Numerical Simulation of F rost Heave in Soils around the Buried Oil

Pipeline in Island Talik Permafrost Region

Zaiguo Fu1, 2 , Yu Zhao1 , Bo Yu1, * , Yasuo Kawaguchi2
1: National Engineering Laboratory for Pipeline Safety, Beijing Key Laboratory of Urban Oil and Gas
Distribution Technology, China University of Petroleum, Beijing, China
2: Department of Mechanical Engineering, Tokyo University of Science, Chiba, Japan
* Correspondent author: yubobox@vip.163.com
Abstract
Differential frost heave and thaw settlement in soils around the buried pipeline in permafrost regions are the
main sources threatening the safe operation of pipelines. To analyze the freezing characteristics and frost
heave in the soils around a buried oil pipeline in island talik permafrost region, a two-dimensional
computational model of the soil temperature fields was established based on the heat transfer process with
phase change. The classic segregated potential frost heave model was adopted to forecast the amount of frost
heave. The calculation parameters were derived from the China-Russia Crude Oil Pipeline (CRCOP)
engineering. We solved these models by using numerical methods and obtained the freezing characteristics
of the soils around the pipeline under the effects of various factors within the pipeline operation life cycle.
The developments of the frost heaves in four typical sections of island talik were predicted accordingly. The
results indicated that the annual maximum freezing circle around the pipe line narrowed with the elapse of
time in operation. The maximum frost penetration under the pipe decreased with the increase of oil
temperature, thickness of thermal insulation layer and water content of soils. The largest amount of frost
heave of the soils under the pipe line emerged in the second year after the start of its operation. The
maximum amounts of frost heave in the sections of cold island talik with 35% and 55% water contents
around the pipeline without thermal insulation layer were larger than those in the sections with lower water
contents.
Keyword: buried pipeline, island talik, phase change, frost heave

1. Introduction
The buried oil pipeline may face huge challenges for safe operation due to severe frost heave and thaw
settlement deformations which are induced by the freezing and thawing process of soils around the pipe
transporting oil with varying temperatures [1]. The freezing and thawing process is also affected by the
change of the atmospheric environment. In permafrost region, the island-shaped section of soils containing
perennial non-frozen water, resulted from the outside disturbance or the consisting of groundwater
passageway, is called island talik permafrost region [2]. The soil in island talik is susceptible to the
disturbance imposed by the environment and buried oil pipeline after the start of its operation. The watercontained soil might be frozen which can lead to differential frost heave and soil deformation and finally can
destroy the pipeline structure [3].
In order to obtain the freezing and thawing characteristics of soils in permafrost region, the complicated
heat and mass transfer process including moisture transfer, heat conduction and phase change has been
widely studied experimentally and numerically [4, 5]. The experimental studies mainly consisted of the
measurements of soil temperature, water content of soil, mechanical characteristics of soils and the moistureheat coupling process. But there are no direct experimental studies on soil freezing and heaving under the
thermal effect of a buried pipeline in permafrost regions. On the other hand, the theoretical and numerical
studies usually concentrated on the moisture-heat-stress coupling model of soils and tried to obtain the
moisture, thermal and stress parameters simultaneously [6, 7]. But it is difficult to clearly describe the
various thermodynamical and mechanical parameters of soils and their interrelationship in theory. These
pioneering studies can provide basic information for determining the freezing and thawing characteristics of
1

