Applied Thermal Engineering 98 (2016) 220227

Contents lists available at ScienceDirect

Applied Thermal Engineering

j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / a p t h e r m e n g

Research Paper

Cooling performance of two-phase closed thermosyphons installed at

a highway embankment in permafrost regions
Fan Yu a, Jilin Qi b,*, Mingyi Zhang a, Yuanming Lai a,c, Xiaoliang Yao a, Yongzhi Liu a,d,
Guilong Wu a,d
a State Key Laboratory of Frozen Soil Engineering, Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences,
Lanzhou 730000, China
b
Beijing High Institution Research Center for Engineering Structures and New Materials, Beijing University of Civil Engineering and Architecture, Beijing
100044, China
c School of Civil Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China
d Beiluhe Observation Station of Frozen Soil Environment and Engineering, Cold and Arid Region Environmental and Engineering Research Institute, Chinese
Academy of Sciences, Lanzhou 730000, China

H I G H L I G H T S

Long-term monitored data are presented for a road section of the Qinghai-Tibet Highway before and after installing the TPCTs.
Cooling scope and period of the TPCTs are analyzed; cooling effects for the soil layers are examined.
Cooling performance of the TPCTs is discussed from the perspectives of embankment deformation and crack formation.
Some suggestions are proposed for a better design of the TPCTs in highway constructions in permafrost regions.

A R T I C L E

I N F O

Article history:
Received 10 August 2015
Accepted 25 November 2015
Available online 24 December 2015
Keywords:
Two-phase closed thermosyphon
Ground temperature
Permafrost degradation
Embankment deformation
Crack formation
Thaw consolidation

A B S T R A C T

Two-phase closed thermosyphon (TPCT) is a popular way to prevent permafrost layers from degrading,
and consequently ensure the stabilities of engineering constructions in permafrost regions. Although TPCTs
have been numerically and experimentally investigated for many years, long-term eld monitored data
concerning the cooling performance of TPCTs are limited. This paper presents the ground temperatures,
embankment deformations and some related meteorological factors for a road section of the Qinghai
Tibet Highway before and after installing the TPCTs in permafrost regions. Based on the monitored data,
three main aspects are analyzed: 1) cooling scope and period of the TPCTs; 2) cooling effects for the soil
layers, especially for the permafrost layers; and 3) remedying effects with respect to embankment deformation and crack formation. Some corresponding suggestions are proposed for a better design of TPCTs
in the construction of roadways in permafrost regions.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
In permafrost regions, engineering constructions have been suffering different levels of deformations caused by permafrost
degradation [1,2]. Taking the QinghaiTibet Highway as an example,
31.7% of the road sections from Golmud to Lhasa (520 km) in the
permafrost regions faced roadbed diseases in 1999 [3], among which
85% of them were caused by thaw settlement [4]. Under global
warming and the thermal effect caused by road embankments, the
underlying permafrost layers are degrading [5]. Thaw consolidation has been considered to be the main cause of embankment

* Corresponding author. Tel.: +8613811703629; fax: +86 931 4967292.

E-mail address: jilinqi@bucea.edu.cn (J. Qi).
http://dx.doi.org/10.1016/j.applthermaleng.2015.11.102
1359-4311/ 2015 Elsevier Ltd. All rights reserved.

deformation for a long time [2]. In recent years, creep of warm permafrost layers has also been proved to be another main cause [6,7].
It is found that the two deformation causes are exactly the main
reason why the QinghaiTibet Highway was reconstructed and repaired for many times [8,9]. In addition, 79% of the permafrost
regions on the QinghaiTibet Plateau were characterized by warm
permafrost, where the mean annual ground temperatures (MAGTs)
were higher than 1.5 oC [10]. The easy-to-thaw characteristic determines that some cooling methods should be taken to protect the
permafrost layers from degrading, and thus to ensure the stabilities of the engineering constructions.
Two-phase closed thermosyphon (TPCT) is one of those cooling
methods, which has already been used in permafrost regions for
many decades. For instances, the TPCTs were successfully used in
the Trans-Alaska Pipeline System between 1974 and 1977 [11]. The
TPCTs were also proved by in-situ geothermal observation between

