permafrost - transportation research boardonlinepubs.trb.org/onlinepubs/sr/sr247/sr247-025.pdf ·...

LChapter 25

FM

RUPERT G. TART, JR.

PERMAFROST

1. INTRODUCTION

T he term permafrost was first proposed by Muller as a conveniently short word to refer

to perennially frozen ground:

Permanently frozen ground or permafrost is de-fined as a thickness of soil or other superficial deposit, or even of bedrock, at a variable depth beneath the surface of the earth in which a temperature below freezing has existed contin-uously for a long time (from two to tens of thousands of years). Permanently frozen ground is defined exclusively on the basis of temperature, irrespective of texture, degree of induration, water content, or lithologic char-acter.... The expression "permanently frozen ground" however is too long and cumbersome and for this reason a shorter term "permafrost" is proposed as an alternative. (Muller 1947, 3)

The formation and existence of permafrost are largely controlled by climate and terrain but are also greatly affected by human intervention. Permafrost occurs in the northern and southern latitudes that are generally within about 35 de-grees of the North and South Poles. Permafrost underlies about one-fifth of the land surface of the world, including about half of Canada (Brown 1970). The distribution of permafrost in the Northern Hemisphere is shown in Figure 25-1.

Permafrost is also found at high elevations in lower latitudes. This high-altitude permafrost is

called alpine permafrost and is found, for example, in the Rocky Mountains of the United States and in the Alps of Europe. Alpine permafrost is more widespread and thicker on north-facing slopes, but variations in snow cover and vegetation on differ-ent slopes complicate the situation. In a totally different environment, permafrost is known to occur offshore in many northern areas of the world, including the eastern Canadian Arctic and under the Beaufort Sea (Samson and Tordon 1969; MacKay 1972; Hunter et al. 1976; Oster-kamp and Harrison 1976).

Geocryology is defined as the study of earth ma-terials at temperatures below freezing (Bates and Jackson 1980, 256). The geocryology of landslides is of greatest significance in cold regions where the mean annual air temperature is near or below freezing and the depth of freezing is significant with, respect 'to the landslide structure. In areas where the ground is permanently frozen, natural thawing or thawing induced by human activities can be a significant destabilizer. Earth material properties can change radically upon freezing (Andersiand and Anderson 1978; Johnson 1981; Eyles 1983; Harris 1987). Depend-ing on specific site conditions, freezing can be ei-ther a stabilizer or a destabilizer of earth materials (Johnston et al. 1981). Shear strengths of soils in-crease when soils freeze. Freezing ground can also cap and restrict groundwater flow, increasing pore pressures in soils remaining unfrozen below the freeze front. Conversely, shear strengths of soil

620

- -EXPLANATION

Ii.i.. e.

permafrost

ir permafrost

.jj V'i b$J1__

4 .

permafrost

II ,g o u 000

FIGURE 25-1 Distribution of permafrost in Northern Hemisphere (National Academy of Sciences 1983).

rseasonal frost Temperature (°C)

I - - -3 -2 1 0°C •1 •2 +3 .4 +5

t I I I I 1 I I

permafrost tablet

mean annual ground temperature

maximum mean

minimum mean monthly temperature monthly temperature -----

I " level of zero annual I amplitude

base of permafrost

60

geothermal gradient

0 2

25

622

Landslides: Investigation and Mitigation

FIGURE 25-2 Groundtemperature regime in permafrost (after Brown et al. 1981).

masses reduce as soils thaw, and thawing can re-lieve pore pressures that develop during periods of freezing. Thawing may promote increased drainage of saturated soils and thus may have the effect of increasing the shear strength of the mass. Therefore, specific sites must be examined closely to determine whether stability or instability may be caused by thermal changes.

2. OCCURRENCE OF PERMAFROST

The temperature within the earth increases grad-ually with depth. Geothermal gradients ranging from about 1°C per 22 m to 1°C per 160 m, de-pending on the type of soil or rock and the effect of past climate periods, have been measured in permafrost regions of the Northern Hemisphere (Jessop 1971; Judge 1973). The depth of per-mafrost can be estimated if the mean annual ground temperature is known and the local geo-thermal gradient can be estimated.

Figure 25-2 presents an example of ground tem-perature distribution in a permafrost region. This temperature profile is sometimes called a "trumpet" curve because of its characteristic shape. Annual fluctuations in air temperature produce correspond-ing fluctuations in ground temperature, but vegeta-tion and snow covering the surface decrease these

fluctuations, which also decrease with increasing depth, reflecting the thermal characteristics of the soil or rock. The limits of summer and winter tem-perature extremes converge below the zone of an-nual surface-temperature influence at a point known as the depth of zero annual amplitude and fol-low the geothermal gradient below that depth. As shown in Figure 25-2, the annual temperature fluc-tuations cause the annual melting and refreezing of the near-surface materials. This zone is called the active layer, and the top of the permanently frozen materials is called the permafrost table.

The permafrost region of North America (Fig-ure 25-1) is divided into two principal zones: the continuous permafrost zone in the north and the discontinuous permafrost zone in the south (Brown and Péwé 1973).

2.1 Continuous Permafrost

In the continuous permafrost zone, permafrost ex-ists everywhere beneath the land surface except in newly deposited unconsolidated sediments where the climate has just begun to influence the ground thermal regime. In this zone permafrost can extend to great depths. For example, in areas on the North Slope of Alaska, permafrost extends more than 350 m below the ground surface, and in the northern parts of the Canadian Arctic Archipelago it ex-tends to depths as great as 600 m (Judge 1973). At the southern limit of the continuous permafrost zone, permafrost is typically 60 to 100 m thick. The active layer is relatively thin, varying from 0.3 to 1 m. The permafrost temperature at the depth of zero amplitude ranges from —5°C in the south to —15°C in the extreme north.

2.2 Discontinuous Permafrost

In the discontinuous permafrost zone, permafrost exists together with layers of unfrozen ground known as taliks. The extent of permafrost terrain in an area depends on the history of the climate for the area and the condition of the ground sur-face. In the southern fringe areas of this zone, per-mafrost occurs as scattered islands ranging in size from a few square meters to a few hectares. Regions favorable to permafrost are usually cov-ered with a thick vegetative mat that insulates the permafrost strata from the current above-freezing mean annual surface temperatures. When this mat

Permafrost

623

is removed or disturbed, the permafrost degrades rapidly. When the insulating mat remains in place, the permafrost can be preserved for many years. Further north, the permafrost becomes more continuous and underlies a greater variety of ter-rain types.

Discontinuous permafrost ranges in thickness from a few centimeters in the southernmost areas to 60 to 100 m at its boundary with the continu-ous permafrost zone. Thicknesses may vary rapidly and irregularly over short distances. The depth of the active layer may exceed 3 m depending on local climate and surface conditions. The active layer may not always extend to the permafrost table; at some locations a zone of unfrozen ground known as the residual thaw layer is found between the base of the active layer and the permafrost table (Brownetat. 1981).

2.3 Saline Permafrost

Because permafrost is defined by temperature, it is not necessary for the soil fluids to become frozen solid for the strata to be referred to as permafrost. Salts in pore water are the most frequent cause of permafrost containing unfrozen soil fluids. The salt concentration of seawater is sufficient to cause a depression of the freezing point of about 1.7°C, and as the salt concentration increases, further de-pression of the freezing point occurs. As saline per-mafrost pore water freezes, salt is rejected in front of the freeze front, resulting in the formation of pockets of highly saline unfrozen pore water be-tween strata containing fully solidified ice. These conditions are common along the Arctic Ocean coastline in Alaska and adjacent areas in Canada where they have caused some difficulties for engi-neers designing facilities for existing small com-munities or associated with petroleum extraction (McPhail et al. 1976; Walker et at. 1983; Nixon and Lem 1984; Sellmann and Hopkins 1983).

