number of blows required to drive the MC sampler 12 inches, or fractions thereof, constitutes the penetration resistance value (N) as shown on the boring logs. The penetration resistance value is a useful index to evaluate the consistency and relative density or hardness of the subsurface materials encountered.

Relative undisturbed soil samples were obtained using Shelby tube samplers in general accordance with ASTM D1587, “Standard Practice for Thin Walled Tube Geotechnical Sampling of Soils.”

5.1.2 Seismic Investigation

Collier Geophysics (Collier) performed seismic investigations at nine (9) locations along the roadway with the objective of evaluating information related to the slides’ shear planes of movement and aid in subsurface modeling for slope movement. The seismic investigation was performed at the sites presented in Table 1. The seismic data collected were analyzed using seismic refraction tomography (SRT) for evaluation of P-wave velocity (Vp) and one-dimensional multichannel analysis of surface waves (MASW) for evaluation of bulk S-wave velocity (Vs) at each site. The seismic investigations were performed between September 29 th and October 1st, 2021. Locations of the seismic lines used to record data, interpretation of the data at each site, and additional information are included Collier’s Geophysical Report presented in Appendix E.

5.1.3 Ground Penetrating Radar

Bridge Diagnostics, Inc. (BDI) subcontracted with Collier to perform 3D GPR along the roadway and paved areas of the South Unit Scenic Loop Road and assessed for pavement thickness and possible areas of voids/loss of support beneath the roadway. The GPR data was collected on September 21st, 2021, by BDI and overseen by a Collier geophysicist. The full width of the roadway as well as parking/pullout areas were included in the investigation. The BDI report is contained in an appendix to the Collier report and presented in Appendix E. Results of the GPR investigation are discussed in more detail in our Pavement Design Report.

5.2 Subsurface Conditions

Subsurface conditions encountered in our exploratory borings generally consisted of fill derived of native materials over native clay with occasional sand, underlain by the Sentinel Butte

Formation.

The existing fill consisted of sandy lean and fat clay with some clayey sand and clinker (locally called scoria) with some gravel and ranged from nil to 10 feet thick. Existing fill was encountered in 30 of our borings. Field penetration data indicated the clay fill was soft to very stiff in consistency and the clayey sand fill was loose to medium dense in relative density.

The native soils below the existing fill generally consisted of sandy lean clay and fat clay and was about nil to 22 feet thick. Field penetration data indicated the consistency of the clays was very soft to very stiff. The clay contained carbonaceous layers at times.

The native soils were underlain by the Sentinel Butte Formation consisting largely of claystone, with occasional layers of sandstone and minor layers of siltstone. Sentinel Butte Formation was encountered in 28 borings at depths of 1.2 to 27 feet. Most of the claystone classified as a fat clay. The Sentinel Butte Formation ranged in stiffness from residuum to very hard, but the majority of it was relatively soft compared to harder, more competent bedrock formations. The

Sentinel Butte Formation contained many carbonaceous layers and some relatively thick layers of lignite.

Overall, most of the subsurface materials encountered in our borings were clay-like materials

(native lean/fat clay and claystone). The sand was less prevalent. The sandier material (native sands and sandstone) was generally encountered near borings YA-G-07, YA-G-08, YA-P-07, and YA-P-08.

The surveyed boring locations are presented in Appendix A at the end of this report. The investigation boring logs are presented in Appendix C.

5.3 Laboratory Testing

Selected soil and bedrock samples were tested in the laboratory to evaluate the classification and engineering properties of the materials encountered in the borings. The testing was conducted in general accordance with recognized test procedures, primarily those of the

American Society for Testing and Materials (ASTM) and American Association of State Highway and Transportation Officials (AASHTO). The following tests were performed in general accordance with locally recognized standards:

Description and Identification of Soils (Visual-Manual Procedure) Moisture Determination Unit Weight

Atterberg limits One Dimensional Swell-Collapse Unconfined Compressive Strength Water-Soluble Sulfates

Water-Soluble Chlorides Resistivity pH Consolidated-Undrained (CU) Triaxial Test Direct Shear Test Double-Hydrometer Test Pinhole Dispersion Test

