Appendix G - GeoTechnical Report - Amendment 7.pdf

Thule Air Base FY12 Dormitory Permafrost and

Geotechnical Investigation

ERDC-CRREL LR 12-01

August 31, 2011

ERDC-CRREL Letter Report LR 12-01

Approved for public release; unlimited distribution

Thule Air Base FY12 Dormitory Permafrost and Geotechnical Investigation

Kevin Bjella, MSc., P.E.

Cold Regions Research and Engineering Laboratory - CRREL

Engineer Research and Development Center – ERDC

U.S. Army Corps of Engineers - USACE

Building 4070, Fort Wainwright, Alaska 99703

August 31, 2011

Prepared for:

New York District

Army Corps of Engineers

US Army Corps of Engineers ®

US Army Cold Regions

Preface

This report and field work were completed by Kevin Bjella, Research Civil Engineer, U.S. Army Cold

Regions Research and Engineering Laboratory (CRREL), and performed under MIPR # W16ROE-1069-

3123.

The author gratefully acknowledges the assistance of Dr. Edel Cortez (CRREL) for his review of this report, and also Marc Beede (CRREL) for his valuable field assistance. The author would also like to thank Greenland Contractors personnel for their support during the field work of this project.

For more information contact:

Kevin Bjella

907-361-5171 kevin.bjella@us.army.mil

This report reflects the personal views of the author and does not suggest or reflect the policy, practices, programs, or doctrine of the U.S. Army or the Government of the United States. The contents of this report are not to be used for advertising or promotional purposes. Citation of brand names does not constitute an official endorsement or approval of the use of such commercial products.

i

TABLE OF CONTENTS

1. INTRODUCTION

1.1 Scope and Purpose

1.2 Project Description

2. SITE CONDITIONS

2.1 Physical Setting

2.2 Site Description

2.3 Climate

2.4 Soils

2.5 Geology

2.6 Glaciology

2.7 Permafrost and Ground Ice

2.7.1. Matrix and Segregated Ice

2.7.2. Wedge Ice

2.7.3. Relict Ice

2.7.4. Active Layer

3. FIELD EXPLORATION

3.1 Borehole Drilling

3.2 Ground Penetrating Radar (GPR)

3.3 Laboratory Testing

3.4 Subsurface conditions

3.4.1. Drilling data

3.4.2. GPR data

3.4.3. Synthesis of Subsurface conditions

4. CONCLUSIONS AND RECOMMENDATIONS

4.1 General

4.2 Foundation recommendation

4.2.1. Alternative I

4.2.2. Alternative II

4.2.3. Alternative III

4.3 Bearing Capacity and long term strength

4.4 Earth Work

4.5 Long Term Temperature Monitoring

ii

APPENDIX A

APPENDIX B

APPENDIX C

BIBLIOGRAPHY

1. INTRODUCTION

1.1 Scope and Purpose

This report presents the results of the permafrost and geotechnical investigation of the proposed location for the FY12 Dormitory, at Thule Air Base (TAB), Greenland. This work was performed under

Military Interdepartmental Purchase Request (MIPR) W16ROE-1069-3123.

The scope of this work was to investigate the subsurface soil and rock conditions at the proposed footprint of the new dormitory, present those findings, and make recommendations for the foundation design. Ground penetrating radar (GPR) was performed in April 2011, and subsurface drilling was performed in July and August 2011.

1.2 Project Description

The US Army Corps of Engineers, New York District (CENAN) is tasked with the construction of a

3-story, 72-person dormitory at the Air Base. The proposed building is composed of a steel super-structure with insulated metal panel cladding, and the footprint of the structure is 15.8-m x 91.5-m. The

30% Submission Design Analysis suggests insulated concrete spread footings and support columns for the foundation design. The anticipated life span of the structure is 50 years.

2. SITE CONDITIONS

2.1 Physical Setting

The Air Base (Pituffik) is situated in the NW trending valley of Pitugfiup kugssua (North River).

This 12 km long valley is ice free, and drains the Greenland Ice Cap on the east, to North Star Bugt on the west. North Mountain bounds the lower reach of this valley on the north, with hilly terrain northward to

Wolstenholme Fjord. South Mountain bounds the lower reach of this valley on the south, with a relatively flat plateau southward to Parker Snow Bugt. The region is void of vertical vegetation and is restricted to low herbs, grasses, dwarf shrubs, mosses and lichens. Generally the surface in the valley surrounding the

Air Base has been disturbed by construction activities, and few undisturbed areas exist where surface soils and vegetation can be seen as they existed prior to the Air Base construction in 1951. As the valley rises in elevation toward the inland ice cap, the vegetation gives way to a very rocky, boulder strewn landscape.

2.2 Site Description

The project site is bound by Amundson Ave. to the west, 1st Street to the north, Crescent Road to the east, and 2nd Street to the south. The site is currently occupied by Building 113 with numerous above-ground utility connections. These include insulated piping and electrical cables, which limited the access to most of the site to drilling and GPR. The site is on a high bench of the North River, with a gentle slope towards the northwest. The North River is approximately 250-m to the north. Building 113 sits on top of a level pad constructed of local fill material, averaging 1.5-m in thickness.

2.3 Climate

The climate of Pituffik Valley is High Arctic marine modified by the proximity to the inland ice cap.