3. CHARACTERISTICS OF FROZEN GROUND

From an engineering standpoint the physical prop-erties of permafrost are a prime concern. Frozen rock formations can have strength characteristics similar to those of unfrozen formations when the rock quality is high. Formations with low rock quality can contain significant volumes of ice, which can produce high strengths when frozen but much tower strengths when thawed. In addition,

there can be a volume change in the formation as a result of the thawing and subsequent draining.

Most soils show markedly different properties in their frozen and unfrozen states. Ice-rich strata have properties similar to ice, whereas the proper-ties of dry, clean granular soils can be similar to those of thawed granular soils. Field investigation techniques for frozen ground, including sampling, borehole logging, and field testing of frozen soil properties, have been discussed by Ladanyi and Johnston (1978).

3.1 Classification

A classification of frozen soils that has been widely used in North America was presented by Pihlain-en and Johnston (1963) and Linnell and Kaplar (1963). Linnell and Kaplar suggested the follow-ing three steps for description and classification of frozen soils:

The soil phase is identified independently of the frozen state using the Unified Soil Classification System, Soils characteristics resulting from freezing of the material are added to the description, and Ice found in the frozen materials is described.

The frozen-soil characteristics (step 2) are based on two major groups: soils in which segregated ice is not visible to the naked eye (Group N) and soils in which segregated ice is visible (Group V). Depend-ing on the degree to which the ice in Group N im-parts solidification and rigidity to the soil mass, soils may be subdivided into those that are poorly bonded (Group NO or well bonded (Group Nb).

Figure 25-3 shows the key elements of the Lin-nell and Kaplar frozen soil classification system. This system has been adopted in ASTM D4083 and can be used with all frozen soils to estimate their engineering properties when frozen and upon thawing. Those strata with little or no ice will be relatively stable upon thawing. For exam-ple, some clayey silts that were overconsolidated when frozen have shown very slight strength changes when they thawed.

In contrast, strata with ice lenses larger than 25 mm will not be thaw stable since these strata will undergo significant volume changes and strength reductions as the ice melts and drains from the formation.

624 Landslides: Investigation and Mitigation

DESCRIBE SOIL INDEPENDENT OF FROZEN STATE Classify Soil by The Unified Soil Classification System

Major Group Subgroup

Description (b) Designation Description (b) Designation

MODIFY SOIL DESCRIPTION BY DESCRIPTION OF FROZEN SOIL (a)

Segregated ice not visible by eye

N

Poorly bonded or friable

Nt

Well Bonded

No I excess ice I

Nbn

I Excess ice

. Nbe

Individual ice crystals Vx or inclusions

Segre9ated ice visible by eye (ice less than 25 mm thick)

V

Ice coatings on particles Vc

Random or irregularly oriented, ice formations

Vr

Stratified or distinctly oriented Vu ice formations

MODIFY SOIL DESCRIPTION Ice greater Ice with soil inclusions

ICE + Soil Type

BY DESCRIPTION OF SUBSTANTIAL ICE STRATA (a)

than 25 mm thick

ICE Ice without soil inclusions

ICE

Reference ASTM D4083.

Description of ice within frozen soil is based on visual examination of the sample in the field.

FIGURE 25-3 Summary of classification procedure for frozen soil (Linnell and Kaplar 1963).

3.2 Strength Properties

Behavior of frozen soil is controlled primarily by temperature and ice content and salinity. As the ice content increases, the soil mass acts more like pure ice. As ice content decreases and temperatures rise, the frictional components of the soil particles begin to contribute to the ground behavior. The degree of bonding of fine sands and silts can be evaluated by blowcounts obtained in the standard penetration test. Strata composed of fine sands and silts that are poorly bonded because of depression of the freezing point by saline pore fluids or because of tempera-tures near 0°C tend to have blowcounts of less than 100 blows per 30 cm (100 blows per foot in cus-

tomary units). Figure 25-4 presents some typical data on blowcount versus salinity from a coastal de-posit of silty fine sand with an average temperature of about —2°C (280F). A temperature-depth profile .of the site is presented in Figure 25-5.

3.3 Creep

A major consideration ivith frozen ground is the creep characteristics of ice and frozen soil. Ice will flow under sustained load; thus ice-rich soils must be investigated to determine their creep charac-teristics. As the temperature of frozen ground approaches 0°C, the rate of creep increases. Pre-dictions of creep rates for frozen ground were made

Permafrost

625

100

90

80

E 70

60

0

w 0.

50

40

0 30

20

10

Blowcounts obtained using 136 kg hammer, dropping 76 cm, driving a • 6.3 cm ID - 7.6 cm OD sampler.

Average ground temperature

-2degrees C

• ___ •

• . $ ________ __

• . • •

• $

0 10 20

30 40 50

60 70

by Nixon and Lem (1984) as well as by many oth-ers. The soil and ice parameters are not well de-fined and can be time consuming and difficult to obtain. As shown in Figure 25-6, the data can be scattered, and very slight changes in near-freezing temperatures can cause significant changes in the creep rate.

3.4 Frost Action

Salinity, parts per thousand

tenance. A number of landslides have been caused by these processes.

Three basic conditions must exist for frost ac-tion to occur (Johnston et al. 1981):

The soil must be frost-susceptible, There must be a water supply, and The soil temperature must be sufficiently low to allow some soil water to freeze.

FIGURE 25-4 Blowcount versus salinity data at saline permafrost site in western Alaska.

Frost action in soils includes the detrimental processes of frost heaving, which result mainly from the accumulation of moisture in the soil in the form of ice lenses at the freeze front during the freezing period, and thaw-weakening, or the decrease in strength when seasonally frozen soil thaws. Frost ac-tion is an extremely complex process. Many studies of frost action and related problems have been undertaken (Johnson 1952; Jessberger 1970). Frost heaving and thaw-weakening of soils produce sig-nificant damage and frequently require costly main-

Frost-susceptible soils are most commonly identi-fied by criteria based on grain size. A grain-size cri-terion originally proposed by Casagrande (1931) appears to offer the most useful rule of thumb for identifying frost-susceptible soils (Linnell and Kaplar 1959). It has been used by the U.S. Army Corps of Engineers to create a frost design soil classification. This frost classification system is fre-quently used in the design of pavements and is presented in Figure 25-7.

Casagrande stated:

TEMPERATURE (°C)

-8 -6 -4 -2 0 2 4 6

2

4

6

12

14


Under natural freezing conditions and with a sufficient water supply one should expect con-siderable ice segregation in non-uniform soils containing more than three percent of grains smaller than 0.02 mm and in very uniform soils containing more than ten percent smaller than 0.02 mm. No ice segregation was observed in soils containing less than one percent of grains smaller than 0.02 mm, even if the ground water level was as high as the frost line. (Casa-grande 1931, 169)

As noted by Casagrande, the migration of water in both unfrozen and frozen soils is complex and basic to frost action and heaving.

Thaw-weakening is also a complex process and may occur in soils, especially clays, whether or not ice lenses have formed during an earlier freezing period. Ice-rich permafrost has very low perme-ability. The permeability is controlled by temper-ature and typically ranges from 10 cm/sec to less than 10.11 cm/sec, as shown in Figure 25-8.

FIGURE 25-5 Ground temperature data at saline permafrost site in western Alaska (Krzewinski and Tart 1985). REPRINTED WITH PERMISSION OF AMERICAN SOCIETY OF CIVIL ENGINEERS

FIGURE 25-6 Creep strain data from port site in western Alaska superimposed on trends presented by Nixon and Lem (1984).

1000

100

10

1-4-1 .00E-04

o ppt

000

18 ppt

25 ppt

35 ppt

46 ppt @ -4 deg C 65 ppt @ -4 deg C

ta points taken from authors files.

ies taken from Nixon and Lem (14.13, Fig 3) Thickened line represents 0 ppt data @

-2.5 deg C.