We performed sieve analysis and Atterberg limit tests on seventy-five (75) samples obtained from the borings. Forty-six (46) samples had an AASHTO soil classification of A-7-6 with group indexes of 3 to 127, seventeen (17) classified as A-6 with group indices of 3 to 21, seven (7) as

A-4 with group indices of 0 to 6, one (1) classified as A-2-6 (0), two (2) classified as A-2-4 (0), one (1) classified as A-1-b (0), and one (1) classified as A-1-a (0). Thirty-one (31) samples classified as CH, twenty-seven (27) as CL, three (3) as ML, eight (8) as SC, four (4) as SM, one

(1) as SW-SM, and one (1) as GW in the ASTM system.

We performed sieve analysis (without Atterberg limit tests) on an additional eight (8) samples of possible road base material, which contained 11 to 45 percent fines (passing No. 200 sieve).

One-dimensional swell or collapse potential of cohesive soils, laboratory test ASTM D4546, was performed on two (2) samples. The samples swelled 0.8 and 3.5 percent when wetted under

1,000 psf applied pressure.

Two (2) unconfined Compressive Strength Tests (ASTM D2166) were conducted on samples collected using a MC sampler. The unconfined compressive strengths were 1,651 and 3,725 psf.

Knight Piesold conducted consolidated-undrained (CU) triaxial tests (ASTM D4767), double hydrometer tests (ASTM D4221), pinhole dispersion test (ASTM D4647), and direct shear tests

(ASTM D3080). Knight Piesold performed CU triaxial testing on two Shelby tube samples from borings YA-A-02 and YA-B5-01. The double hydrometer and pinhole tests were performed to confirm the results of the Crumb Tests (generally following ASTM D6572-21 Method A) indicating that dispersive soils are prevalent at the site. The direct shear tests were performed to provide additional peak and residual soil strength parameters for slope stability analyses. All test results obtained from Knight Piesold are presented in Appendix D.

The material descriptions and laboratory test results were used for the geotechnical engineering analyses and the development of structure recommendations.

Details of laboratory test results are presented in Appendix D, and they are integrated with the boring logs in Appendix C.

5.4 Groundwater

We did not encounter groundwater during drilling. Our observations represent groundwater conditions at the time of field exploration and may not be indicative of conditions at other locations or times. Groundwater levels may fluctuate with varying seasonal and weather conditions. S&W encountered groundwater during drilling and in measurements taken from vibrating-wire piezometers (VWP). The groundwater levels they observed are presented in their

2021 reports.

5.5 Geologic Hazards and Geotechnical Concerns

Geologic hazards and geotechnical concerns encountered on this site include dispersive soils, unstable slopes, erosion, seismicity, and sulfate attack and corrosion.

5.5.1 Dispersive Soils

Based on our review of published geologic maps and information provided by CFLHD, dispersive soils are present in this region. These soils are highly erodible and vulnerable to sheet flow, piping, and gully erosion. Dispersive soils occur when the soil structure collapses as individual clay particles disperse into solution when the soil is wetted. Dispersion of the clay particles in the soil accelerates erosion and can lead to the formation of voids and tunnels

(“sinkholes”). We performed field Crumb Tests (generally following ASTM D6572-21 Method A) on soil collected in the upper 15 feet of each of our borings to evaluate the presence of dispersive clays along the roadway. The Crumb Test results indicated the majority of the clay soils we encountered are dispersive. Knight Piesold conducted pinhole and double hydrometer tests to confirm the presence of dispersive clays as discussed in Section 5.3. Proper design and construction of surface and subsurface drainage is critical to quickly remove water from the pavement, proposed retaining walls, and slopes supporting the roadway to reduce the risk of significant movement due to loss of support if the dispersive clays are wetted (SCS 1991).

5.5.2 Unstable Slopes

Based on our field reconnaissance observations, review of site grading plans, and LiDAR difference models, we believe fill slopes steeper than 1V:2H (vertical:horizontal) may be unstable in the project area. Retaining walls or re-grading of fill slopes to 1V:2H or flatter are feasible mitigation options at locations adjacent to the roadway where over-steepened fill slopes are present. Site specific evaluations of fill slopes 1V:2H or steeper are required to determine the appropriate mitigation option to achieve the project goals.