soils in permafrost region. However, they are not sufficient and convenient for calculating the frost heave in
soils around an actual buried pipeline engineering.
Furthermore, in order to predict the amount of the soil deformation and the danger level for the buried
pipeline in permafrost regions, many works related to the thaw settlement and frost heave in soils around the
buried pipeline have also been carried out. Konrad et al. [8] proposed a procedure to calculate the frost heave
based on a finite difference formulation of heat and mass transfer in saturated soils and on a segregation
potential concept. Oswell [9] conducted several large-scale frost heave tests and pointed out that soil frost
heave under pipeline depends on three conditions: the freezing temperature in unfrozen soils, a source of
water to migrate to the freezing front and the frost susceptible soil that permit free water to migrate. Palmer
et al. [10] examined the interaction between soil frost heave and upheaval buckling of pipeline from the
aspect of mechanics. Konrad et al. [11] simulated the frost action including the stress-strain-temperature
relationship using a two-dimensional segregation potential method. Selvadurai [12, 13] established a threedimensional finite element model of the interaction between buried pipeline and soil based on the simplified
moisture-heat coupling effect. It focused on the induced bending stress on pipe caused by frost heave. These
studies provide valuable information for predicting the soil deformation and understanding its effects on
pipeline.
Recently, owing to the design and operation of China-Russia Crude Oil P ipeline (CRCOP), some
institutions in China such as the cold and arid regions environmental and engineering research institute of
Chinese academy of sciences have also performed certain research related to the soil deformation around the
buried pipeline in permafrost regions. The researchers conducted geological survey and investigations of the
temperature distribution of soils and transported oil along the pipeline. Among them, Zhang et al. [14] used
the finite element method to estimate the thermal effects of the CRCOP pipeline. The results indicated that
the thaw of permafrost under the pipeline cannot be prevented. Li et al. [15] analyzed the freezing and
thawing process of the soils around the buried pipeline by numerical simulation. They concluded that the
insulation can restrain the heat transfer between pipeline and soils during the early operating period. Ji et al.
[16] measured the frost heaving features of soils in typical geomorphologic units along the pipeline and
found that the frost heaving ratio of the silty clay with humus was the maximum one among the tested
samples. These findings provide a great deal of useful information for understanding basic situations of
CRCOP engineering.
However, it is worth mentioning that, owing to the complicated atmospheric environment, various
compositions of soils and different engineering background, great differences exist in the situations of
permafrost region (including the island talik) and pipeline which are located in various geographical
locations. Accordingly, different thermodynamical and mechanical models are required for different actual
pipeline projects. With respect to CRCOP, nobody has performed comprehensive analyses of the freezing
and heaving rules of the soils in island talik under the influences of different oil temperatures, different
thicknesses of thermal insulation layer and different water contents of soils, as far as we know. Thus, a
systematic method to obtain more accurate and reliable results of soil frost heave for CRCOP is very
necessary.
Against this background, the freezing characteristics of soils around the buried pipeline in four typical
sections of island talik with various oil temperatures, thicknesses of thermal insulation layer and water
contents were investigated quantitatively on the basis of the simulation of unsteady soil temperature fields.
The finite thermal effect domain and the equivalent heat capacity method were adopted to analyze the heat
transfer process with phase change. The amounts of soil frost heave were calculated by the segregated
potential frost heave model based on the obtained frost penetrations and temperature gradients. The research
provides an effective method for calculating the freezing size and frost heave in soils around the buried
pipelines in island talik permafrost regions.
2. Modeling
As described in our previous research [17] for forecasting the thaw settlement of soils , the heat transfer of
soils around the buried oil pipelines in permafrost regions including the island talik is a complicated process
involving phase change, moisture migration and coupled deformation. To simplify the calculation, some
assumptions are made in the modelling as follows: (1) the replaced soil in the pipe ditch, the pipe wall and
the coating have no effects on the long-term soil temperature fields; (2) the oil temperature on the crosssection of pipe is uniform; (3) the soil is isotropic and the heat conduction of soil along the direction of pipe
axis is ignored; and (4) the effects of moisture transfer on the long-term soil temperature fields are also
ignored [18].
2

2.1 Geometric model

Some researchers adopted a finite thermal effect domain for simulating the soil temperature fields around the
buried pipeline in non-permafrost regions [19]. For the pipeline in island talik permafrost region, the similar
method was adopted to determine the appropriate size of the thermal effect domain. The sketch of the
geometric model of the buried oil pipeline in permafrost region is presented in Figure 1. Half of the thermal
effect domain was chosen to be the computational domain. According to the results of geological survey for
CRCOP [20], the unilateral horizontal distance was set to be 15 m for the buried pipeline with a large
diameter. The vertical depth was set to be 20 m due to the stable year-round temperature gradient at the
location. The soil around the pipeline was divided into three layers by the soil types in the domain. The first
layer at the depth of 3 m below the ground was composed of sand loam. The second layer at the depth of 310 m consisted of silty clay mainly. The third layer at the depth of 10-20 m was composed of bedrock
mainly. The nominal buried depth and diameter of the pipe were 1.5 and 0.813 m respectively. Partial
sections of pipeline were wrapped with thermal insulation layer of 80 mm thickness.

FIGURE 1: Sketch of the buried oil pipeline in permafrost region.