High temperatures can exceed 10°C, with low temperatures approaching -40°C. The mean annual air temperature (MAAT) is 12.0°F with occasional hurricane force winds. The average air thawing degree days (ATDD) for 1953 to 2009 is 423°C, and the air freezing degree days (AFDD) for the same period is

4447°C. The average length of the thawing season is 125 days from mid May to mid September. Due to the recorded amount of climate change in the last three to four decades in high latitude regions, and because the future amount of climate change is uncertain, the ATDD was increased by a factor of two, resulting in a approximate 2.0°C increase in MAAT. The basis for design utilizing ATDD of 846°C

(MAAT of 14.2°C) was performed for thaw depth calculations.

2.4 Soils

The soil is predominately glacio-fluvial in origin, deposited during glacial occupation and retreat.

Therefore the soils range from silt size, to cobble and boulders but are predominantly gravel size. The depth dimension, lateral dimension, particle size distribution, and ice content can be very localized due to this depositional process. A particular depositional interval may not be traceable to within meters.

However, in proximity to the North River, alluvial deposition has reworked the glacio-fluvial sediments.

The alluvial reworking can provide greater continuity, where a particular interval may be traceable for tens of meters if not longer, and this appears to be the case at this project site.. The project site appears to be located at a high bench of the North River, therefore alluvial reworking of the sediments is expected.

The temperature of the soil measured at a CRREL site near the airfield with barren fill for surface material during the dates of drilling are shown in Table 1. Based on the results of four CRREL temperature measuring sites at TAB, the temperatures at the proposed site are not expected to deviate from those at the site near the airfield.

Table 1. Subsurface temperatures July 29th, 2011.

Depth-m Temp °C 1 3.4 2 -0.5 3 -3.1 4 -4.1 5 -5.0 6 -5.7 7 -6.3

2.5 Geology

The glacially sculpted geology of the region is dominated by an alternating sequence of sedimentary rocks, termed the Thule Supergroup. More specifically, the Upper Thule Supergroup is sediments comprised of the Narssarssuk Group and the Dundas group (Dawes, 2006). These siliciclastic redbeds and pale carbonaceous sediments are visible in the glacially eroded slopes of North and South Mountain, Dundas Fjeld, and Saunders Ø. These lithified sediments are essentially undeformed, however half-grabben structures are numerous, displaying vertical displacements of tens to hundreds of meters. A regional swarm of basic dykes and sills are noted in this group, with a prominent exposure of Diorite having been cut for the TAB runway. This strata generally dips to the south-southwest at shallow angles, with the strike trending west-northwest, consistent with the trend of the North River Valley. This dipping strata is easily imaged with ground penetrating radar, and near the North River the bedrock surface, which is buried under glacial and alluvial sediments, can be seen dipping away from the river towards the south.

Prominent exposures of the mudstone (slate) sequences can be found in the valley and leading up to, and at the BMEWS site. The layers are approximately 2mm in thickness, dark colored, and smooth.

Weathering provides ready detachment at the bedding boundary, and exposures are easily disturbed by hand and foot traffic. It is expected that bedrock is relatively shallow at the project site. Borehole drilling into the weathered shale bedrock exposes clay size material embedded with angular shale fragments < 1.0 cm, and this will grade to pure clay with very low moisture (ice content).

2.6 Glaciology

The glacial history of the region described by Funder (1990) suggests that during the late

Pleistocene, multiple glacial advances covered the Pituffik Valley. The greatest extent of glaciation and the earliest recorded in the existing sediments, the Agpat Glaciation, extended westward beyond Saunders

Ø. The extent of the following inter-glacial is undetermined, but the ice is hypothesized to have retreated to Wolstenholm Fjord. The following glacial advance, the Narssarssuk Glaciation, filled Pituffik Valley and extended southwestward terminating on the plateau to the south of the valley, with the following inter-glacial hypothesized to have retreated to Wolstenholm Fjord. The next and last glacial advance, in the early Holocene, failed to progress from Wolstenholm Fjord, and no further advances in the valley are noted. Due to the multiple glacial events, a large amount of outwash sediment overlie the sedimentary bedrock within Pituffik Valley. The glacial sediment thickness varies from 6-m to 18-m, with progressively thickening depth moving towards the existing ice cap margin, and thinning depth with grading to beds of sand, when approaching the beaches of North Star Bugt. Glacial tills of larger cobble and erratic are evident in the valley. Therefore the air base is situated on a varying thickness of glacial sediments, with an average of 3-m to 10-m at the airfield.

During the glaciations, alluvial sediments have been deposited along the channel of the North River.

This deposition has re-worked the outwash and till into more stratified layers of silts, sands and gravels.

These would be located up to the highest benches of the North River, to include the proposed location of the FY12 Dormitory.

2.7 Permafrost and Ground Ice

A periglacial environment has existed in the High Arctic, through the Pleistocene and Holocene, and continues today, including the Pituffik Valley. During glacial retreat the sediments and rock are exposed to low temperatures, allowing for the establishment of permanently frozen ground (permafrost) and associated ground ice features. In the High Arctic such as Pituffik, the permafrost is continuous in lateral extent, and is estimated to be more than 300m in depth. Currently the permafrost temperature, at the depth of zero annual amplitude, is approximately -10°C (Bjella, 2010). This permafrost was created epigenetically, where the rock and sediments were placed in full thickness, then subsequently permanently frozen. The post-glacial continual reworking of the alluvial sediments of the North River allows for the creation of syngenetic permafrost on the margins of the channels during flooding events.