5.5.3 Erosion

Clays normally have a flocculated or aggregated structure because of the electrochemical attraction of the particles to each other and to water. Calcium and magnesium cations are predominate in the double layer in these clays. The balance of electrochemical forces in these soils favors a strong attraction of the particles to one another. This explains these soils’ typical cohesive and non-erosive behavior.

Dispersive clays, however, are predominated by sodium cations in their pore water. Because sodium cations only have one positive charge compared to two charges for calcium and magnesium ions, an imbalance in the electrochemical forces occur in dispersive clays. This imbalance causes particles in a dispersive clay to be repulsed rather than attracted to one another. Consequently, dispersive clay particles tend to react as single grained particles and not as an aggregated mass of particles.

Because clay particles are very small and have a low mass, the particles are easily detached and transported by water. This explains the extremely low resistance to erosion of these dispersive soils. Dispersive clays are most easily eroded by water with a low ion concentration, such as rainwater (SCS 1991).

Slopes along the roadway range from gentle to very steep, and some have high erosion potential. We observed evidence of existing culverts with outlets high on the slopes that produced significant erosion near the roadway. The presence of dispersive clays increases the risk of significant erosion and “sinkholes” that can lead to roadway and slope instability and failure. Figures 2 and 3 show an example of this issue near Station 283+00. Figure 2 shows a riprap apron constructed below the culvert outlet and significant erosion of the slope downstream of the riprap. Figure 3 shows an apparent head scarp, just above the dead vegetation and downstream of the culvert outlet, indicating the erosion progressed upslope from the riprap apron in close proximity to the road.

Figure 2. Erosion at Station 283+00

Figure 3. Culvert outlet near Station 283+00

CULVERT OUTLET HEAD SCARP

RIPRAP APRON

RIPRAP APRON

EROSION

5.5.4 Seismicity

No active faults are known to exist in the immediate vicinity of the project site. Based on site class definitions provided in Table 3.10.3.1-1 of AASHTO (2020), the site is categorized as Site

Class D. Based on the recommendations in Table 3.10.6-1 of AASHTO (2020), the site classifies as Seismic Zone 1.

The peak ground acceleration (PGA) and the short- and long-period spectral acceleration coefficients (SS and S1, respectively) for reference Site Class B were determined using the seismic design maps from the USGS website for an event with a 7 percent probability of exceedance in 75 years. An event with the above probability of exceedance has a return period of about 1,000 years. Since the site is designated at Site Class D, site-specific adjustments are necessary. The seismic design parameters for Site Class D are shown in Table 2. These values may be used to construct the Design Response Spectrum for use in the seismic design of the retaining wall structures.

Table 2. Seismic Design Parameters

Site Class PGA (0.0 sec) SS (0.2 sec) S1 (1.0 sec) B 0.019 0.046 0.017

Site Class AS (0.0 sec) SDS (0.2 sec) SD1 (1.0 sec) D 0.03 0.073 0.04

5.5.5 Sulfate Attack and Corrosion

The concentration of water-soluble sulfates measured in the laboratory on twenty-four (24) representative samples of clayey soils ranged from 0.005 to 0.685 percent, with thirteen (13) samples with a concentration higher than 0.2 percent. These concentrations of water-soluble sulfates represent a Class S3 (severe) sulfate exposure based on ACI 318, Building Code

Requirements for Structural Concrete and Commentary requirements. The degree of attack is based on a range of Class S0 (not applicable) to Class S3 (very severe). Based on the sulfate test results, ACI recommends a water to cement ratio of 0.45 and use of ASTM C150 Type V cement or equivalent.

The pH, water-soluble chlorides, and electrical resistivity were measured for fourteen (14) samples. Test results measured pH values ranging from 8.1 to 9.5. Water-soluble chlorides were measured between 0.0002 and 0.0011 percent. The soil resistivity measurements were

158 to 371 ohm-centimeters. AASHTO (2020) recommends that soil resistivity of less than

2,000 ohm-cm should be considered as indicative of potential pile deterioration or corrosive situations. Considering the soldier piles will be in contact with the on-site soils, we recommend painting the entire pile with a corrosion resistant material after fabrication or including sacrificial steel to mitigate the effects of pile deterioration and corrosion. Based on the Post-Tensioning

Institute (PTI) Recommendations for Prestressed Rock and Soil Anchors (2014), ground anchors will require Class 1 corrosion protection. A qualified corrosion engineer should review this data to determine the appropriate level of corrosion protection for other materials in contact with the on-site soils.