2.2 Governing equations
As introduced in our previous research [17], the equivalent heat capacity was adopted to replace the latent
heat of phase change according to the temperature regime of soil. The governing equation of heat conduction
is written as

cp

T
T

T
(
) (
),
t x x
y y

(1)

where c p denotes the heat capacity of soil and is the thermal conductivity according to the temperature
condition and the corresponding freezing-thawing state. They are expressed as follows when T is between
the temperature of fully frozen T1 and the temperature of phase transition T2 .

cp c f

cu c f
T2 T1

= f

(T T1 )

u f
T2 T1

L W
1 W T

(T T1 )

(2)

(3)

where, c f and cu are heat capacities of frozen and unfrozen soil, respectively, f and u are thermal
conductivities of frozen and unfrozen soil, respectively, L is the latent heat in ice-water phase transition.
3

T is a small temperature range, W is the total water content and W is the change of frozen water
content in a small temperature range.
The heat conduction equation of the thermal insulation layer outside the pipe wall in the polar coordinate
can be written as

*c p *

T 1 * T
1 * T

( r
) 2
(
),
r r
r
r

(4)

where, the superscript * denotes the properties of the insulation material.

The segregated potential frost heave model [21, 22] adopted in this research is given as Eq. (5). The frost
heave was assumed to be composed of segregation frost which is induced by water migrating and freezing
from outside, and of in-situ frost causing by freezing of pore water. The former frost makes volume of soils
increase approximate 1.09 times. Meanwhile, the latter one makes volume increase about 0.09 times.

H h f 0.09Z 1.09 SP0 gradT (t ) t 0.09Z

(5)

where H is the value of frost heave, h f is segregation frost and SP0 is segregated potential which can
be calculated by SP0 a exp(bPe ) , a and b are the parameters related to the soil type , Pe is external
physical loads, gradT (t ) is the temperature gradient at a given time, t is freezing time, is porosity of
soil, Z is the displacement of the freezing front in time t .
2.3 Boundary conditions
The method for setting the boundary conditions is similar to that used in our previous research [17].
However, the parameters are significantly changed due to the changed state of soils in island talik. For the
upper boundary of the model of soil temperature fields, it is assumed to be the first boundary condition
because the annual variation of the ground surface temperature is found to be regular based on the geological
survey. It is easy to describe it by a periodic function. In contrast, it requires a complicated function to
describe the ambiguous convective heat transfer coefficient if we set the boundary to be the second boundary.
Considering the atmospheric temperature has risen by 2.4 K in the last 50 years, the periodic function is
written as

f (t ) T ' A sin(

2 t
2.4
)
t,
360 2 50 360

(6)

where, T ' is the annual mean surface temperature, A is the annual amplitude of the change of the surface
temperature, t is time. Two representative functions were obtained by regression analysis using least square
method for the surface temperature in island talik permafrost region along the CRCOP. The average
temperatures were 274.15 and 275.15 K with both amplitudes being 19 K, respectively. The corresponding
regional soils were defined as low-temperature (cold) and high-temperature (hot) unfrozen soils,
respectively.
The left boundary not including pipe wall is a symmetrical boundary, while the right boundary is assumed
to be not thermally affected by the buried pipe. They are formulated as

T
0.
x

(7)

The boundary of pipe wall in the model is assumed to be the first boundary condition. The temperature is
same as the uniform oil temperature based on the aforementioned assumption. It is described by a sinusoidal
function obtained by regression analysis via the least square method as

T 272.65 5.2sin(2 t / 360 3) ,

(8)

where, /3 is the initial phase. This phase differs from that in the periodic function of the surface temperature.
It indicates that the change of oil temperature lags behind that of the surface temperature.
The lower boundary of the model is assumed to be the second boundary condition according to the yearround stable temperature gradient there via geological survey [20]. This boundary is described as a constant
heat flux as Eq. (9). The stable temperature gradient at the location of 20 m below the surface is -0.04 K/m
along the section of CRCOP in China [23].

T
(9)

2.4 Other parameters

The basic thermophysical parameters of various materials involved in the model of soil temperature fields
are set as same as those in our previous research [17]. The other thermophysical parameters and equivalent
heat capacities of the silty clay with various water contents are referenced from [24]. The equivalent heat
capacities of the three types of soils with the average water contents in various layers corresponding to
different temperature ranges are shown in Table 1. For calculating the segregated potential during the
freezing process, the parameters a and b of the sand loam around the pipe are set to be 1.4510-8 m2 /(sK)
and 0.01510-3 Pa-1 , respectively. The porosity of soil is assumed to be 10% generally [2].
TABLE 1: Equivalent heat capacities (J/(kgK)) of soils with average water contents [24].
Temperature regimeK
Soils