F. Yu et al./Applied Thermal Engineering 98 (2016) 220227

1990 and 1997 to be a reliable technology for rehabilitating a housing

with unacceptable settlement on the Ellesmerer Island, Canada [12].
In China, the TPCTs have been frequently employed to insure the
stabilities of the QinghaiTibet Highway [13,14], the Qinghai
Tibet Railway [15,16], and the ChaidaerMuli Railway [17]. In
addition, other cooling methods were combined with the TPCTs for
better cooling effects, such as crushed rock revetment [18,19], forcedair ventilation [20] and insulating material [21]. In these engineering
constructions, the evaporator sections were embedded in the soil
layers, while the condenser sections were exposed in the air. The
TPCTs only work when the temperature difference exceeds the
threshold, i.e. the start-up temperature difference. Zhang et al. [22,23]
found that it was about 0.2 C by means of indoor physical model
test and a self-designed experimental apparatus.
In order to enlarge the cooling range underneath the embankments, inclined TPCTs have been numerically and experimentally
investigated [23,24]. Besides the inclining angle, there are three other
inuencing factors of the cooling performance of the TPCTs, including aspect ratio (the ratio between evaporator length and internal
diameter), lling ratio (the ratio between the volumes of working
uid and evaporator section) and working uid [23]. The inuencing factors have been extensively investigated by previous researchers
[23,2528].
With respect to the previous numerical modeling work, for a simplied calculation, the thermal resistance of the TPCTs were usually
ignored and the heat-convention coecient between air and the
radiating ns was taken as a constant when the TPCTs were working
[21,2931]. On the other hand, the heat transfer coecients for every
part of the TPCTs were introduced by Pan and Wu [32], based on
which a coupled air-TPCT-soil model was put forward [33]. For ease
of calculation, however, some empirical formulas were adopted for
calculating the heat resistances in the coupled model, which may
lead to some unavoidable errors. Under this circumstance, a fullycoupled model is urgently needed. The governing equations in the
three zones have already researched respectively. The three governing equations in the air zone are continuity, momentum and
energy equations [34,35]. Also, the three equations are usually used
for simulating two-phase ows in TPCTs [3638]. In the soil layers,

221

the heat transfer equation with phase-change considered is usually

adopted [3941].
From the above analyses, it can be found that the TPCTs have been
considered to be a popular and ecient way to ensure the stabilities of the engineering constructions in permafrost regions, especially
in warm permafrost regions. There are many inuencing factors for
the cooling performance of the TPCTs, which needs further investigation. A fully-coupled air-TPCT-soil model is urgently needed for
a better prediction. To date, long-term eld monitored data about
the cooling performance of the TPCTs are currently lacking. This
would be favorable for understanding the cooling performance of
the TPCTs, and for establishing the fully-coupled model for the stabilities of the engineering constructions in permafrost regions.
In this paper, more than 10-year monitored data including ground
temperatures, embankment deformations and some related meteorological factors are presented for a road section installed with
TPCTs in permafrost regions. The main objective is to examine the
cooling performance of the TPCTs from the perspectives of thermal
and mechanical (deformation) stabilities.

2. Monitoring program
The monitoring program was carried out at a road section of the
QinghaiTibet Highway. The highway mileage is K3187 + 000, while
the place is named as Kaixinling. The longitude and latitude are
N335729 and E922058 respectively, and the geographic position can be seen in a previous paper [42]. An illustration of the
instrumented road section is shown in Fig. 1. Ground temperatures and embankment deformations were started to monitor in
2003. Due to an unacceptable embankment deformation, TPCTs were
installed around September 2009. The working uid is anhydrous
ammonia, while the container is made by carbon steel. As shown
in Fig. 1, the lengths of the three parts of the TPCTs, condenser, adiabatic and evaporator, are 4, 3 and 5 m respectively. Some other
parameters of the TPCTs are summarized in Table 1. A meteorological station was established in Aug. 2008. Detailed information about
these three monitoring items is introduced in the following.

Fig. 1. Schematic cross-section (a) and plan (b) of the instrumented road section installed with TPCTs.