6 CONSTRUCTION RECOMMENDATIONS

The following sections present recommendations for site preparation, excavation, subgrade preparation, fill placement, and surface drainage on the project. Earthwork should be observed and evaluated by a qualified representative of CFLHD. The evaluation of earthwork should include observation and testing of engineered fills, subgrade preparation, foundation bearing soils, and other geotechnical conditions exposed during the construction of the project.

6.1 Demolition

Backfill materials surrounding culverts and culvert headwalls structures that are planned for replacement should be removed and replaced with moisture-conditioned, compacted fill.

Sections of the reinforced soil slope (RSS) section at Site J that conflict with the proposed retaining wall should be removed prior to construction of the wall and pavement section. As-built drawings of the RSS (Project ND FTNP THRO 11(9)) are included in Appendix G for reference. Clean portions of the fill can be reworked and used as new, moisture-conditioned, compacted fill. We do not recommend reuse of the clinker found at Site J for backfill below or behind new structures.

6.2 Site Grading Considerations

Strip and remove existing vegetation/organics, debris, and other deleterious materials from the proposed structure and pavement areas and any other areas to receive fill. All exposed surfaces should be free of mounds and depressions, which could prevent uniform compaction.

Stripped materials consisting of vegetation and organic materials should be wasted from the site or used to revegetate landscaped areas after completion of grading operations. Once properly cleared, all exposed areas which will receive fill should be scarified to a minimum depth of 6 inches, moisture-conditioned, and compacted in accordance with Section 204 (Excavation and

Embankment) of FHWA FP-14.

Based upon the subsurface conditions encountered, subgrade soils exposed during construction are anticipated to be relatively stable. However, the stability of the subgrade may be affected by precipitation, repetitive construction traffic, and other factors. If unstable conditions are encountered or develop during construction, stability may be improved by scarifying and drying the subgrade soils. Over-excavation of wet zones and replacement with granular materials or crushed rock may be necessary. Use of fly ash, kiln dust, cement, or geotextiles could also be considered as stabilization techniques. Laboratory evaluation is recommended to determine the effect of chemical stabilization on subgrade soils prior to construction.

We recommend slopes between the pavement and top of retaining walls be graded to 1V:3H, or shallower where possible. Slopes steeper than 1V:2H may be unstable in this project location and should be flattened to 1V:2H or flatter where possible. Site specific evaluations of designed slopes 1V:2H or steeper are required to determine the appropriate mitigation option to achieve the project goals.

6.3 Fill Materials and Placement

Engineered fill should be placed as specified in FHWA FP-14 Section 204 (Excavation and

Embankment) or Section 209 (Structure Excavation and Backfill), whichever is appropriate.

Import material should consist of select backfill as specified in a Special Contract Requirement

(SCR). Import fill material should be reviewed by the geotechnical engineer prior to importing to the site. Our study did not identify material sources or disposal sites within the Park. Imported fill will have to be sourced from outside the Park, and excess material will have to be wasted outside of the Park.

Culverts may be placed below pavements. The compaction of backfill in trenches can significantly affect the serviceability and lifetime of pavements, flatwork, and structures above culverts. We recommend trench backfill be placed as discussed above for structures and should be observed and tested by a qualified representative of CFLHD during construction.

Backfilling culverts with lean concrete will improve the performance of pavements and flatwork by reducing the potential for settlement in the pipe trench and reducing risk of erosion due to dispersive soils. FHWA FP-14 Section 209.09 discusses backfill options for pipe culverts.