All other lines @ - deg C.

linity in parts per thousand is shown on each line.

1 .00E-03 1 .00E-02 1 .00E-01 1 .00E+00 1 .00E+O1

Creep Strain, per year

Permafrost

627

FIGURE 25-7 Summary of U.S. Army Corps of Engineers method of defining frost susceptibility of soils (ASTM 1993).

FROST GENERAL GROUP SOIL TYPE

Fl Gravelly soils

F2 (a) Gravelly soils (b) Sands

F3 (a) Gravelly soils Sands, except very fine silty sands Clays, P1> 12

F4

j

a) All silts Very fine silty sands Clays, P1 < 12 Varved clays and other fine-grained, banded sediments

PERCENTAGE FINER THAN

TYPICAL 0.02 mm

UNIFIED SOIL BY WEIGHT

CLASSIFICATION

3 to 10

GW, GP, GW-GM, GP-MG

10 to 20

GM, GW-GM, GP-GM 3 to 15

SW, SP, SM, SW-SM, SP-SM

Over 20 GM,GC Over 15 SM,SC

- CL, CH

- ML,MH Over 15 SM

- CL, CL-ML - CL and ML; CL, ML, and SM;

CL, CH, and ML; CL, CH, ML, and SM

3.5 Seasonal Freezing and Thawing

In regions where there are prolonged winter peri-ods with daily average air temperatures at or below freezing, the ground surface freezes. Conduction is the dominant ground heat-transfer process; how-ever, the description of conductive heat flow must also be accompanied by an evaluation of the liber-ation or adsorption of heat resulting from phase changes in the water. '(Ihenever a substance such

as water changes from a solid to a liquid (by thaw-ing) or from a liquid to a gas (by evaporation), considerable volumes of heat energy are required. Conversely, considerable heat energy is released when a gas converts to a liquid (by condensation) or from a liquid to a solid (by freezing). The term latent heat is used to define the energy components associated with these phase changes.

For estimation of seasonal freezing and thaw-ing, ground surface temperature variations 'are

FIGURE 25-8 Hydraulic conductivity of frozen soils (Nixon 1992).

LEGEND 10- -. - - - - - ..

1111 III o-O MORIN CLAY

o 108 INUVIK CLAY

E 10 - - - - - - MORIN CLAY

10-10 'N - - - . - - - . . y TOMOKOMAI SILT-CLAY - P JAPANESE CLAYS

10-11 0

S SILTLOAM

10r12 - - - - . i_ S•,ç - - . i

0 _____ £ CHENA SILT

O 10

0.1 0.2 0.5 1 2 5 ic U MANAITABASHI CLAY

1 TEMPERATURE, -T (degrees C) _______ X X KAOLIN

628


usually not considered in detail, but average val-ues for the entire freezing or thawing period are used instead. In practice, the average seasonal temperatures are represented by the freezing or thawing index. The air freezing (Iai) and thawing

('at) indexes are determined by computing the sum of the deviations of mean daily air tempera-tures from the freezing point during the thawing or freezing season. For time scales on the order of months, average ground surface temperatures can be correlated reasonably well with average air temperatures. For design purposes ground surface temperatures are estimated from air temperatures using an empirically defined n-factor, the ratio of the ground surface freezing or thawing index (I, f or 'St) to the air freezing or thawing index (Iai or

fl1 = If /If nt = IsC/Iat (25.1)

The magnitude of the freezing and thawing n-factors (n1 and n) depends on climate condi-tions as well as type of surface. In summer, vegeta-tion-covered but treeless surfaces that remain wet will have average ground surface temperatures about equal to the corresponding average air tem-peratures, and the n-factors will thus be close to unity. Snow is an excellent insulator, and its pres-ence causes the mean annual ground temperature to be several degrees warmer than it would be oth-erwise. Dry or well-drained surfaces, such as road-ways, because of their sensitivity to variations in solar and long-wave radiation, have surface tem-peratures that vary, over a wider range than do air temperatures. Care must be taken in estimating n-factor values (Lunardini 1978).

The depth of annual freezing is also dependent on the characteristics of the ground, especially its moisture content. The latent heat of a soil mass is dominated by the latent heat of water, which is the amount of heat that must be removed from water before it becomes ice and can cool to tem-peratures below 0°C. An identical amount of heat is required to melt ice. This amount of heat is much greater than the specific heat of water, the amount of heat it takes to warm the water 1 degree if there is no phase change. The latent heat of the mineral soil particles is zero since no phase change occurs. Thus, as the soil moisture content increases, soil masses freeze and thaw more slowly.

3.6 Ground Movements Resulting from Freezing

The process known as frost creep has been sub-jected to considerable study since it is readily mea-sured in the field. The potential amount of downslope movement due to frost creep (C) re-sults when the soil freezes and heaves a distance H perpendicular to the slope and a subsequent thaw causes the soil to settle vertically. The values of C and H and the slope angle e are then related ac-cording to the following equation:

C=HtanO (25.2)

Field measurements of frost creep suggest that it may be less important than first believed and that other mechanisms may combine with frost heave to create downslope movements. For example, Jahn (1975) measured 30 mm of downslope movement on a 4-degree slope in Spitsbergen, yet the measured frost heave of 0.15 in would only account for about 10 mm of the downslope movement. Most investi-gators suggest that viscous flowage of saturated soils is an important aspect of downslope movements of thawed soils, even on gentle slopes. This flow of sat-urated unfrozen earth material is called solifluction, which is discussed in more detail in Section 4.1.1.1.

3.7 Drainage Cutoff Caused by Freezing

Groundwater flow can be cut off or diverted by frozen zones with much lower permeabilities. This frequently occurs when the surface freezes and the groundwater is trapped beneath a surface cap and the permafrost table. Under these artesian condi-tions, quite high pressures can build up. These pressures may be further increased as the zones are progressively invaded by advancing freeze fronts. The buildup of pore-water pressure reduces the shear strength of the soil strata in which the pres-sures develop. Weaknesses in the frozen surface capping zone can result in seepage zones that may flow both summer and winter. In summer these zones form springs, but in the winter the discharg-ing water freezes and builds up icing deposits—lay-ers of solid ice that may block or damage transportation facilities or cover slopes, which may become unstable when the icing deposits melt in the following summer and provide a large supply of water (Figure 25-9).

Permafrost

629

FIGURE 25-9

Example of slope icing. R.O. TART. JR.

9-4

3.8 Ground Movements Resulting from Thawing

Frozen soils frequently contain moisture contents in excess of saturation, even for very loose soils. When these soils thaw, the soil matrix compresses as the excess moisture is forced out of the pores. This compaction is known as thaw-consolidation. The amount of compression is proportional to the ratio of the in-place frozen density to the after-thaw density. lithe moisture content of the frozen soil can be determined, assuming that the soil has ex-cess moisture, the frozen density can be calculated. The after-thaw density can be estimated by deter-mining the soil type and the overburden pressure. Thaw strains of 25 percent are common in frozen fills and natural ground formations with high mois-ture contents. Thaw-consolidation should be antic-ipated in soils with frozen moisture contents of 20 percent or higher unless the soils contain a clay component. Taber (1943) described the role of thaw-consolidation in the processes of solifluction and skin flow, two major types of slope movement in originally frozen soils (see Section 4.1).

3.9 Shear-Strength Reduction During Thaw

Excess water is frequently drawn into freezing soil strata, causing these strata to become saturated and loose. When thawing occurs, the shear strengths of these loose saturated soils are very low until the soils have been allowed to drain and dry. Provision or adequate drainage is frequently a problem in the topographically low-relief regions common to many areas in the Arctic of North America. Low ,,,,it temperatures inhibit rapid drying.