222

F. Yu et al./Applied Thermal Engineering 98 (2016) 220227

Table 1
Some parameters of the TPCTs.
Parameters

Values (mm)

Inner diameter of pipe

Outer diameter of pipe
Fin height
Fin space
Fin thickness

79
89
25
10
1

2.1. Ground temperatures

Two thermistor cables were installed at the left shoulder and in
the middle of the road section on September 2003, and another one
was added at the right shoulder on April 2013. As shown in Fig. 1,
the three cables reached the depth of 20 m below the asphalt pavement. The thermistors have an accuracy of 0.05 C, which were
developed by State Key Laboratory of Frozen Soil Engineering,
Chinese Academy of Sciences. The data were collected by a CR3000
datalogger (Campbell Scientic Inc., USA) twice a month. The depths
of permafrost tables are obtained by interpolating calculation based
on the monitored ground temperatures. 0 C is taken as the thawing
temperature for ease of discussion, since it varies with soil type, load,
water and salt contents, etc. [43].
2.2. Embankment deformations
Twenty steel nails and a 15-m-deep steel pipe were embedded
into the asphalt pavement on October 2003, as shown in Fig. 1b. It
is assumed that the deformation of the soil layers deeper than 15 m
can be negligible, and thus the steel pipe can be taken as a bench-

mark. The elevation differences between the steel nails and the
benchmark are the embankment deformations. An electronic total
station was used to monitor the deformations once or twice a month.
It was suspended from August 2008 to August 2009 due to the replacement of asphalt pavements and the installation of the TPCTs.
2.3. Metrological factors
A meteorological station was installed nearby the road section
on August 2008. Air temperatures and wind speeds at the heights
of 2 m and 10 m were monitored using HMP45C-L11 and 05103L11 sensors (Vaisala, Finland). In addition, air humidity and radiation
and precipitation were also continuously measured. Since air temperatures and wind speeds are the top two inuence factor for the
heat transfer coecient of the TPCTs [44], only they are adopted
in this paper.
3. Results and discussions
The monitored data including ground temperatures, embankment deformations, air temperature and wind speeds at the height
of 2 m are shown in Fig. 2. Fig. 2a shows the monthly average values
of air temperatures and wind speeds at the height of 2 m. The accumulated values of embankment deformations (Fig. 2b) correspond
with the ground temperatures below left/right shoulders and in the
middle of the road section (Fig. 2ce). The Kriging interpolation
method is adopted to obtain the continuous ground temperatures
based on the monitored data at different depths. The embankment deformations at left shoulder refer to the average values of
the monitoring points 10# and 15#, that in the middle refer to the
average values of the monitoring points 8# and 13#, while that at

Fig. 2. Monitored data including air temperatures and wind speeds (a), embankment deformations (b) and ground temperatures below the left shoulder (c), in the middle
(d) and below the right shoulder (e).

F. Yu et al./Applied Thermal Engineering 98 (2016) 220227

223

Fig. 3. Average ground temperatures in the middle (a) and below the left shoulder (b) at the road section of the QinghaiTibet Highway in 2004, 2009, 2010 and 2013.

the right shoulder refer to the average values of the monitoring points
6# and 11# (Fig. 1b). It is worth mentioning that right/left shoulders are determined when facing the direction to Lhasa (Fig. 1b).