6.3.1 Shrink/Swell Factors

Estimated shrink/swell factors of the materials encountered within the project limits are provided in Table 3. Shrink/swell factor estimations are based on laboratory unit weight measurements on samples obtained during the subsurface investigation, published data, and previous experience. The soils encountered during excavation may vary significantly across the site and may vary from soils encountered in the subsurface investigation. Variances may also occur from topsoil stripping, clearing and grubbing, fill compaction requirements, and construction practices. The estimated factors provided only account for differences in density between potential cut locations and fill slope embankments and do not account for waste or stockpiling.

Table 3. Shrink/Swell Factors

Material Cut Slope (In-Situ)

Density (pcf) Fill Embankment

Density (pcf) % Shrink or Swell1

Clayey Sand Fill 105 115 -10 Sandy Clay Fill 100 110 -10

Native Sandy Clay 95 110 -16 Native Clay 100 110 -10

Sandy Claystone (Sentinel Butte) 115 115 0 Claystone (Sentinel Butte) 100 110 -10 Siltstone (Sentinel Butte) 110 110 0

Sandstone (Sentinel Butte) 100 115 -15 1 A negative number represents shrinkage

6.4 Excavation and Trench Construction

Excavations into the on-site soils will encounter a variety of conditions. All excavations must comply with the applicable local, State, and Federal safety regulations, and particularly with the excavation standards of the Occupational Safety and Health Administration (OSHA).

Construction site safety, including excavation safety, is the sole responsibility of the contractor as part of its overall responsibility for the means, methods, and sequencing of construction operations. Yeh’s recommendations for excavation support are provided for the Client’s sole use in planning the project. In no way do they relieve the contractor of its responsibility to construct, support, and maintain safe slopes. Under no circumstances should the following recommendations be interpreted to mean that Yeh is assuming responsibility for either construction site safety or the contractor’s activities.

We believe the clay and Sentinel Butte Formation material encountered on this site will classify as Type B material and the sand and gravel will classify as Type C material using OSHA criteria. OSHA requires that unsupported cuts be no steeper than 1V:1H for Type B material and 1V:1½H for Type C for unbraced temporary excavations in dry conditions. Flatter slopes will be required below groundwater or if seepage is present. A Registered Professional

Engineer should design any excavations greater than 20 feet. Please note that the contractor’s

OSHA-qualified “competent person” must make the actual determination of soil type and allowable sloping in the field.

The soils encountered by the proposed excavations may vary significantly across the site. The preliminary classifications presented above are based solely on the materials encountered in widely spaced exploratory test borings. The contractor should verify that similar conditions exist throughout the proposed area of excavation.

As a safety measure, we recommend that all vehicles and stockpiles be kept to a lateral distance equal to at least the depth of the excavation from the crest of the slope. The exposed slope face should be protected against the elements and monitored by the contractor on at least a daily basis.

Trench boxes or other means of temporary shoring may be needed during installation of pipe culverts.

6.5 Drainage Considerations

Moisture conditions in the subgrade soils influence the performance of pavements, foundations, and other improvements. The risk of wetting these subgrade soils can be reduced by proper planning and maintenance of surface grades and drainage. Excessive wetting of the subsurface before, during, and after construction can cause movement of pavements and foundations. The risk is even greater with the presence of dispersive soils at this site and proper drainage will be critical to the performance of the pavements and retaining walls. We recommend the following measures be implemented during construction and maintained after construction.

• Positive drainage away from the structures should be provided. Proper design of drainage should include preventing the ponding of water on or immediately adjacent to structures and pavements. Surface features that could retain water in areas adjacent to the roadway or structures should be sealed, backfilled, and graded to drain, or eliminated, such as areas similar to that shown in Figure 4.

• Backfill against structures, culvert trenches, and curb and gutter should be moisture-conditioned, well-compacted, and free of all construction debris to reduce the possibility of moisture infiltration and migration. Infiltration can be reduced by placing seals between curbs and pavement.

• Concentrated runoff should be avoided in areas susceptible to erosion and slope instability. Slopes and other stripped areas should be protected against erosion by re-vegetation or other methods. We recommend culverts outlet well below the roadway, at the toe of the slope if possible. There is high risk of erosion that can propagate up to the roadway if culverts outlet on the upper or middle portions of slopes, as demonstrated near Station 283+00 as discussed in Section 5.5.3. Riprap is effective at slowing the velocity of water transported below a culvert outlet such that the velocity does not cause erosion in non-dispersive soils, however it does not prevent dispersive soil erosion. Our observations made during site reconnaissance indicate that there is higher risk of erosion at culverts that outlet on the upper or middle portions of a slope and less risk further downslope where the grade flattens out. Figure 5 shows a culvert near Station

219+00 that outlets near the bottom of the slope leading down from the roadway.