Ice exists in the permafrost and is the primary concern for the engineering of foundations for structures. If allowed to thaw, the greater the amount of ice present, the lower the shear strength of the soil will be upon thawing. This lower shear strength results in differential thaw settlement, and must be accounted for in design.

2.7.1. Matrix and Segregated Ice

The freezing process during creation of the permafrost can create excess ice in the soil if the proper conditions exist. From a conservative standpoint, it is generally assumed that soils are saturated when they become permanently frozen. Therefore the 10% volume expansion that occurs to fresh water upon freezing will cause individual soil particles to be moved apart within the matrix ice. Fine grain soils contain more water (ice) at saturation than do coarser grain soils. Also, if fine grain soils exist along with an additional supply of water, the matrix ice can continue to be created into what is termed segregated ice due to the wicking action of the fine grain soils, creating near horizontal layers or lenses of ice millimeters to centimeters in thickness, and centimeters to meters in length. During the sediment deposition, either glacial or fluvial, the amount of water available was variable. Therefore it is not uncommon to find very high ice content soils directly above, below, or adjacent to soils with lower ice content.

2.7.2. Wedge Ice

Cold temperatures in the winter contract the surface soils and allow cracks to develop that are meters deep, tens of meters in length, and millimeters in width. Surface water in the spring and summer migrates into the cracks and freezes, creating a thin vein of ice. This location of ice intrusion is a zone of weakness, and the process will repeat at the same location for hundreds and thousands of years with the resultant ice vein becoming an ice wedge. These wedges can be meters in depth, one to two meters in width, and meters to tens of meters in length. Most often the wedges will interconnect into polygon shapes that are meters in diameter, and these shapes can be seen at the surface because of the mechanical distortion that occurs along with the ice intrusion. The Pituffik Valley has a high occurrence of wedge ice and the probability is very high for its existence at the project site.

2.7.3. Relict Ice

During glacial retreat, the margin of the ice sheet is a very dynamic with glacial till being exposed and deposited, large volumes of water flowing off the ice sheet carrying outwash, and ice fracturing and being deposited on top and in loose piles of sediment. This fractured ice can be deposited and buried with sufficient expediency to prevent the ice from melting prior to permafrost creation. Therefore chunks of ice, termed relict ice (Corte 1962), can exist in the permafrost and in a very random manner. The North

River channel however, moves such large volumes of water and sediment, that the probability of relict ice surviving in the benches that are now exposed near the river, and at the proposed FY12 Dormitory location, is relatively low.

2.7.4. Active Layer

The active layer is the near surface soils that undergo the annual process of thawing in the summer, and complete refreezing during the winter, and the bottom of the active layer is the top of the permafrost table. The depth of annual thaw is dependent on vegetative cover, soil type, soil moisture, and solar aspect. For Pituffik Valley the active layer ranges from 30cm depth in vegetative areas under thick organic cover, to 1.5-m to 2.0-m in non-vegetated areas. It is very common to find a thick sequence of segregated ice at the base of the active layer. This is due to the permafrost acting as an aquitard, and the collected water generally exceeds saturation of the soil.

3. FIELD EXPLORATION

3.1 Borehole Drilling

A total of eight boreholes (Fig. 1) were completed at the proposed site from July 29th and July 30th, 2011, and all were logged by the author. The drill rig utilized was a Geoprobe 7822DT, combined rotary and direct push technology (DPT) machine. This combined ability was specifically chosen to allow for direct ‘undisturbed’ sampling of the fine grained frozen soils and ice, while having the ability to quickly convert to air rotary or down-the-hole hammer, to allow penetration of ice-cemented gravels and larger cobble and boulders. Borehole logs are presented in Appendix A.

The MC5 Macro Corer DPT rod and barrel system utilizes a 5.7 cm diameter, 1.5-m sample barrel that allows for sample capture into a clear acrylic sample tube. If conditions are favorable, 1.5-m of recovery is possible with minimal thawing of the samples due to friction. DPT is a misnomer as the barrel is not pushed into the soil, but rather it is hammered into the soil at very high frequency. After retrieval the plastic tube is cut lengthwise into halves to allow for sampling and logging. With above freezing ambient air temperatures the sample will begin to thaw, allowing for a visual approximation of ice-rich

vs. ice-poor intervals. It was possible to obtain nearly full recovery in three full barrels (4.5-m) from the surface down in five of the six holes, and nearly four full barrels (6-m) in three of the holes. Occasionally refusal was encountered during DPT sampling, and changing to air rotary allowed for penetration of the resistive zone, then DPT was resumed to obtain continuing samples.

When it was determined that DPT would no longer be effective for that particular hole, the system was changed to air rotary and the drilling resistance was monitored during hole advancement to gain additional information. The air rotary system consisted of NWJ rod (7 cm diameter) with either tri-cone roller bit or down-the-hole hammer engagement. The compressed air was supplied by a Thule Air Base provided Atlas Copco XAHS-1146 screw compressor, with 540 l/s flow at 20 bar. BH-6 was drilled specifically to ascertain depth to bedrock, and only two DPT sample barrels were obtained (3-m) from the surface down.

Figure 1. Borehole locations around Bldg. 113.