263.15~253.15

268.15~263.15

270.15~268.15

271.15~270.15

272.15~271.15

272.65~271.15

272.95~272.65

273.15~272.95

275.15~273.15

sand loam
silty clay
bedrock

982
1158
982

1156
1693
1476

1782
2650
2364

3367
6727
3658

5578
6758
6160

18699
12142
16081

30353
37137
39563

66724
68372
1267

1273
1466
1272

3. Numerical Method
3.1 Grids
Figure 2 shows the grids of partial computational domain near the buried pipe. The unstructured triangular
grids in Cartesian coordinate are generated for the soil area automatically via the Delaunay triangulation
method and the structured quadrilateral grids in polar coordinate are generated for the thermal insulation
layer area. Since the temperature gradient is greater in the region near the pipe, denser grids are generated in
the region close to the pipeline without overlapping each other. There are 4570 triangular elements in total.
Frame 001 17 May 2011 fluent6.3.26

-1
-1.2
-1.4
-1.6

-1.8
-2

-2.2
-2.4
-2.6
-2.8
-3

0.5

1.5

FIGURE 2: Grids of computational domain around the buried pipe.

3.2 Establishment of initial temperature fields
The initial temperature fields can be used to verify the model and calculate unsteady soil temperature fields
further. When we calculated the initial temperature fields, the climate warming was ignored in the surface
temperature function of Eq. (6) and the boundary of pipe wall was assumed to be adiabatic.
5

-5

-5

Depth (m)

Depth (m)

The widely applied commercial software FLUENT was adopted in this study. This numerical method can
be easily used to practical engineering application. Several user defined functions were written and inserted
to describe the thermophysical parameters of various types of soils and the time-dependent boundary
conditions. The heat conduction equation was discreted by the second order upwind scheme and the time
step was chosen as 86400 s. The calculation of initial temperature fields had lasted for a period of 50-100
years until the soil temperature reached constant at the same position in the area with seasonal varying
temperature at the same time in continuous years, meanwhile, the temperature fields in other areas of the
computational domain remained stable.
The section of CRCOP in China was put into operation in autumn, and one week was needed to fill up the
pipeline with oil. It is difficult to determine the exact initial time for different typical sections along the
pipeline. However, as long as the initial boundary conditions are guaranteed to correspond with the
corresponding time, the initial temperature fields of any time in one year do not affect the prediction results
in the long-term operation of pipeline. In the current study, July 15th, the day in the hottest month was
selected as the initial time for the convenience of calculation. The initial temperature fields of soils on the
starting date of the pipeline are shown in Figure 3. Figure 3(a) corresponds to the hot unfrozen soils with
temperature from 275.15 to 280.15 K under pipeline and Figure 3(b) represents the cold unfrozen soils with
average temperature being 274.65 K. These features also agree well with those of the temperature
distribution in island talik permafrost regions introduced in the classic literature [2].

-10
297.00
292.00

-10
295.00
288.00

290.00

286.00

286.00

284.00

284.00

-15

282.00

-15

282.00

280.00

280.00

278.00

278.00

276.00
274.00

276.00

-20

275.00

10

-20

15

Horizontal distance (m)

273.00

10

15

Horizontal distance (m)

(a)
(b)
FIGURE 3: Initial temperature fields of: (a) hot unfrozen soils and (b) cold unfrozen soils in island talik
permafrost region.

269.15 273.15 277.15 281.15 285.15 289.15 293.15 297.15

0

Temperature (K)

-2
-4

Measurement
Calculation

Depth (m)

-6
-8
-10
-12
-14
-16
-18
-20

FIGURE 4: Distribution of soil temperature in different depths near Jia Station

6

3.3 Validation
The temperature distribution of the plumb line that is 10 m of horizontal distance from the central axis of the
pipeline in the computational domain was chosen as the evaluation object. Figure 4 presents the calculation
results along with the actually measured temperatures in different depths in an island talik near Jia station
along CRCOP in July of summer. It can be found that the calculation results of the natural temperature
distribution of soils are in good agreement with the measured values with the maximum deviation being not
more than 1.5 K. This validates the accuracy and reliability of the model and numerical calculation.
4. Results and Discussions
According to the water-bearing characteristics of soils in island talik permafrost region along the section of
CRCOP in China, the four typical talik sections of 20%, 25%, 35% and 55% water contents in soils were
selected as research objects to analyze the freezing characteristics and to calculate the amounts of frost
heave. The temperature fields of hot and cold unfrozen soils as established in Section 3.2 were adopted as the
initial temperature fields. The oil temperatures in operation were considered to be constant (267.15, 269.15
and 271.15 K) and varying in a sinusoidal pattern as expressed in Eq. (8). The thicknesses of thermal
insulation layer were assumed to be 0, 10, 80 and 120 mm respectively. The developing processes of the
annual maximum freezing circles (isothermal line of 273.15 K, freezing interface) around the buried pipeline
under the conditions of above various parameters were simulated and analyzed to determine the change of
the maximum frost penetration and frost heave during the pipeline operation life cycle of 50 years.
4.1 Freezing characteristics
The developments of the annual maximum freezing circles and frost penetrations emerging in typical years
within the pipeline operation life cycle under different factors such as oil temperature were investigated.
Figure 5 presents the results of the initial cold unfrozen soils of 20% water content around the pipe line
without heat preservation under the conditions of 267.15 K and periodic oil temperature. In this study, we
also performed numerical calculations under other conditions to compare the freezing characteristics. Figure
6 shows the variations of the annual maximum frost penetrations under the pipeline with different oil
temperatures, different thicknesses of thermal insulation layer and different water contents during the
operation period, respectively.