Relatively minimal erosion has occurred at the outlet where there is a relatively flat slope. Further downslope from the outlet, where the grading becomes steeper, a

“sinkhole” has been formed.

• Precautions should be taken to avoid wetting or drying of open excavations for foundations and other earthwork.

Figure 4. Surface feature with potential for ponding behind roadway curb

Figure 5. Culvert outlet near Station 219+00

7 SLOPE STABILITY

We utilized the results of our site observations and field reconnaissance, subsurface exploration, and laboratory testing to evaluate the slopes at the locations presented in Table 1.

We performed slope stability analysis on the existing slopes to understand possible causes of the existing recently active landslides and identify other areas of potential slope instability. We then coordinated the proposed retaining wall layouts and grading with HDR to evaluate slope stability with the mitigation measures in place.

7.1 Global Stability Analysis Background

In order to analyze the global stability at each site, we first defined the slope geometry and the strength of the materials that make up the different strata within the subsurface profile. We

SINKHOLE EROSION

FEATURE

CULVERT OUTLET

used survey information (existing and proposed grades) provided by HDR and subsurface data from Yeh’s exploratory borings and seismic investigation and S&W’s exploratory borings and inclinometer data to create representative cross-section models at each site. We selected cross-sections based on the most critical location(s) along a slope.

The global stability at each site was then analyzed using the limit equilibrium method of slices.

The slope stability analysis for each cross-section was completed using the Slide2 software by

Rocscience. The limit equilibrium method of slices divides the slope profile into many vertical

“slices”, calculates the driving and resisting forces for each slice, sums the forces, and calculates a factor of safety as the ratio of the sum of resisting forces to the sum of driving forces. We used three limit-equilibrium methods of slices to evaluate slope stability based on the models we created: Simplified Bishop, Spencer, and Morgenstern-Price.

We use the term “factor-of-safety” (FS) in the following discussions to describe the ratio of the strength available to resist sliding compared to the forces causing sliding within a slope. Thus, a FS of 1.0 can be interpreted as resisting forces equal to driving forces, or a slope that exists at equilibrium and is marginally stable. A FS less than 1.0 indicates resisting forces less than driving forces, or a slope below the equilibrium limit (the slope is failing). A FS more than 1.0 indicates total resisting forces are higher than total driving forces (the slope is stable).

The AASHTO LRFD Bridge Design Specifications (9th Edition, 2020) Section 11.6.3.7 indicates the overall (global) stability of earth slopes with or without a foundation unit should be investigated using a resistance factor of:

• 0.75 where the geotechnical parameters and subsurface stratigraphy are well defined

• 0.65 where the geotechnical parameters and subsurface stratigraphy are highly variable or based on limited information

For overall (global) stability, resistance factors are inverted to get a factor of safety corresponding to that calculated by limit equilibrium; thus 0.75 is equal to a FS of 1.3 and 0.65 is equal to a FS of 1.5. We recommend a FS of at least 1.3 for remediation of landslides.

The slope stability analyses were performed for long-term (drained) conditions representing normal post-construction conditions. We used results from CU triaxial, double hydrometer, direct shear, unconfined compression, and soil classification tests in addition to published data and previous experience to establish soil strength parameters for the soil types at each site.

The peak shear strength of soils was used for undisturbed soils; however, in areas of slope instabilities, the residual shear strength was used for materials disturbed within the slope. We assigned values we believe are conservative and input them into the slope stability evaluation program, Slide2. Cross-sections for sites with failing, or moving, slopes were modeled to back-calculate a FS of approximately 1.0 to verify the geotechnical strength parameters used in each model representing the existing marginal stability of the slope. Once the models were calibrated, we updated the models to include the retaining walls and evaluate the global slope stability of the stabilized slopes. Detailed results of our slope stability analyses are presented in

Appendix F.