4. TYPES OF PERMAFROST SLOPE MOVEMENTS

In permafrost areas landslides are found in both frozen and unfrozen ground. Landslides involving unfrozen ground are mostly found in areas of re-cent sediments adjacent to bodies of water. Unfrozen Cretaceous shales are also affected by landsliding in Canada's Northwest Territories. The discussion in this chapter will be restricted to landslides in frozen or partially frozen ground. Most landsliding involving frozen ground begins when thawing occurs; shearing through frozen ground has been noted in some natural slopes but is relatively rare. There is also the potential that dynamic loadings related to earthquakes may cause liquefaction of thawed cohesionless soils sit-uated between the frozen near-surface portions of the active layer and the underlying permafrost (Finn et al. 1978).

4.1 Classification of Permafrost Landslides

Typical landslide features and associated mass-wasting phenomena have been widely described in the literature (Bird 1967; MacKay 1966; MacKay and Mathews 1973; Washburn 1973). Because of the great degree of human activity along the Mackenzie River Valley of northwestern Canada, the characteristics and distribution of landslides in this region have received consider-able attention (Isaacs and Code 1972; Chyurlia 1973; Code 1973; McRoberts and Morgenstern 1973). McRoberts and Morgenstem (1973, 1974a) proposed a classification of permafrost landslides based solely on descriptive features (Figure 25-10). The major divisions in this classification—flows, slides, and falls—closely correspond to the stan-

T::&

630


FIGURE 25-10 Classification of permafrost landslides (after Brown etal. 1981).

MAJOR MASS MOVEMENTS IN PERMAFROST TERRAIN

I I I FLOW SLIDE FALL

I I I I I I Solifluction Bimodal Skin Multiple Multiple Block Rotational

Retrogressive Retrogressive (Unfrozen)

dard terminology for describing landslides, as de-fined in Chapter 3 of this report. However, the subdivisions of these categories contain some non-standard terms that define some of the unique landsliding conditions found in permafrost areas.

4.1.1 Flows

Flow-dominated movements on natural and cut slopes are common in permafrost areas, especially when the slopes are composed of fine-grained frozen soils. Studies by Pufahl (1976) demon-strated the complexities of the heat balance on the surface of exposed and thawing permafrost. Denudation of vegetative cover on ice-rich per-mafrost slopes usually results in thawing. Because moisture contents are frequently well above 20 percent, flowing or creeping, or both, of the thawed zone is a natural consequence. This move-ment continues until the next winter, when freez-ing begins to restabilize the slope.

The McRoberts and Morgenstem classification (Figure 25-10) considered only naturally occurring landslides and included four types of flow move-ments: solifluction and skin, bimodal, and multi-pie retrogressive flows.

4.1.1.1 Solifluction

Solifluction is defined as "the slow downslope flow of saturated, unfrozen earth materials" (National Research Council of Canada 1988, 78). Soli-fluction has been observed on natural slopes and on slopes modified by human construction. In each case, freezing in winter seals the sloping ground surface, which then prevents surface drainage of pore water and causes water contents to reach levels of saturation and higher. As the saturated zone warms in spring and summer, the slope slowly creeps downward. As the upper zone thaws and drains, it becomes more stable but con-tinues to move on top of the ice-rich materials, which remain frozen. A classic example of soli-

FIGURE 25-1 1 Solifluction lobes on alpine permafrost slope in eastern Kyrgystan. R.G. TART, JR.

Permafrost

631

fluction is shown in Figure 25-11. Solifluction is similar to flow movements but is much slower and not as deep-centered. Figure 25-12 shows Solifluc-tion around a deeply embedded pipe pile. Inclino-meters placcd closc to this pile showed the greatest movement at the surface with the magni-tude of movement decreasing with depth (Figure 25-13). The movement is greatest during the pe-riod of spring thaw and slows as drainage occurs through the summer. It stops when the ground refreezes in the autumn.

4.1.1.2 Skin Flows Skin flows involve the detachment of a thin cover of vegetation and mineral soil and subsequent movement of this mass over a planar undulating surface. These failures have much in common with the debris flows occurring in residual or colluvial soils (see Chapter 20). They form long ribbonlike shapes down the slope that may coalesce to involve large areas and are most common in ice-rich sedi-ments in the discontinuous permafrost zone, partic-ularly following a forest fire (Hardy and Morrison 1972; Hughes 1972; MacKay and Mathews 1973).

Skin flows develop when the pore pressures generated during thawing reduce the shearing re-sistance of the soil materials to low levels. Accordingly, skin flows may develop on slopes as flat as 6 degrees (McRoberts 1973), although they are frequently found on steep slopes.

4.1.1.3 Bimodal Flows Bimodal flows were so named by McRoberts and Morgenstern to indicate that two modes of mass movement occur. These dual modes of movement are reflected by the characteristic low-angle tongue of flowed debris and the steep head scarp. Bimodal flows have many characteristics in com-mon with the landslides occurring in sensitive clay deposits (see Chapter 24).

In most active bimodal flows, permafrost in the head scarp is exposed to the atmosphere and de-grades by ablation. The surfaces of the tongues have very low slopes and the water contents of the materials forming the tongue may approach the liquid limit of these soils. Figure 25-14 shows plan and section views of a bimodal flow landslide along the Mackenzie River.

Bimodal flows are initiated by some process or combination of processes that removes the insu-lating cover of vegetation and thawed soil, expos-

/ .. - -;•.is

4

*

ing the ice-rich permafrost in such a manner that ablation begins and is sustained. Following initia-tion, ablation can generate substantial rates of

mass wasting. Observations by McRoberts (1973) suggested that head scarps may regress at rates ranging from 3 to 20 cm/day during the season of

thaw. With such rates it is evident that bimodal flows can pose a threat to adjacent structures and transportation facilities.

4.1.1.4 Multiple Retrogressive Flows Multiple retrogressive flows are dominated by flowage but may retain some portion of their pre-failure relief. Their form suggests that a series of retrogressive failures have occurred in the head scarp. Rotational and nonrotational sliding may also be observed in portions of these flows, which may cover quite large areas and are usually rela-tively shallow.

4.1.2 Slides

McRoberts and Morgenstern subdivided per-mafrost slide movements into three classes: block, multiple retrogressive, and rotational slides (Figure 25-10). Block slides frequently involve a single large block that has moved downslope with little or no back tilting. Multiple retrogressive slides are characterized by a series of more or less arcuate blocks that are concave toward the toe and step upward from the toe to the head scarp. The blocks may exhibit back tilting.

FIGURE 25-12 Solifluction flow around 45-cm pipe pile. R.G. TART, JR.


-5.0

Displacement (cm)

0.0 5.0 10.0 15.0 -10.0

Temperature (C)

-5.0 0.0 5.0 10.0

I _________________

5.0 II

10.0 10.0' o 0

o 0

- —o--l1/1/90 . —D-11/1/90

-+- 10/30/9 1 -+- 10/30/9 1

—o— 5/6/92 V

—o— 5/5/92

15.0 -*-- 7/29/92 15.0 -%- 7/29/92

—a— 10/28/92 -- 10/28/92

-*- 5/20/93 —x— 5/20/93

—o-- 10/20/93 —o— 10/20/93

20.0 20.0

FIGURE 25-13 Inclinometer displacements for slope in discontinuous permafrost showing solifluction flows.

Multiple retrogressive and block slides occur in materials whose water contents appear to be lower than those that result in flows. These slides are as-sociated with shear failure in frozen and unfrozen soil and are frequently quite large features 45 to 60 in high. McRoberts and Morgenstem (1974b) dis-cussed the mechanics of slides in frozen soil ex-tensively and suggested that the base of these failures often lies in unfrozen clay, at least for the sites they studied in the Mackenzie River Valley.