(a)
(b)
FIGURE 5: Development of the annual maximum freezing circle around the pipeline with different oil
temperature: (a) 267.15 K and (b) periodic.
From Figure 5, it can be seen that the annual maximum freezing circle around the pipeline in island talik
permafrost region narrows as time elapses when the pipeline is operated with negative and periodic oil
temperatures. The variation trend of the freezing circle is smooth during the operation period. The size of the
circle with 267.15 K oil temperature is larger than that with the periodic oil temperature. Moreover, the
annual maximum freezing circles with 269.15 and 271.15 K oil temperatures are also found to be smaller
7

than that with 267.15 K oil temperature in the other same situations based on our simulation (not shown here,
due to the limited space). It indicates that the lower the oil temperature is, the larger size of the annual
maximum freezing circle can be obtained. In addition, with the other same parameters, the annual maximum
freezing circle of soils around the buried pipe with 80 mm thermal insulation layer is found to be much
smaller than that of the case without heat preservation. The heat exchange between soils and the pipe can be
suppressed by the thermal insulation layer obviously. The thicker the layer is, the harder the unfrozen soil
around the pipeline turns to be freezing. The annual maximum freezing circle of the soil with 20% water
content is found to be slightly larger than that of the 25% water content case with the other same parameters
due to the smaller heat capacity of soils with lower water content.

1.6

271.15 K
269.15 K
267.15 K
periodic

Frost penetration (m)

0.8
0.0
-0.8
-1.6
-2.4
-3.2
0

10

15

20

25

30

35

40

45

50

Time (year)

(a)
1.2

0 mm
40 mm
80 mm
120 mm

Frost penetration (m)

0.8
0.4
0.0
-0.4
-0.8
-1.2
0

10

15

20

25

30

35

40

45

50

Time (year)

(b)
-0.4

20%
25%
35%
55%

Frost penetration (m)

-0.6
-0.8
-1.0
-1.2
-1.4
-1.6
0

10

15

20

25

30

Time (year)

(c)
8

35

40

45

50

FIGURE 6: Variation of annual maximum frost penetration under pipeline with different: (a) oil
temperatures, (b) thicknesses of thermal insulation layer, and (c) water contents in soils.
From Figure 6(a), it can be seen that the maximum frost penetration emerges in the second year after
operation as 2.8 m under the pipeline with the oil temperature of 267.15 K for the initial cold unfrozen soil of
20% water content. In contrast, the 271.15 K oil temperature has less effect on the freezing of soil. With the
periodic oil temperature as described in Eq. (8), the freezing and thawing of soils around the pipe line may
develop alternately because of the positive and negative effect of the transported oil. Thus, the maximum
frost penetration in the second year under this condition as 1.34 m is between those of the 271.15 K and
269.15 K cases. Furthermore, the thermal insulation layer seems to have significant effects on the frost
penetration as shown in Figure 6(b). The annual maximum frost penetrations are almost identical for the
cases with 80 and 120 mm thermal insulation layers. It indicates that an economical thickness of the
insulation layer exists for the application in CRCOP. However, they are smaller than that of the 40 mm case
due to the more effective heat insulation.
From Figure 6(c), we can see that the annual maximum frost penetration under the pipe decreases with the
increasing of water content. For the initial cold unfrozen soil of 55% water content, the maximum frost
penetration as 1.03 m also appears in the second year after operation of the pipeline with the periodic oil
temperature, which is 0.3 m smaller than that of the 20% water content case. In addition, according to the
features of the freezing circles shown in Figure 5, the maximum frost penetration at the position far away
from the buried pipe is found to decrease year by year due to the rise of the atmospheric temperature. These
results imply that close attentions should be paid to the probable large increase of the frost penetration in
island talik permafrost region when the weather or oil turns extreme cold at the project site and when the
thermal insulation layer is broken or soaked.
4.2 Frost heave in typical sections
In this section, the oil temperature is set to vary in a sinusoidal pattern as Eq. (8) according to the design
conditions of the section of CRCOP in China. Based on the monthly frost penetration and average
temperature gradient of the freezing front in the freezing season, the amounts of the monthly frost heave
under pipeline in three typical sections of island talik with 20%, 25% and 35% water contents in the second
year after operation were calculated for two initial temperature conditions of hot and cold unfrozen soils by
using Eq. (5). In particular, for the section with 55% water content, the designed thermal insulation layer was
ignored to predict the maximum frost heave of soils under the cold initial temperature for considering the
worst factors. Table 2 lists the monthly frost penetration and average temperature gradient of the freezing
front in the freezing season in the second year after operation for the four typical talik sections. Table 3
presents the monthly accumulated frost heave under the same conditions.
TABLE 2: Monthly frost penetration (m) and average temperature gradient (K/m) of the freezing front in
typical sections of island talik.
Temperature
condition