7.2 Existing Slope Stability

We modeled the existing conditions at each site to evaluate the slope stability of the existing slopes. Strength parameters based on the results from our field investigation and lab testing were checked against back-calculated strength parameters for slope with a FS of approximately

1.0, which represents a marginally stable slope condition. The weak layer algorithm of Slide2 allowed modeling of the slip surface at sites with failing, or moving, slopes using strength properties supported by lab testing and inclinometer data. Appendix F contains the output files from Slide2 for our stability analysis. Input data for each soil type are summarized in Table 4 below.

Table 4. Input Parameters for Slope Stability Analysis

Soil Description Unit Weight (pcf) Cohesion (psf) Friction Angle (degrees) Clay 110 50-100 20-27

Sandy Clay 110 50 24 Weathered Claystone 110 0-50 11-18 Displaced Claystone 110 50 24

Claystone 110 100 24-29 Sandstone 110 20 30

Select Backfill 135 0 34

7.3 Slope Stabilization

We used the calculated strength parameters from the existing slope conditions to evaluate the feasibility of several mitigation methods by including them in our analysis. We considered the use of toe-berms, deep pavement patches, drilled shaft (shear pile) walls, and soldier pile and lagging walls with ground anchors. In conversations with CFLHD and the Park, the project team decided to use soldier pile and lagging walls with ground anchors based on project conditions such as access, availability of material, constructability, cost, and previous successful use in other National Park Service (NPS) projects. Additionally, if additional slope movement occurs below the walls after construction is complete, this mitigation method allows for the addition of lagging and another row of ground anchors to be placed in the future.

We modeled soldier pile and lagging walls with ground anchors at each wall site in Slide2. We used the same cross-sections we used to evaluate the existing slopes and included proposed grading and wall locations agreed upon with HDR.

As we performed our analysis, we encountered areas where the slope below the retaining walls still had a FS near or below 1.0, indicating potential instability. With concurrence from the design team personnel representing CFLHD and NPS, a minimum FS of 1.1 for the slopes below the retaining walls was agreed to be adequate. We targeted a minimum FS of 1.3 for the global stability of the slopes with the retaining walls in place.

We were able to achieve a minimum FS of 1.3 at each of the seven retaining wall sites using the soldier pile and lagging walls with ground anchors for mitigation of the movement at each site.

The table in Appendix F summarizes the slope stability results at each wall location and includes details such as the size, strength, angle, and spacing of ground anchors and soldier piles and assumptions made for the slope grading. Details pertinent to each site are discussed in the following sections.

7.3.1 Site F

Site F is located at MP 26.17 near Scoria Point along the South Unit Scenic Loop Road. We observed tension cracks parallel to the roadway centerline during site reconnaissance indicating the slope is moving. There have been several cold patch repairs attempted in the last several years, and a “sinkhole” near the location was discovered in March 2020 by S&W staff. The slope below the roadway is high and steep and shows evidence of creep (S&W 2021).

We evaluated the existing slope conditions at Station 154+25 and calculated a FS of 0.87 using the Morgenstern-Price analysis method for the existing slope. To increase the FS to at least

1.3, we modeled the proposed geometry with the retaining wall 15 feet from the edge-of-pavement (EOP) with a 1V:3H grade from the guardrail to the top-of-wall (TOW) and a 2 percent grade from EOP to the guardrail. We assumed the guardrail is 4 feet off the EOP. Soldier pile and ground anchor specifications are presented in Appendix F. The mitigation methods increased the global FS of the slope to 1.31, with a FS of 1.11 for the slope below the wall.

7.3.2 Site G1/G2

Site G1/G2 is located near MP 25.86. We observed tension cracks parallel to the roadway centerline during site reconnaissance, as well as steep slopes below the roadway (greater than

1V:2H). We evaluated this site at two cross-section locations, Stations 168+46 (Site G1) and

173+00 (Site G2).

The FS of the existing slope at Site G1 was 1.01 using the Morgenstern-Price analysis method.

To increase the FS to at least 1.3, we modeled the proposed geometry with the retaining wall 10 feet off EOP with a 2 percent grade from the EOP to TOW.