Figure 25-15 shows one example of a multiple retrogressive slide. It appears to be the result of groundwater conditions in a talik beneath the ac-tive layer. Thawing presumably provided addi-tional excess moisture to the sliding mass, and the result was an almost circular landslide. Figure 25-16

shows the presumed positions in the spring and in the fall of the groundwater source believed to have contributed to this landslide.

4.1.3 FaIls

Permafrost-related falls involve the gravity-influenced rapid downward dropping of detached blocks of frozen materials. Falls are common along the banks of rivers and lakes and parts of the Arctic Ocean coastline where thawing and erosion undercut the permafrost-bonded bank materials. Large blocks of frozen material period-ically break off, facilitating bank recession. Recession rates as high as 10 rn/year have been reported (Walker and Amborg 1963).

Permafrost

633

PLAN

TRIM LINE NORTH SCARP

MACKENZIE

\

RIVER

VET- THRUST RIDGES( OPP1EDVEGE1ATIO'T3

OPE

SMALL GULLY 4

CROSS SECTION

I LOBE ANGLE OVERALL 7- 75 HEAD SCARP 400 I

w INORTH SOUTH SCARP 13° TO 1401 NORTH SCARP IPERPENDICULAR TO RIVER SCARP z 0

25 < LIMIACKENZIE RIVER >

0

0 100

TEST P I T

N

-4-.

HEAD SCARP HANNA ISLAND LANDSLIDE

STADIA SURVEY 71 JUNE 1972

RIVER ELEVATION ASSUMED - 0 lee'

0 25 50 lee' I I

SCALE VERTICAL, HORIZONTAL

PROFILE BEHIND HEAD SCARP

NORTH SCARP FOOT

200

300

FIGURE 25-14 Bimodal flow near Hanna Island, Mackenzie River, Northwest Territories, Canada (BROWN ETAL 1981).

HORIZONTAL DISTANCE. feet

4.2 Selected Examples of Permafrost Landslides

In the following sections a number of slope failures in permafrost regions are discussed. Taken mostly from the author's own experience, they have been selected to illustrate typical slope stability prob-lems encountered in permafrost regions.

4.2.1 Flow Movements in Mine Wastes

Flows frequently occur as a result of the placement of ice-rich overburden or waste materials from mining operations. In many cases, the waste is frozen when placed in a stockpile and can take more than one season to thaw completely. Rapid thaw can result in large flows, as shown in Figure 25-17. At this site ice-rich mine wastes stored on a sloping surface during winter began to flow as the waste thawed.

4.2.2 Flowage of Fill Material

Figures 25-18 and 25-19 show a flow failure at a location where a fill containing ice was placed on sloping natural ground, probably during the win-ter. As the fill thawed, the saturated mass began to flow. During subsequent winters the surface of the fill froze, confining incoming groundwater within the fill. During the spring and summer each year, as the thaw front moved into the loose ice-rich fill, the fill began to flow.

4.2.3 Solifluction and Creep Movements

Permafrost areas frequently include landslides that develop as a result of either shallow or deep-seated creep. Figure 25-20 presents the profile of a per-mafrost slope that is creeping and includes incli-nometer deflections at various locations. The inclinometer deflections show the effects of deep

634


FIGURE 25-1 Multiple retrogres-sive earth slide in discontinuous permafrost. R.G. TART. JR.

creep and reduced movements in the the ground surface. In this case the ice-rich zone is sliding and creeping over a denser, drier zone. The creep fail-ure surface developed at a depth at which the per-mafrost temperature had warmed sufficiently to creep but not enough to allow for adequate drain-age and increased strength.

4.2.4 Landslides in Soil and Rock Cut Slopes

Figure 25-21 shows a permafrost rock cut slope that has become unstable as it has thawed, releas-ing both blocks of bedrock and masses of satu-rated, unstable soil. The darker materials in the center of Figure 25-21 are weathered rock and soils that were frozen at the time of construction but began sloughing and sliding as the slope thawed. This release is typical of a thawing com-posite soil-and-rock-slope failure. The slope ap-peared stable during construction, but the insulating overburden was removed, exposing the frozen and weathered bedrock to thawing. After the thaw front was able to penetrate the rock structure, blocks were released along the surfaces

of structural weakness, including both foliation surfaces and joints. Additional volumes of thaw-ing soils added considerably to the total volume of unstable material.

Another example of this type of landslide can he seen in Figure 25-22. A mining operation de-veloped a 1 10-rn-high cut slope through old lake deposits to uncover a placer gold deposit in an an-cient buried channel. The excavation approached maximum depth about midsummer. The lake de-posits were mostly fine-grained soils with high water contents, although there were also some coarser-grained, and thus better-drained, strata. Seepage from the fine-grained strata could be ob-served during and after the landsliding, which is shown in more detail in Figure 25-23. The fine-grained soils with high water contents would become ice-rich when frozen. Although no con-firming evidence is available at this time, it is pos-sible that because of the high latent heat of the fine-grained ice-rich lake deposits, they were frozen at the time of cutting, whereas the well-drained de-posits had thawed. For this reason, the fine-grained strata became unstable upon thawing, and the bet-ter-drained strata remained relatively stable.

Permafrost

635

GROUNDWATER FLOW

04

FENG O FREEZE FRONT

- ROUND

- ->- MOVING DOWNWARD OR

IMPERVIOUS ZONE

FIGURE 25-16 Schematic section views showing prob-able causes of multiple retrogressive slide.

GROUNDWATER FLOW BECOMES BLOCKED

PORE PRESSURE BUILDUP DURING FREEZEUP

THAW FRONTS MOVING TOGETHER

THAWED

040

DETERIORATING FROZEN Op ZONE THAW

FROZEN GROUND OR

IMPERVIOUS ZONE / -

FAILURE SURFACE

SLOPE RELEASE DURING BREAKUP

Any ice-rich slope will begin to move as it thaws. The movements. may include surficial sloughing and flow movements as well as more deep-seated landslides. Steeply inclined thawing slopes are frequently more stable than gently in-clined slopes, and almost vertical slopes are often the most stable because vegetation mats fall more readily over the edges of steep slopes and begin to develop a new insulating cover that reduces thaw-ing and increases stability. In the thawing per-mafrost slope pictured in Figure 25-24, minor surficial sloughing continues as a new vegetation mat slowly develops.

4.3 Associated Slope Stability Problems

The most common problem associated with slope stability in permafrost areas is the stabilization of earthworks, especially dams and dikes.

4.3.1 Stability of Embankments

If fill slopes are constructed in permafrost areas, the fill materials must be allowed to thaw before their placement if they cannot be protected from thaw-ing after construction. It is virtually imj,ossible to compact frozen chunks of soil unless the moisture

636

Lands/ides: Investigation and Mitigation

FIGURE 25-17 Flow of mine wastes R.O. TART, JR.

!Ih

71 MW

FIGURE 25-18 Flow failure of pipeline workpad. R.G. TART, JR.

content of these chunks is less than 5 percent. Even then compaction will be inefficient and pos-sibly ineffective. Therefore, any earthwork project that will thaw during its lifetime must he thawed at the time of construction to prevent instabilities caused by subsequent thawing. An example of a flow failure of a fill was discussed in Section 4.2.2 and shown in Figures 25-18 and 25-19.

4.3.2 Thermal Piping or Erosion

Frozen fine-grained soils containing adequate water contents in the form of ice are quite imper-meable. Naturally occurring deposits and con-structed embankments containing such materials have otten been used as dams in permafrost re-gions. Thcsc matcrials can thaw and thcn fail cat-astrophically by a process known as thennal piping. Erosion (or piping) of fine-grained soils refers to the progressive removal of smaller soil particles as flow passes through the erosive soil mass. If left unchecked, this veIl-known process may lead to complete destruction of the soil structure. Thermal piping refers to the addition of heat to a frozen mass by the infiltration of warmer ground-

SEASONAL THAW

0 5 10 METERS I I

SCALE

FIGURE 25-19 Profile of flow failure of pipeline workpad.