Month

hot

Nov.
Dec.
Jan.
Feb.
Mar.

20%
0.030/ * 8.51
0.31/ 4.57
0.62/ 0.79
0.91/ 0.72
1.07/ 0.00

Typical sections
25%
35%
0.038/8.37
0.054/9.28
0.031/5.04
0.42/ 4.54
0.63/ 2.95
0.75/ 2.00
0.89/ 0.80
0.99/ 0.23
1.03/ 0.00
1.08/ 0.00

Nov.
0.05/ 9.86
0.053/6.39
Dec.
0.40/ 5.07
0.40/ 4.86
Jan.
0.73/
2.17
0.74/ 1.96
cold
Feb.
1.06/ 0.21
1.02/ 0.05
Mar.
1.34/ 0.00
1.21/ 0.00
*The left side of the slash represents the value of frost penetration, while the right
average temperature gradient.

55%

0.053/9.26
0.058/13.08
0.31/ 5.54
0.32/ 7.16
0.62/ 2.94
0.65/ 1.25
0.91/ 0.86
0.91/ 0.00
1.03/ 0.00
1.03/ 0.00
side represents the value of the

According to the results shown in Table 2, the monthly frost penetration under the pipeline is found to
increase from the last November to March in the second year for all the typical sections due to the effects of
the decreasing oil temperature and atmospheric temperature. For the sections with 20% and 25% water
9

contents, the values of frost penetrations under cold initial temperature condition is larger than those under
the hot initial temperature condition. Nevertheless, the monthly frost penetrations for the section with 35%
water content under the hot and cold initial temperature conditions do not have the similar rules. Actually,
since the hot unfrozen soils of 35% water content have a larger thermal diffusing coefficient, the frost
penetration is slightly larger than that of the soils under cold initial temperature condition. For the sections of
island talik with 35% and 55% water contents under cold initial temperature condition, the monthly frost
penetrations are also quite close due to the almost uniform thermophysical parameters. Because their larger
heat capacities, the corresponding frost penetrations are slightly less than those of the sections with 20% and
25% water contents. The average temperature gradients of the freezing front are larger in the initial two
months. In addition, it can be seen in Table 2 that the average temperature gradients for the four sections
decrease month by month. It can be attributed to the decreasing temperature difference between the adjacent
soils during the heat transfer process.
TABLE 3: Monthly accumulated frost heave ( 10-2 m) under pipeline in typical sections of island talik.
Temperature
condition

hot

cold

Month
Nov.
Dec.
Jan.
Feb.
Mar.
Nov.
Dec.
Jan.
Feb.
Mar.

20%
09.90
20.00
21.80
23.40
23.60
11.40
22.40
26.90
27.60
27.80

Typical sections
25%
35%
10.00
12.20
21.70
23.70
28.20
28.60
30.10
29.30
30.20
29.40
7.70
12.20
18.80
26.30
23.10
33.50
23.50
35.70
23.60
35.80