3.2 Ground Penetrating Radar (GPR)

Ground penetrating radar surveys were conducted at the proposed site April 23rd and April 28th, 2011. A Geophysical Survey System (GSSI) SIR-3000 GPR was utilized with 200 MHz, and 400 MHz antennas. The 200 MHz antenna provided the best resolution vs. depth of imaging, and all the images illustrated in this report are of that frequency unless noted. The radargrams are presented in Appendix B.

Two grids were laid out in two areas adjacent to the existing Building 113. The first grid was 46-m x

20-m, and was positioned on the Northeast side of the building, between the utilidor running parallel to the building, and 1st Street. This grid was limited in location directly behind the building due to a water conveyance utilidor running perpendicular to Building 113, towards 1st street (Fig. 2). Drifted snow was greater than 1m in depth across portions of this grid, and required removal prior to the survey. The western corner of the grid was created as the origin (0,0) and the eastern corner adjacent to 1st street was created as (46-m,20-m). Along the Y-axis, -1m intervals were located; and along the X-axis, 2-m intervals were located.

BH-5

BH-4

BH-3

BH-6

BH-2

BH-1

* BH-4 Also includes BH-4A and BH-4B N

The second grid was located along the southwest wall of the building in the designated parking area.

Berms of plowed snow and utlidors limited the ability to create a larger grid. No snow removal was performed in this area. This grid was 50-m x 6-m, with (0,0) located at the western corner of the grid, and

(50m, 6m) located at the eastern corner of the grid (Fig. 2, 11). Along the Y-axis, 1-m intervals were located; and along the X-axis, 2-m intervals were located.

Both grids were surveyed in the same manner starting at (0,0) and surveying along the X-axis to the southern corner. The next line was advance 1m along the Y-axis and surveyed in the opposite direction, and this was repeated across the grid. The survey was then run orthogonal to the previous, starting at (0,0) and moving along the Y-axis. For both grids the Y-axis lines were run at 10-m intervals and were all run in the same direction. Selected radargrams, Figures 4 thru 11, are shown in Appendix B.

Figure 2. Grid layouts for GPR surveys at Bldg 113.

Grid 1

Grid 2

N

3.3 Laboratory Testing

During logging, samples were obtained from specific intervals of interest. The samples were field logged (Appendix A), bagged and sealed. Later that day they were weighed and placed in a sample oven at 110°C (ASTM D2216), which was available at Thule. Moisture content was then determined after sample drying by dividing the weight of the moisture, by the weight of the dry sample, and is expressed as a percentage. A few bulk samples were returned to our laboratory in Fairbanks for particle size distribution (ASTM D422) and frozen unit weight. However, due to the required delivery date of this report, the samples had not arrived in time for the analysis to be conducted. It is recommended that grain size analysis be obtained from Baker (2004) for FY05 Dormitory, as it is not anticipated that soil conditions differ dramatically between the two sites.

Due to the difficult logistics from Thule AFB, it was decided that the shipping of intact frozen samples to our laboratory could not be counted upon, therefore no dry unit weights were measured. For a geotechnical investigation at the FY05 Dormitory site 60m from this project site (Baker, 2004), one dry unit weight was measured at 1034 kg/m^2 with a moisture content of 31%. Another investigation conducted near the airfield (Shannon and Wilson, 1983) measured eight dry unit weights with an average of 1872 kg/m^2 and an average moisture of 16%. Salinity of the soils was not tested as Baker (2004) revealed that although Thule Air Base is in close proximity to the sea, there are not elevated salinity levels at this location between the North River and the airfield. No strength testing was conducted for the bedrock samples.

3.4 Subsurface conditions

3.4.1. Drilling data

A total of eight subsurface borings were completed during this investigation (Fig. 1). The boring logs are located in Appendix A, and selected photos of the cores are shown in Figures 12 thru 24 in

Appendix C. All borings except BH-5 were along a line at the back of Building 113. This line projects from the southeast end of the site, parallel to the existing building, and to the northwest. This line approximately falls along the centerline of the proposed dormitory. It did not appear that a substantial amount of fill had been placed along this line, and the drill logs and GPR confirmed this. The surface elevation gradually decreases moving from southeast to northwest with BH-1 approximately 1.5-m higher than BH-4. Building 113 is located on man-placed fill, and BH-5 at the front of the building is approximately 2.5-m higher than BH-4.

Five of the borings were designated to give detailed information from the surface down to refusal, and the remaining three were used as confirmation. Soil moisture (ice content) was visually examined and logged for each 1.5-m Macro core recovered, and 20 specific samples were measured analytically from key intervals within the ‘detailed’ boreholes. The gravimetric moistures for 9 of the 20 samples were greater than 50%, and these samples were all taken from between just below the active layer at 1.5-m to

1.8-m, and down to 5.5-m to 6.0-m. The stratigraphy encountered in this high ice content zone is highly variable, and the individual specific layers are not easily correlated to adjacent boreholes. It was expected that these sediment horizons would be heterogeneous within relatively short distances and smaller scales

(tens of centimeters), specifically with regard to gravel size, sand size, and ice content. Generally speaking however, the soils within this horizon are silty gravels with some sand constituent, they frequently contain segregated clear ice at varying depths, and generally they are gray indicating a lack of oxidation (reduced environment). In BH-4 foliated white ice (wedge ice) was encountered just below the active layer. Subsequent confirmation boreholes BH-4A and 4B were drilled either side of BH-4 and it was found the ice wedge has a minimum width of 1.0-m and minimum vertical length of 2.5-m. The moisture contents for the entire depth of BH-3 indicate ice-poor permafrost at that location. This is an anomaly in comparison to the other boreholes, and is good evidence for the heterogeneity of the permafrost conditions.