0.0

-5.0

Displacement (cm)

0.0 5.0 10.0 15.0

•

20.0 25.0

I 0.0

-10.0

Temperature (C)

-5.0 0.0 5.0 10.0

II I 13

5.0 5.0

X.

10.0 1 + * 1 10.0

±+ *A ô .9 + o * x —a-- 2/5/91 2 —a— 2/5/91 + * X • -+-- 6/4/91 -+-- 6/4/91

—a-- 10/23/91 . —o-- 10/23/91 15.0 . —w— 6/2/92 15.0 -*-- 6/2/92

—a-- 10/15/92 —a— 10/15/92 -*- 5/26/93 —x-- 5/16/93

9/30/93 —o-9/30/93

20.0 20.0

FIGURE 25-20 Inclinometer displacements on warm permafrost slope creeping near base of permafrost.


FIGURE 25-21 Cut slope failures due to thawing permafrost affecting both soil and bedrock (upper left), which produced rock falls that damaged concrete barriers (Jersey barriers) placed adjacent to highway shown in foreground. R.G. TART, JR.

water, which results in thawing of the ground ice. The ice itself may have formed a network of pas-sages through which water may preferentially flow once the ice has thawed. If the soils are relatively fine-grained, these flows may also subsequently erode (or "pipe") the now highly erodible fine soil particles in the thawed soil mass.

In the area shown in Figure 25-25, a water sup-ply lake was lost when thermal piping resulted in a breach in an embankment that retained the lake.

5. METHODS FOR CONTROLLING STABILITY OF PERMAFROST SLOPES

There are two primary types of methods for con-trolling the stability of permafrost slopes: physical restraint and thermal modification. In some cases, combinations of these two approaches can pro-vide the most cost-effective means of control.

5.1 Physical Restraint Methods

Physical restraint ineiliods are similar to those used in thawed soils and can include, but are not limited to, the following:

OBSERVED FAILURE SURFACE

UNDIFFERENTIATED - GLACIAL DEPOSITS


LAKE DEPOSITS (PROBABLE ICE RICH ZONES)

BEDROCK SURFACE -


FIGURE 2 5-22 Open-pit mine cut-slope failure due to thawing discontinuous permafrost.

- - T \CHANNEL

FIGURE 25-23 Closer view of open-pit mine cut-slope failure shown in Figure 25-22. R.G.TART JR.

- - 1 r" -rnfl ~

loin N_

- 1 FIGURE 2 5-24 Partially stabilized cut slope in

-

permafrost. R.O. TART, JR.

Permafrost

639

Drainage improvements, Buttresses, Retaining walls, Slope geometry modifications, and A variety of other methods.

These methods are covered in Chapter 17 of this report.

5.2 Thermal Modification Methods

Thermally modified slopes are unique to cold re-gions. Thermal modifications involve

Maintaining the current thermal state of the slope, Removing heat from the slope to increase freez-ing, or Adding heat to the slope to increase the rate of thawing.

In considering thermal modification, it is very im-portant to know the thermal properties of the earth mass being modified and to perform the ap-propriate heat flow calculations to predict the rate and amount of modification that will take place.


lce Wedge

114

FIGURE 25-25 Permafrost dam on North Slope of Alaska breached by thermal piping. R.G. TART, JR.

5.2.1 Maintaining Current Thermal State

Thawing of permafrost is frequently caused by construction activities that disturb the ground cover of vegetation and organic materials. These materials are usually very good insulators and in some cases do not allow the thaw front to pene-trate during the short summer seasons characteris-tic of the arctic regions. When these materials are crushed or removed and the underlying soils are allowed to thaw, slope instabilities and settlement usually result. This disturbance of the insulating cover can be mitigated in some instances by re-placing the cover with new vegetation or some other type of insulating material.

An example of this type of modification is shown in Figure 25-26. Wood chips from a sawmill were placed on the surface of a slope in a per-mafrost area to provide additional insulation. Throughout the fall day on which this picture was taken, the wood chips remained frozen, but the surrounding ground surface was thawed, as mdi-

cated by its darker color. Seepage was also observed on a section of the slope where the insulation was not being used. Additional protection at this loca-tion was provided by passive heat-extraction de-vices, discussed further in the following section.

5.2.2 Heat Removal from Ground

Heat may be removed from the ground by active or passive methods. Active refrigeration is one common heat-removal technique. This type of heat removal is used, for example, in indoor ice rinks; another example is the artificial freezing used to stabilize tunnels during construction in nonpermafrost areas. Because of its cost, however, it is usually used only for temporary stabilization. Several Trans-Alaska Pipeline pump stations that are located on permafrost have refrigerated foun-dations to prevent thaw settlement of the pump-station buildings. Passive methods include heat pipes and various types of passive air ventilation systems. Heat pipe systems have been used to sta-

Permafrost

641

bilize slopes in Alaska, as discussed at the end of this section. Although other passive heat removal systems are in use, they have primarily been used to stabilize building foundations. The most com-mon of the other systems is a passive air ventila-tion system in which a set of surface manifolds is Ut jeitied so as to passively maintain a sufficient pressure differential to cause a continuous flow of cold air through subsurface ducts.

I leat pipes ate dosed iubes filled wiih a fluid that boils at temperatures below freezing and then condenses at cooler temperatures. When the ground temperatures are near freezing, the fluid boils and rises to the top of the tube. During the winter, the cold air passing over the heat pipe fins causes the fluid to condense and flow to the bottom of the tube, where it receives heat from the ground to warm it, causing it to boil and rise once again. Thus these systems remove heat from the ground in a continuous cycle. In the summer, when the air temperatures are higher than the ground tempera-tures, the heat pipes are inactive. This is a passive heat extraction process because no external energy is required to make this system operate.

At the location shown in Figure 25-26, heat was also removed from the ground by passive ther-mal heat pipes. Two finned heat pipes can be ob-served on top of each pipe-pile supporting the pipeline and also in free-standing heat pipes be-tween the piles. Each of these heat pipes removes heat from the ground in order to maintain the un-derlying permafrost and thereby increase the sta-bility of the slope. Figure 25-27 shows finned heat pipes in a fill slope that overlies ice-rich per-mafrost near 0°C. Near the toe of the slope the underlying permafrost began to thaw and initiate slope movements. Finned heat pipes were in-stalled along the toe of the slope to remove heat from the permafrost and retard the thawing, thus providing more stability to the slope.

removing vegetation, covering the area with a dark material in summer or with a temporary heated fa-cility, flooding, and other methods.

5.3 Design of Cut Slopes

Performance of highway cut slopes in permafrost soils has been studied in Alaska, the Yukon, and the Northwest Territories (Lotspeich 1971; Smith and Berg 1973; Pufahl et al. 1974; McPhail et al. 1976). These studies and experience in permafrost terrain led McRoherts (1978) to suggest that cut slope designs and configurations can be best eval-uated in terms of four standard classes, which he termed Types I to IV.

FIGURE 2 5-26 Permafrost slope insulated with wood chips (lighter-colored ground surface along pipeline), Trans-Alaska Pipeline. Heat pipes are also visible on each pipe pile supporting pipeline and as free-standing elements between piles. R.O. TART. JR.

5.2.3 Heat In put to Ground to Promote Thawing

In some cases, the addition of heat to the ground to promote thawing is the most cost-effective method of stabilization. In areas of discontinuous per-mafrost, it may be possible, by surface modification, to increase the heat flow into an area enough to cause it to thaw and thus allow thaw strain to occur before construction. This can be accomplished by

5.3.1 Type!

Type I consists of low cuts, 1 to 3 in high, that are capable of self-stabilization provided there is a rea-sonably well-developed organic cover at least 0.2 m thick overlying the mineral soil. The cut should be made vertically. Any large trees should be re-moved from the upslope areas for a horizontal dis-tance approximately 1.5 times the height of the cut. Removal of the trees promotes the movement of the organic mat downslope and over the face of


Fintied 1001 :4 being used to f [ stabilize slope

ice-rich permafrost.