55%

18.60
38.40
42.00
42.20
42.30

It can be seen in Table 3 that the amounts of the frost heave under the pipeline increase month by month in
the freezing season for all the typical sections of island talik. The frost heaves develop quickly in the first
three months from the freezing moment but smoothly in the later two months till March. This also indicates
the fast freezing induced by the large temperature gradient during the early operating period. The maximum
amount of the frost heave of the section with 20% water content is 0.278 m under conditions of cold initial
temperature and periodic oil temperature. It is larger than that under the condition of hot initial temperature.
For the section of island talik with 25% water content, the result under hot initial temperature is 0.302 m.
Moreover, the amounts of this case are larger than those under the cold initial temperature in the same
months due to the greater temperature gradients in such conditions. As the water content increases, the
largest amounts of the accumulated frost heaves of the sections with 35% and 55% water contents under the
cold initial temperature increase to 0.358 and 0.423 m, respectively. Thus, the relationship between the
maximum frost heave and water content of soils during the pipeline operation period is not simply linear
because of the combined effects of the temperature gradient at the freezing front and of the frost penetration
affected by the complicated process of heat transfer including phase change in soils.
In the design of the section of CRCOP in China, the allowable deformation of the pipeline structure in the
length of 40 m is suggested to be 0.25 m [24]. The obtained maximum frost heaves in the typical sections
exceed the critical value except for the soils of 20% water content under hot initial temperature and the soils
of 25% water content under cold initial temperature. In addition, the pipeline in the sections with more than
50% water contents is designed to be wrapped by a thermal insulation layer of 80 mm to protect the
structure. According to the forecast results, heat preservation for the buried pipeline in the island talik
permafrost region is strongly recommended. It is also worth mentioning that the used thermal insulation
layer can become invalid due to its damage and water immersion after operation. This might seriously impact
on the soils around the pipeline and cause freezing. Subsequently, a certain heave deformation may occur
under the pipeline and result in danger. The value of soil frost heave can be applied to further mechanical
calculation about the buried pipe. The research results will provide references for risk evaluation and site
management of the buried pipelines in island talik permafrost regions.
5. Conclusions
10

The model of the soil temperature fields around a large-diameter buried pipeline in the island talik
permafrost region was established with taking account of the phase change in the heat transfer process. The
characteristics of soil freezing and heaving were determined on the basis of the segregated potential frost
heave model and the CRCOP engineering. The annual maximum freezing circles and the maximum frost
heaves under the conditions of two initial soil temperature and other factors were obtained by numerical
calculation. The following conclusions were drawn from the present study.
(1) The annual maximum freezing circle around the buried oil pipeline narrows with the elapse of time. It
also increases with the decreases of oil temperature and water content of soils under the same other
operation conditions. The development of the annual maximum frost penetration has the same pattern.
The thermal insulation layer can effectively restrain the freezing of unfrozen soils around the pipeline in
island talik permafrost region. Close attentions should be paid to the damage and failure of the insulation
layer.
(2) The combined effects of temperature gradient and frost penetration result in a complicated relationship
between the maximum frost heave and water content in soils. The maximum amounts of frost heave
emerging in the sections of island talik with 35% and 55% water contents under cold initial temperature
are larger than those in the sections with 20% and 25% water contents. The maximum frost heaves may
exceed the allowable value of deformation except for the section with 20% water content under hot
initial temperature and the section with 25% water content under cold initial temperature.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
Acknowledgements
The support by the National Science Foundation of China (No.51176204; No.51325603) to this study is
acknowledged and highly appreciated. The first author would like to thank Japanese Ministry of Education,
Culture, Sports, Science and Technology for its support in the form of the scholarship.
References
[1] H. J. Jin, Design and construction of a large-diameter crude oil p ipeline in Northeastern China: A special issue on
permafrost pipeline, Cold Regions Science and Technology, vol. 64, no. 3, pp.209-212, 2010.
[2] X. Z. Xu, J. C. Wang, and L. X. Zhang, Physics of frozen soil, China: Science P ress, pp.2-56, 2010. (in Chinese)
[3] H. J. Jin and M. C. Brewer, Experiences and lessons learned in the engineering design and construction in the
Alaska Arctic, Journal of Glaciology and Geocryology, vol. 27, no. 1, pp.140-145, 2005. (in Chinese)
[4] R. L. Harlan, Analysis of coupled heat-flu id transport in partially fro zen soil, Journal of water Resources
Research, vol. 19, no. 5, pp.1314-1323, 1973.
[5] W. Kunio and W. To mo mi, Measurement of unfro zen water content and relative permittiv ity of frozen unsaturated
soil using NMR and TDR, Cold Regions Science and Technology, vol. 59, no. 1, pp.34-41, 2009.
[6] M. Shen and L. Bran ko, Modeling of coupled heat, mo isture and stress field freezing soil, Cold Regions Science
and Technology, vol. 14, no. 3, pp. 237-246, 1987.
[7] Z. G. Fu and B. Yu , Research on moisture-heat-stress fields of the soil around oil pipeline in permafrost region,
Oil & Gas Storage and Transportation, vol. 29, no. 8, pp. 565-570, 2010. (in Ch inese)
[8] J. M. Konard and N. R. Morgenstern, Frost heave prediction of chilled p ipelines buried in unfrozen soils ,
Canadian Geotechnical Journal, vol.21, no.1, pp.100-115, 1984.
[9] J. M. Oswell, Pipelines in permafrost: geotechnical issues and lessons , Canadian Geotechnical Journal, vol. 48,
no. 9, pp.1412-1431, 2011.
[10] A. C. Palmer and P. J. Williams, Frost heave and pipeline upheaval buckling, Canadian Geotechnical Journal,
vol.40, no.5, pp. 1033-1038, 2003.
[11] J. M. Konrad and M. Shen, 2-D frost action modeling using the segregation potential of soils , Cold Regions
Science and Technology, vol. 24, no. 3, pp. 263-278, 1996.
[12] A. P. S. Selvadurai, J. Hu, and I. Konuk, Co mputational modelling of frost heave induced soilpipeline interaction
I. Modelling of frost heave, Cold Regions Science and Technology, vol. 29, no. 3, pp. 215-228, 1999.
11