Below the 5.5-m to 6.0-m depth, it was found that drilling resistance increased dramatically, with either DPT or air rotary. With this increase in drilling resistance, was a corresponding drop in moisture content. Analytical samples were taken at or near the bottom-of-the-hole in BH-1, BH-3, and BH-4, and all were < 15%. In addition, visual logging indicated very ice-poor sediments in those boreholes, and also at that depth in BH-2 as well. The increase in resistance for the DPT type drilling completely stopped advancement, generally in the same manner that encountering a cobble or boulder would stop advancement. BH-6 was drilled as a confirmation hole to penetrate beyond the DPT refusal depth, and ascertain the thickness and material being encountered. In BH-6, DPT refusal occurred at 3.0-m, most probably due to a cobble. Tricone air rotary was then initiated, and good advancement was noted to approximately the depth of previous DPT refusals in previous holes, at 5.5-m to 6.0-m. During this interval the process was intermittently very bouncy with consistently wet returns. This indicates firmly cemented, high ice content sediments, with transitions from large or abundant gravels and sands, to small or less abundant gravels and sands. This corresponds to the previously drilled holes and the high ice content horizon mentioned above. When the advancement became less then 1.0cm/min, the drill system was changed to down-the-hole hammer. This process utilizes a pneumatic hammer driven with high air flow, embedded with carbide buttons to engage the soil/rock. Initially the advancement was excellent, but after 1.5-m the resistance increased to the point that the rate dropped to < 2.5cm/min. The air return was strong with almost no wet cuttings and sporadic dust or vapor, indicating good communication and possibly drilling of dry bedrock material.

3.4.2. GPR data

The GPR surveys conducted at Building 113 were performed in April while the active layer was completely frozen to insure no attenuation from liquid water would affect the signal. In all the surveys, two distinct soil horizons are visible, separated by a reflector(s) with a high dielectric contrast (Appendix

B, Figs. 4 thru 11). The first horizon occurs from the surface down to approximately 5.0-m. In Grid 1, this horizon is not strikingly sharp, with subtle individual layers, most probably due to the very high ice content throughout the horizon. However, in Grid 2, this same horizon displays distinct layering probably due to the fact that man placed fill was hauled in to create the level pad for Bldg. 113. The bottom of the active layer is discernable in most images as a faint continuous horizontal reflector at 2.0-m to 2.5-m..

Correlating the Grid 1 images to the drilling, it can be seen that this horizon is the ice-rich glacio-fluvial/alluvial sediments mentioned previously from the drilling data. The wavy and discontinuous nature of the imaged stratigraphy explains the fact that individual gravel layers were difficult to correlate between boreholes.

At approximately the 5.0-m depth, sharply contrasting, horizontal to slightly dipping, nearly continuous reflector(s) are displayed in all the images. These reflectors have black/white/black banding, or negative/positive/negative, which can indicate that a higher dielectric constant material is overlying a lower dielectric constant material. As was noted in the drilling results, the deeper subsurface material generally becomes less moist (ice-poor) below the 5.0-m to 6.0-m depth. Dielectric constants for ice-rich permafrost overlying ice-poor, or dry soil/rock, would be consistent with this polarity signature. Sharply imaged stratigraphy is visible below this reflector for approximately 1.0-m to 1.5-m, beyond that depth the reflections are muddled, which can indicate that the limit of signal return has been achieved, or the material has become homogeneous, such as with thinly bedded intact sedimentary rock. Shale and

Dolostone outcrops have been imaged with GPR by the author elsewhere in the Pituffik Valley and similar results occur yielding a bright reflector at the bedrock surface with muddled to non-existent reflections from below. This bedrock reflector appears to become deeper when moving from the northwest to the southeast, which is consistent with the S-SW dipping sedimentary structure of the valley.

3.4.3. Synthesis of Subsurface conditions

The boreholes indicate that below the bottom of the active layer at 1.5-m to 1.8-m, and to a depth of

5.0-m to 6.0-m, exists heterogeneous, ice-rich, silty gravels with sands. This alluvial/glacio-fluvial horizon not only contains very high ice content (excess ice) sediments, but also contains segregated ice up to tens of centimeters in thickness and length, and also contains wedge ice, meters in depth and width, and tens of meters in length. The GPR indicates this sediment horizon extends at approximately this depth and thickness across the entire site. These sediments are extremely thaw-unstable due to the high ice contents.

These ice-rich sediments overlay what likely is weathered bedrock, more specifically a shale sequence of the Upper Thule Supergroup. It is known from previous borings conducted by the author, that the contact between the sediment horizon, and the bedrock is not discrete. There generally exists ice-poor sandy gravel to gravelly sands that overlie the contact with the weathered bedrock. Additionally, mechanical weathering of the shale produces a mixture of angular shale pieces with increasing amounts of blue/gray clay size material with depth, before obtaining competent bedrock. The thickness of the weathered contact is approximately 1.5-m to 2.0-m, and this corresponds to the GPR reflections mentioned earlier, that are visible below the 5.0-m to 6.0-m bright reflector. The ice content of the sediment/bedrock weathered contact continually decreases to trace levels when approaching the competent bedrock. Therefore, at some depth within this weathered contact the soil/rock is thaw-stable.