F inned heat pipes

J•'• :

the cut without tearing. The natural draping of the organic mat over the cut face acts as an insu-lator and promotes reestablishment of vegetation. In cuts higher than 1.5 m, a reinforcing mesh should be applied to the surface of the organic mat to provide it with an adequate tensile strength. Areas with thin organic cover may have to be de-signed so that they can slough without impeding the transportation facility, or they may be de-signed as Type IV cuts (see below) when ice-rich soils are involved.

5.3.2 Type II

Type II cuts are in frozen ice-poor soil or bedrock. Such cuts can be designed so that the slope angle is compatible with the thawed (unfrozen) proper-ties of the materials. Higher rates of surface ero-sion can be expected because of severe freeze-thaw conditions and slope establishment of vegetation.

5.3.3 Type III

Type III cuts include any made in a terrain in which ground ice occurs as a matrix of ice wedges or poly-

gons in otherwise relatively ice-poor mineral soil. These cuts should be vertical if possible, since fewer ice wedges are likely to be encountered. If the ice wedges are relatively widely spaced and a thick or-ganic mat is present, the cuts may self-heal as with Type I cuts. Considerable water may be produced during the melting of the ice wedges and the self-healing process. Revetments may be required along the roe of slope to stabilize it and allow the water to drain. Wider-than-normal ditches may be required to provide space for accumulations of slope debris, the revetments, and drainage. Since this extra width increases the amount of excavation required, a more economical alternative may be excavation of the ice wedges over some reasonable distance based on geocryologic considerations and backfill-ing with selected materials.

5.3.4 Type IV

Type IV cuts include those in ice-rich permafrost soils, which are the most difficult cuts to design and manage. These cuts may be subject to contin-uous exposure, melting, and retreat of the cut

Permafrost

643

faces, essentially forming bimodal flows. Generally such slopes can only be stabilized by providing an adequate insulation cover to prevent, or at least reduce, the thawing process. The insulation must be free-draining and sufficiently flexible to accom-modate thaw settlements; hence mixtures of sand, gravel, crushed rock, and artificial insulation ma-terials are generally used.

Considerable experience is required to recog-nize terrain conditions and recommend appropri-ate cut-slope design responses. Conditions pro-moting the self-healing of cuts are often found in tills, alluvial or colluvial silts and sands, and bed-rock. Conditions causing non-self-healing cuts are more common in continuous permafrost zones, ice-rich soils, and glaciolacustrine sediments. A clear understanding of the conditions within a particular cut may not emerge until the actual ground-ice conditions are exposed during con-struction. Inspection of adjacent natural slopes is often the best guide to the expected performance of cut slopes at any location.

REFERENCES

Andersland, O.B., and D.M. Anderson. 1978. Geo-technical Engineering for Cold Regions. McGraw-Hill, New York, 551 pp.

ASTM. 1993. Annual Book of ASTM Standards, Vol. 04.08: Soil and Rock; Dimension Stone; Geosyn-thetics, Philadelphia, Pa., pp. 617-621.

Bates, R.L., and J.A. Jackson, eds. 1980. Glossary of Geology, 2nd ed. American Geological Institute, Falls Church, Va., 751 pp.

Bird, J.B. 1967. The Physiography of Arctic Canada. Johns Hopkins Press, Baltimore, Md.

Brown, R.J.E. 1970. Permafrost in Canada: Its Influence on Northern Development. University of Toronto Press, Toronto, Ontario, Canada, 246 pp.

Brown, R.J.E., and T.L. Péwé. 1973. Distribution of Permafrost in North America and Its Relation-ship to the Environment: A Review, 1963-1973. In Proc., Second International Conference on Per-mafrost, Yakutsk, USSR, North American Con-tribution, National Academy of Sciences, Washington, D.C., pp. 7 1-100.

Brown, R.J.E., G.H. Johnston, J.R. MacKay, N.R. Morgenstem, and W.W. Shilts. 1981. Permafrost Distribution and Terrain Characteristics. In Permafrost: Engineering Design and Construction (G.H. Johnston, ed.), Associate Committee on

Geotechnical Research, National Research Council of Canada, John Wiley & Sons, Toronto,

pp. 3 1-72. Casagrande, A. 1931. Discussion of Benkelman, A.C.,

and P.R. Olmstead, A New Theory of Frost Heaving, HRB Proc., Vol. 11, Part 1, pp. 168-172.

Chyurlia, J. 1973. Stability of River Banks and Slopes along the Liard River and Mackenzie River N.W.T. Report 73-3. Environmental-Social Program, Northern Pipelines, Department of Indian and Northern Affairs Canada, Ottawa, pp. 111-152.

Code, J.A. 1973. The Stability of Natural Slopes in the Mackenzie Valley. Report 73-9. Environmental-Social Program, Northern Pipelines, Department of Indian and Northern Affairs Canada, Ottawa, 18 pp.

Eyles, N. 1983. Glacial Geology: An Introduction for Engineers and Earth Scientists. Pergamon Press, Elmsford, N.Y., 409 pp.

Finn, W.D., R.N. Yong, and K.W. Lee. 1978. Lique-faction of Thawed Layers in Frozen Soil. Journal Geotechnical Engineering Division, ASCE, Vol. 104, No. GT1O, pp. 1243-1255.

Hardy, R.M., and H.A. Morrison. 1972. Slope Sta-bility and Drainage Considerations for Arctic Pipelines. In Proc., Canadian Northern Pipeline Research Conference, Technical Memorandum 104, National Research Council of Canada, Ottawa, pp. 249-267.

Harris, C. 1987. Mechanics of Mass Movement in Periglacial Environments. In Slope Stability (M.G. Anderson and K.S. Richards, eds.), John Wiley and Sons, New York, pp. 53 1-559.

Hughes, O.L. 1972. Surficial Geology and Land Classification, Mackenzie Valley Transportation Corridor. In Proc., Canadian Northern Pipeline Re-search Conference, Technical Memorandum 104, National Research Council of Canada, Ottawa, pp. 17-24.

Hunter, J.A., A.S. Judge, H.A. Macauley, R.L. Good, R.M. Gagne, and R.A. Burns. 1976. Permafrost and Frozen Sub-seabottom Materials in the Southern Beaufort Sea: Beaufart Sea Project. Technical Report 22. Environment Canada, Ottawa, 177 pp.

Isaacs, R.M., and J.A. Code. 1972. Problems in Engineering Geology Related to Pipeline Con-struction. In Proc., Canadian Northern Pipeline Research Conference, Technical Memorandum 104, National Research Council of Canada, Ottawa, pp. 147-178.

Jahn, A. 1975. Problems of the Pen glacial Zone. PWN-Polish Scientific Publishers, Warsaw, 223 pp. NTIS: PB 248 901.

644


Jessop, A.M. 1971. The Distribution of Glacial Perturbation of Heat Flow in Canada. Canadian Journal of Earth Sciences, Vol. 8, No. 2, pp. 162-166.

Jessberger, H.L. 1970. Ground Frost: A Listing and Evaluation of More Recent Literature Dealing with the Effect of Frost on the Soil. Report FSTC-HT-23-311-70 translated by U.S. Army Foreign Science and Technology Center, Charlottesville, Va., 494 pp.

Johnson, A.W. 1952. Special Report 1: Frost Action in Roads and Ai'ifields: A Review of the Literature. HRB, National Research Council, Washington, D.C., 300 pp.