[13] A. P. S. Selvadurai, J. Hu, and I. Konuk, Co mputational modelling of frost heave induced soilpipeline interaction
II. Modelling of experiments at the Caen test facility , Cold Regions Science and Technology, vol. 29, no. 3,
pp.229-257, 1999.
[14] J. M. Zhang, G. Z. Qu, and H. J. Jin, Estimates on thermal effects of the China Russia crude oil pipeline in
permafrost regions, Cold Regions Science and Technology, vol. 64, no.3, pp.243-247, 2010.
[15] G. Y. Li, Y. Sheng, H. J. Jin et al., Develop ment of freezingthawing processes of foundation soils surrounding
the ChinaRussia crude oil p ipeline in the permafrost areas under a warming climate, Cold Regions Science and
Technology, vol. 64, no. 3, pp.226-234, 2010.
[16] Y. J. Ji, H. J. Jin, J. M. Zhang, and N. W. He, Experimental study of the frost-heaving ratio of the typical soil
samples along the China-Russia crude oil pipeline, Journal of Glaciology and Geocryology, vol. 30, no. 2, pp.296300, 2008. (in Chinese)
[17] Z. G. Fu, B. Yu, J. Zhu, and W. Li, Thaw characteristics of soil around buried pipeline in permafrost regions
based on numerical simulat ion of temperature fields , Journal of thermal science and Technology, vol. 7, no. 1,
pp. 322-333, 2012.
[18] S. X. Li and T. H. Wu, Permafrost temperature regime: study method and applied analysis , Journal of Glaciology
and Geocryology, vol. 26, no. 4, pp. 377-383, 2004. (in Chinese)
[19] X. G. Cui and J. J. Zhang, Determination of the thermal influence zone of buried hot oil p ipeline on steady
operation, Journal of University of Petroleum, China (Edition of Natural Science), vol. 28, no. 2, pp.75-78, 2004.
(in Chinese)
[20] G. Y. Li, H. J. Jin, Y. Sheng et al., Recent advances in frozen ground engineering geology survey along the China Russia Crude Oil Pipeline route(Mohe-Daqing section), Journal of Glaciology and Geocryology, vol. 30, no. 1,
pp.170-175, 2008. (in Ch inese)
[21] J. M. Konrad and N. R. Morgenstern, The segregation potential of a freezing soil, Canadian Geotechnical
Journal, vol.18, no.4, pp.482-491, 1981.
[22] J. F. Nixon, Field frost heave predictions using the segregation potential concept, Canadian Geotechnical
Journal, vol.19, no.1, pp.526-529, 1982.
[23] H. J. Jin, J. Q. Hao, X. L. Chang et al., Zonation and assessment of fro zen-ground conditions for engineering
geology along the ChinaRussia crude oil pipeline route fro m Mohe to Daq ing, Northeastern China , Cold
Regions Science and Technology, vol. 64, no. 3, pp. 213-225, 2010.
[24] Y. J. Ji, H. J. Jin, G. S. Wang, and J. M. Zhang, Thaw-stability assessment of the permafrost foundation soil along
the proposed China-Russia crude oil pipeline fro m Mohe to Daqing, Journal of Engineering Geology, vol. 18, no.
2, pp. 241-251, 2010. (in Chinese)

12