At the location Building 113 currently occupies, there is approximately 1.5-m of fill material overlying the heterogeneous ice-rich horizon. This fill layer is level on the surface, but the lower surface dips to the northwest with the slope of the hillside.

4. CONCLUSIONS AND RECOMMENDATIONS

4.1 General

It is understood that the 30% design submission proposes a foundation design similar to the FY05

Dormitory at Thule AFB. This also is similar to the foundation design utilized for the new medical facility. This design incorporates the use of individual sub-grade precast concrete footings, supporting insulated columns that elevate the three story structure above grade. These post-and-pad type foundations have a few advantages when placed on ice-rich permafrost. Firstly, the post type of support allows for elevation of the structure above grade, providing air circulation between the bottom of the first floor and the subgrade, and this significantly decreases or eliminates the amount of heat transfer from the building to the foundation soils. Secondly, this requires that the superstructure which transfers the load from the building to the posts, be designed to withstand differential settlement in a more robust manner than with a continuous footing type foundation. In the event of minor thaw settlement, it also provides access for re-leveling. We concur with this choice of foundation. The subgrade soils below the active layer and down to 5.0-m to 6.0-m are extremely ice-rich. A conventionally constructed at-grade heated structure, with any type of foundation, would transfer an adverse amount of heat energy into the soil, causing considerable differential thaw-settlement.

4.2 Foundation recommendation

As a general note, it is recommended that the active layer soils be completely removed and not re-used as structural fill material. This will equate to a minimum 1.5-m to 1.8-m thickness across the entire site and should extend beyond the perimeter of the building by 2.5Lf, where Lf is the largest horizontal dimension of the footer. These active layer soils are non-consolidated and contain a percentage of fines that exceed the maximum value of 3% (Baker, 1984), which classifies them as frost susceptible soils. If they are to be used, it should only be for non-structural fill applications, and preferably not within 2.0-m of the structure at any location.

4.2.1. Alternative I

The general construction method should insure that the foundation footers are placed in such a manner within the compacted fill pad that the permafrost table is absolutely protected from thaw conditions and is assured to refreeze to above the footer, at least one to two footer thickness (tf). This insures that the footer will never be impacted by destructive freeze-thaw cycling, and weakening of the load bearing compacted fill. Placement of rigid board insulation provides thermal resistance, insuring a maximum depth of thaw to above the footers (Figure 3). The following design is recommended with alternatives:

• Excavation and discard of the active layer soils. This requires a minimum of 1.5-m to 1.8-m of soils to be removed below the surface.

• After the active layer soils have been removed, visual observations are required to insure that none of the specific locations where footers will be placed consist of massive ice (wedge ice or relict ice). If massive ice is located within the footprint of the footer, over-excavation and complete removal of this ice is required before select fill placement. This will eliminate the destructive effects of creep settlement of the massive ice.

• Placement and compaction of NSF select material (< 3% fines passing 0.02mm) to at least two footer thicknesses on the top of the permafrost, compacted to 100% optimum density (ASTM D1557

Modified Proctor), and should extend beyond the perimeter of the structure by 2.5Lf.

• Placement of the footers.

• A thickness of NFS material should then be placed and compacted to 95% optimum density (D1557), from the bottom of the footer, to a minimum twice the footer thickness above the footer. The fill will extend beyond the perimeter of the structure a minimum of 2.5Lf.

• Rigid board polystyrene insulation should then be carefully placed together under the entire structure at a minimum depth of 0.6-m below final grade, and extending beyond the perimeter a minimum of

2.0Lf. The insulation shall be two layers of 5cm thick, polystyrene rigid board, butt together and overlapped.

• Continue placement with NSF material compacted 90% optimum density (D1557) to the final grade, insuring that the rigid board insulation does not shift during placement. Final thickness of compacted fill above insulation shall be a minimum of 0.6-m. This fill will extend beyond the perimeter of the structure a minimum of 2.5Lf.

• The columns extending from the structure to the footer should be insulated, and thermally decoupled from the structure if possible to prevent thermal bridging.

*The following excavation depths and replacement thicknesses assume the pre-construction grade is equal to the final grade. Because this project site is slightly sloping, care must be taken to insure that all active layer soils are removed, regardless of the elevation of the final grade.

Figure 3. Alternative I general foundation plan. Not to scale.

4.2.2. Alternative II

As was mentioned previously, dry bedrock has a high probability of being located at a depth of approximately 6.0-m below surface. This depth is very reasonable for an end-bearing and adfreeze combination pile type foundation. This method reduces the need for large scale active layer removal and insulation placement, as the load bearing elements will be driven or placed (excavated or drilled holes) on thaw-stable, high load bearing bedrock. This is the most conservative alternative when considering the uncertain future of climate change.

The availability of the equipment needed to either drive or drill for these piles could be an unsolvable issue. As is discussed in Section 4.4, construction progress must be made before building material arrival via vessel occurs. Installing piles after vessel arrival may not fit the construction schedule.