Johnston, G.H., ed. 1981. Permafrost: Engineering Design and Construction. Associate Committee on Geotechnical Research, National Research Council of Canada, John Wiley & Sons, Toronto, 540 pp.

Johnston, G.H., B. Ladanyi, N.R. Morgenstern, and E. Penner. 1981. Engineering Characteristics of Frozen and Thawing Soils. In Permafrost: Engi-neering Design and Construction (G.H. Johnston, ed.), Associate Committee on Geotechnical Research, National Research Council of Canada, John Wiley & Sons, Toronto, pp. 73-147.

Judge, A.S. 1973. Deep Temperature Observations in the Canadian North. In Proc., Second Inter- national Conference on Permafrost, Yakutsk, USSR, North American Contribution, National Aca-demy of Sciences, Washington, D.C., pp. 35-40.

Krzewinski, T.G., and R.G. Tart. 1985. Thermal Design Considerations in Frozen Ground Engineer-ing. American Society of Civil Engineers, New York, N.Y., pp. 42-43.

Ladanyi, B., and G.H. Johnston. 1978. Field Investi-gations of Frozen Ground. In Geotechnical Engi- neering for Cold Regions (O.B. Andersland and D.M. Anderson, eds.), McGraw-Hill, New York, N.Y., pp. 459-504.

Linnell, K.A., and C.W. Kaplar. 1959. The Factor of Soil and Material Type in Frost Action. Bulletin 225, HRB, National Research Council, Washing-ton, D.C., pp. 81-1 26.

Linnell, K.A., and C.W. Kaplar. 1963. Description and Classification of Frozen Soils. In Proc., First International Conference on Permafrost, Lafayette, md., Publication 1287, National Academy of Sciences, Washington, D.C., pp. 481-487.

Lotspeich, F.B. 1971. Environmental Guidelines for Road Construction in Alaska. Environmental Protection Agency, College, Alaska.

Lunardini, V.J. 1978. Theory of n-Factors and Correla-tion of Data. In Proc., Third International Con-ference on Permafrost, Edmonton, Alberta, National Research Council of Canada, Ottawa, Vol. 1, pp. 41-46.

MacKay, J.R. 1966. Segregated Epigenetic Ice and Slumps in Permafrost, Mackenzie Delta Area, N.W.T. Geographical Bulletin, No. 8, pp. 59-80.

MacKay, J.R. 1972. Offshore Permafrost and Ground Ice, Southern Beaufort Sea, Canada. Canadian Journal of Earth Sciences, Vol. 9, No. 11, pp. 1550-1561.

MacKay, J.R., and W.H. Mathews. 1973. Geomorph-ology and Quaternary History of the Mackenzie River Valley near Fort Good Hope, N.W.T. Cana-dian Journal of Earth Sciences, Vol. 10, No. 1, pp. 26-41.

McPhail, J.F., W.B. McMullen, and A.W. Murfitt. 1976. Yukon River to Prudhoe Bay: Lessons in Arctic Design and Construction. Civil Engineering (New York City), Vol. 46, No. 2, pp. 78-82.

McRoberts, E.C. 1973. Stability of Slopes in Perma-frost. Ph.D. Thesis. University of Alberta, Ed-monton, Alberta, Canada.

McRoberts, E.C. 1978. Slope Stability in Cold Regions. In Geotechnical Engineering for Cold Regions (O.B. Andersiand and D.M. Anderson, eds.), McGraw-Hill, New York, N.Y., pp. 363-404.

McRoberts, E.C., and N.R. Morgenstern. 1973. Landslides in the Vicinity of the Mackenzie River, Mile 205 to 660. Report 73-35. Environmental-Social Program, Northern Pipelines, Department of Indian and Northern Affairs Canada, Ottawa, 96 pp.

McRoberts, E.C., and N.R. Morgenstern. 1974a. The Stability of Thawing Slopes. Canadian Geo-technical Journal, Vol. 11, No.4, pp. 447-469.

McRoberts, E.C., and N.R. Morgenstern. 1974b. Stability of Slopes in Frozen Soil, Mackenzie Valley, N.W.T. Canadian Geotechnical Journal, Vol. 11, No. 4, pp. 554-573.

Muller, S.W. 1947. Permafrost or Permanently Frozen Ground and Related Engineering Problems. J.W. Edwards, Inc., Ann Arbor, Mich., 231 pp.

National Academy of Sciences. 1983. Fourth International Conference on Permafrost: Final Proceedings, Fair-banks, Alaska, National Academy Press, Washing-ton, D.C., 413 pp.

National Research Council of Canada. 1988. Glossary of Permafrost and Related Ground-Ice Terms. Technical Memorandum 142. Ottawa, Ontario, Canada.

Permafrost

645

Nixon, J.F. 1992. New Frost Heave Prediction Method for Design of Northern Pipelines. In Proc., Second International Offshore and Polar Engineering Conference, International Society of Offshore and Polar Engineers, Golden, Cob., pp. 32-39.

Nixon, J.F., and G. Lem. 1984. Creep and Strength Testing of. Frozen Saline Fine Grained Soils. Canadian GeotechnicalJourrial, Vol. 21, pp. 518-529.

Osterkamp, T.E., and W.D. Harrison. 1976. Subsea Permafrost: Its Implications for Offshore Re-source Development. The Northern Engineer, Vol. 8,No. 1, pp. 31-35.

Pihlainen, J.A., and G.H. Johnston. 1963. Guide to a Field Description of Permafrost. Technical Memor-andum 79. Associate Committee on Soil and Snow Mechanics, National Research Council of Canada, Ottawa, 23 pp.

Pufahi, D.E. 1976 The Behaviour of Thawing Slopes in Permafrost. Ph.D. Thesis. University of Alberta, Edmonton, Alberta, Canada, 345 pp.

Pufahl, D.E., NR. Morgenstern, and W.D. Roggen-sack. 1974. Observations on Recent Highway Cuts in Permafrost. Report 74-32. Environmental-Social Program, Northern Pipelines, Department of Indian and Northern Affairs Canada, Ottawa,

53 pp. Samson, L., and F. Tordon. 1969. Experiences with

Engineering Site Investigations in Northern Quebec and Northern Baffin Island. In Proc., Third Canadian Conference on Permafrost, Tech-

nical Memorandum 96, Associate Committee on Geotechnical Research, National Research Council of Canada, Ottawa, pp. 2 1-38.

Sellmann, P.V., and D.M. Hopkins. 1983. Subsea Permafrost Distribution on the Alaskan Shelf. In Fourth International Conference on Permafrost: Final Proceedings, Fairbanks, Alaska, National Academy Press, Washington, D.C., pp. 75-82.

Smith, N., and R. Berg. 1973. Encountering Massive Ground Ice During Road Construction in Cen-tral Alaska. In Proc., Second International Confer-ence on Permafrost, Yakutsk, USSR, North American Contribution, National Academy of Sciences, Washington, D.C., pp. 730-736.

Taber, S. 1943. Perennially Frozen Ground in Alaska. Geological Society of America Bulletin, Vol. 54, pp. 1433-1548.

Walker, D.B.L., D.W. Haley, and A.C. Palmer. 1983. The Influence of Subsea Permafrost on Offshore Pipeline Design. In Proc., Fourth International Conference on Permafrost, Fairbanks, Alaska, National Academy Press, Washington, D.C., pp. 1338-1343.

Walker, H.J., and L. Arnborg. 1963. Permafrost and Ice Wedge Effects on River Bank Erosion. In Proc., First International Conference on Permafrost, Lafayette, md., Publication 1287, National Acad-emy of Sciences, Washington, D.C., pp. 164-171.

Washburn, A.L. 1973. Pen glacial Processes and Environments. Edward Arnold Publishers, Lon-don, England, 320 pp.

permafrost - transportation research boardonlinepubs.trb.org/onlinepubs/sr/sr247/sr247-025.pdf ·...

Documents