4.2.3. Alternative III

It is understood that placing of this structure on relatively shallow embedded footers on seismically unstable soils, predicates that the structure be classified Seismic Category D. Another method for placement that possibly could reduce costly seismic design is over-excavation to bedrock, or to some depth deeper than in Alternative I. Successive lifts of select material are compacted between the bottom of the excavation and the footers. With this method, the footers can be embedded at greater depth in the select fill pad and without the need for the horizontal sheet insulation. Although this may be technically possible provided that robust excavation equipment (heavy dozer with ripper shanks) is available, it is unknown if this method would provide economic advantages. In short, the additional cost of over-excavation and fill placement would need to be economically compared to the required construction protocols to insure compliance with Seismic Category D.

4.3 Bearing Capacity and long term strength

Previous geotechnical investigation for the FY05 Dormitory, within 60m of this project site, reported that the Standard Penetration Tests (SPT) (ASTM 1586) produced blow counts (N values) in excess of 100 per 15.2 cm for 27 of 31 intervals tested in five boreholes. Seven of the intervals required greater than 120 blows per 15.2cm, and four of the intervals required 200 blows per 15.2 cm (Baker, 2004). The blow counts indicate the instantaneous strength of these soils is extremely high, and this would be expected from large particle size sediments such as gravels and larger, cemented in an icy matrix. However, the moisture contents indicate thick intervals of sediments containing excess ice.

Mechanically, this indicates many of the soil particles are not in contact, therefore long term loading will deform the ice and particles visco-elastically. Therefore the design of the foundation on ice-rich soils should be governed by long term strength. The long term strength of ice-rich soils is a function of time and temperature, and the maximum temperature expected at the ice-rich soil and NFS fill interface was calculated to be -2.2°C utilizing the Modified Berggren equation and also TEMP/W two-dimensional soil temperature modeling software.

Laboratory measurement of short term and long term strength is a time consuming process, and requires the delivery of intact frozen samples to the laboratory. Therefore it was determined to omit those tests. A conservative empirical estimation often used in-lieu of laboratory data is given by Johnston

(1981). This estimation is based on compilation of data from ice-rich soils and from investigators, both

North American and Russian. It can be seen that the long term strength is a function of the temperature of the ice-rich soils:

σf = 35 + 28(0-Ts) (kpa) (4.1) where:

σf = long term (50 year) shear strength

0-Ts = the temperature at the critical depth (always positive)

By assuming the shear strength is near zero due to the ice content, Terzaghi’s equation for shallow square foundations can be reduced to:

..FS

N q fc f σ = (4.2) where:

qf = bearing capacity of the soil (kpa)

Nfc = 5.14 is a shape factor for circular or square foundations σ = long term shear strength of the soil

S.F. = is the factor of safety

For a calculated maximum temperature at the permafrost fill interface of -2.2°C, and a factor of safety of 3.0, we obtain an allowable bearing capacity of 165 kpa. This solution does not specifically determine the amount of settlement that will occur, however, because the footers will be placed on a thick layer of compacted fill, the ultimate load that is transferred from the fill to the permafrost should not exceed the above value, it is our opinion that creep settlement will be negligible and within normal tolerances for foundation such as this.

4.4 Earth Work

The following earth work recommendations assume construction utilizing Alternative I, and if

Alternative II or III are chosen, modifications to this earth work plan may need to occur. It is also assumed that footers and columns will arrive as soon as possible via ocean going vessel, which probably cannot occur until sea ice conditions allow in late June, or if icebreaking of the channel occurs.

Excavation of the active layer soils can take place any time prior to the arrival of the footers and columns, however, extreme care should be taken to not allow for thaw of the design elevation of the permafrost/fill interface. It is recommended that excavation proceed to approximately 0.5-m above the design permafrost/fill interface until the foundation materials have arrived or are about to arrive. It is also recommended that a CRREL representative be available during the excavation to ascertain the areas requiring over-excavation of massive ice (wedge or relict ice).

All the fill material shall be clean, select fill gravel allowing for dense compaction, and possessing no more than 4% fines. The fill will extend beyond the perimeter of the building as noted above and may be gradually sloped away from the building at 2:1 horizontal to vertical. Positive drainage away from the perimeter of the building preventing surface flow beneath the structure is required. Care should be taken to insure that no down-wind topographic lows occur that may allow for accumulation of drifted snow which will not be removed under normal snow removal procedures. The accumulated snow will serve as a very efficient thermal insulator, and winter heat extraction will be reduced in those areas allowing for diminished temperature decrease in the subsurface.

4.5 Long Term Temperature Monitoring

For structures located on permafrost, long term temperature monitoring is important for analyzing the performance of the foundation system and aid future engineering decisions. It is recommended that for either foundation alternative, a minimum of six thermistor strings be installed from the surface down to an appropriate depth. For the footer and column foundation, all the strings should extend down to the bottom of the excavation, with some located near the columns, and some located away. For adfreeze pile foundations, at least two strings should be located down the entire length of the pile, while the other four may be obtain a shallower depth and be located away from the piles. When a final decision is made regarding the foundation type, the temperature monitoring design can be determined.

APPENDIX A

Boring Logs

Soil Classification

The soils classification system for this report utilizes certain features from the Unified Soil Classification

System (USCS), ASTM methods D-2487 (Classification of Soils for Engineering Purposes), D-2488

(Description of Soils, Visual-Manual Procedures), and the Description of Classification of Frozen Soils by Linell and Kaplar (1966). The identification of the following characteristics are listed in the order they are reported:

a) Density or consistency (when available)

b) Color

c) Minor constituents

d) Major constituents

e) Trace constituents

f) Geologic characteristics (if applicable)

Color is used to distinguish between different soil layers, or to indicate the degree of weathering and/or oxidized state. Major, minor, and trace constituents of a soil are determined by visual-manual procedures or by measurement. Major constituents are those comprising more than 50% of the soil mass, minor constituents comprise 15% to 50% of the soil mass and are modifiers of the major constituents. Trace constituents are those comprising less than 12% of the soil mass. Visual ice descriptions are used to determine the degree of excess ice and

Table 2. Description and Classification of Frozen Soils.

Segregated ice is not visible by eye

Friable, poorly bonded. Material is easily broken up Nf

Well bonded, Soil particles strongly held together by ice

Excess ice Nbe

No excess ice Nbn

Segregated ice is visible by eye < 50%

Individual ice crystals Vx

Ice coatings on soil particles Vc

Stratified or distinctly oriented ice formations Vs

Random or irregularly oriented ice formations Vr

Ice > than 2.5cm in thickness Ice visible, >50%

ICE +

inclusions

Individual layer ICE

After – Linell and Kaplar, “Description and Classification of Frozen Soils”, International Conference on Permafrost, Purdue University, November 1963.

Table 3. Drill log abbreviations:

fine f.

medium med.

coarse cse.

slightly sl.

with w/ dark dk.

light lt.

occasional occ.

frequent freq.

trace tr.

interbedded int.

not applicable NA air rotary AR tricone TC downhole hammer DH weathered wx

APPENDIX B

GPR Radargrams

(Red arrows indicate location of inferred bedrock, green arrows indicate base of fill material)

Figure 4. Bldg 113, Grid 1, NW-SE Line 13

Figure 5. Bldg 113, Grid 1, NW-SE Line 21

Figure 6. Bldg 113, Grid 1, SW-NE Line 2

Figure 7. Bldg 113 , Grid 1, SW-NE Line 5

Figure 8. Bldg 113, Grid 2, NW-SE Line 1

Figure 9. Bldg 113, Grid 2, NW-SE Line 7

Figure 10. Bldg 113, Grid 2, SW-NE Line 5

Figure 11. GPR Grid and survey line locations.

SW-NE, Line 5

Grid 1 Grid 2

NW-SE, Line 7

NW-SE, Line 1

NW-SE, Line 21 NW-SE, Line 13

SW-NE, Line 2

SW-NE, Line 5

N

APPENDIX C

Photos

Figure 12. Typical surface soils. BH-3 @ 1.0m.

Figure 13. Typical active layer soils. BH-4, silty gravels with sand.

Figure 14. Typical sub-active layer frozen gravels. BH-3, note grey color.

Figure 15. Typical frozen, coarse sandy gravel/clear ice interface. BH-3.

Figure 16. Top of foliated ice (wedge ice) in BH-4.

Figure 17. Continuing wedge ice BH-4.

Figure 18. Clear ice with small gravel inclusions in BH-4.

Figure 19. Typical frozen silty gravels w/sand. BH-2.

Figure 20. Typical ice-rich silty gravels transitioning to ice-poor. BH-1.

Figure 21. Typical very ice-poor silty gravel with wx bedrock in BH-1.

Figure 22. Typical fill under Bldg. 113, silty gravels w/ sand. BH-5.

Figure 23. Frozen ice-rich coarse sandy gravels. BH-5.

Figure 24. Frozen, ice-poor to ice-free wx bedrock. This sample is from another project site, but this is what is expected at depth for more competent shale bedrock at FY12 Dormitory.

BIBLIOGRAPHY

Baker, Michael, Jr. (2004). Thule FY05 Dormitory Replacement Geotechnical Investigation, Thule Air

Base, Greenland. Anchorage, Alaska.

Baker, Michael, Jr. (1984). Geotechnical Investigation Report, Consolidated Cold Storage Building, Thule Air Force Base, Greenland. Beaver, Pennsylvania.

Bjella, Kevin (2010). Thule Air Base, Airfield White Painting and Permafrost Investigation, Phase II.

ERDC-CRREL Letter Report LR 10-05. Cold Regions Research and Engineering Laboratory, Hanover, NH, USA.

Corte, Arturo (1962). Relationship Between Four Ground Patterns, Structure of the Active Layer, and

Type and Distribution of Ice in the Permafrost. CRREL Research Report RR 88. Cold Regions

Research and Engineering Laboratory, Hanover, NH, USA.

Dawes, Peter (2006). Explanatory notes to the Geological map of Greenland, 1:500,000. Thule, Sheet 5.

Geological Survey of Denmark and Greenland, Danish Ministry of the Environment. Copenhagen, Denmark.

Funder, Svend (1990). Late Quaternary stratigraphy and glaciology in the Thule area, Northwest

Greenland. Meddelelser Om Gronland, Geoscience 22.

Johnston, G.H. (1981). Permafrost: Engineering Design and Construction. National Research Council of

Canada. John Wiley and Sons, New York.

Linell, K. and E. Kaplar (1980). Design and Construction of Foundations in Areas of Deep Seasonal

Frost and Permafrost. CRREL Special Report SR 80-34. Cold Regions Research and Engineering

Laboratory, Hanover, NH, USA.

Shannon & Wilson (1983). Geotechnical Studies, Control Tower, Thule Air Base, Greenland. Fairbanks, Alaska.