3.1. Cooling scope of the TPCTs

The cooling scope indicates the soil layers cooled by the TPCTs
horizontally and vertically, which is closely related with the embankment deformation. Since the ground temperatures were missing
during the period of installing the TPCTs in July and August of 2009,
the ground temperatures in 2008 is taken as the initial state to
examine the cooling scope. Because of lacking sucient thermistor cables from the left shoulder to the middle of the road section
(Fig. 1), it is only possible to get some information about the horizontal cooling radius from the thermistor cable in the middle. Fig. 3
presents the average ground temperatures in the middle and below
the left shoulder in 2004, 2009, 2010 and 2013. In the middle of
the road section, the ground temperatures of the upper 4-m soil
layers rstly decreased from 2004 to 2008 and then increased from
2008 to 2013. The similar variational trend was found for the upper
2-m soil layers below the left shoulder. This phenomenon indicates that changes in the ground temperatures of the upper soil layers
may be mainly resulted from climate warming and thermal inuence of the road embankment. Small changes were found for the
ground temperatures of the soil layers below the depth of 4 m in
the middle, indicating that the horizontal cooling radius may be less
than 5 m. Thus, it is suggested that inclined TPCTs should be better
for cooling the permafrost layers in the middle. It can be seen from
Fig. 3b that the vertical cooling range below the left shoulder was
29 m from 2008 to 2010 (area (1)), and was 412 m from 2010
to 2013 (area (2)). Thus, the maximum cooling depth reached 12 m
during the monitoring period.
Another interesting phenomena was that the cooling range moved
downward. The upper limit moving downward from 2 m to 4 m may
be mainly caused by climate warming and the thermal effect of the
road embankment. The lower limit moving downward from 9 m to
12 m may be resulted from the long-term cooling effect of the TPCTs
neutralizing the upward geothermal ows. Therefore, it is found that
the cooling scope of the TPCTs is under the thermal inuences from
climate warming, the road embankment and upward geothermal
ows.

3.2. Cooling period of the TPCTs

For the road embankments installed with TPCTs in permafrost
regions, it is well known that TPCTs start to work when the temperature at condenser section is a little bit lower than that at
evaporator section [21,23]. Because the temperatures at evaporator and condenser sections were not measured directly in this paper,
the working period can only be estimated according to the air temperature and ground temperatures around the TPCTs. Therefore, it
is assumed that the TPCTs started to work once the air temperature was colder than the ground temperatures. Since the evaporator
section ranged from 2 to 8 m below the road surface, the ground
temperatures of these soil layers are employed. Fig. 4 presents the
air temperature and the ground temperatures at the depth of 2, 4,
6 and 8 m below the left shoulder. Due to the different thermal states
at different depths, the cooling period may vary from part to part
at the evaporator section. In 2010, the cooling period at the depth
of 2 m ranged from January to April and then from August to December, while it ranged from January to April and then from October
to December at the depths of 4, 6 and 8 m. Therefore, the cooling
period was eight months at 2 m, which was one month longer than
that at 4, 6, and 8 m. This indicates that some errors may be

Fig. 4. Air temperature and ground temperatures at the depths of 2, 4, 6 and 8 m

below the left shoulder in 2010.

224

F. Yu et al./Applied Thermal Engineering 98 (2016) 220227

triggered when only one value of soil temperature was usually employed for determining the cooling period of the TPCTs in previous
literatures [21,30].

that the ground temperature was slightly decreased at 4 m from

2010 to 2013. The possible reason is that the ground temperature
may also be inuenced by the heat transferred from the asphalt
pavement.

3.3. Cooling effects of the TPCTs

3.3.1. Cooling effect around the adiabatic and evaporator sections of
TPCTs
The adiabatic section ranged from 0 to 3 m, while the evaporator section ranged from 3 to 8 m below the road surface. Fig. 5
illustrates the ground temperature at the depths of 0.5, 1.5, 2, 4, 6
and 8 m in the years of 2008, 2010 and 2013. The ground temperature in 2008 is also taken as the initial thermal state to examine
the cooling effects. For the soil layers around the adiabatic section
(Fig. 5ac), the ground temperatures were greatly inuenced by
climate warming and the thermal inuence caused by the road embankment. Thus, it is dicult to estimate the cooling effects taken
by the TPCTs.
As for the soil layers around the evaporator section (Fig. 5df),
the ground temperatures were mainly inuenced by the TPCTs.
Thus, the black shadows may indicate the cooling effects caused
by the TPCTs at different depths. The larger the area is, the greater
the cooling effect was. Though there was a larger temperature
difference between the air temperature and the ground temperature at 4 m during the cooling period (Fig. 4), it seems that the
strongest cooling effect was found at 6 m (Fig. 5e). From 2008 to
2010, the cooling amplitudes of the ground temperatures at 4, 6
and 8 m were 0.70, 0.74 and 0.26 C respectively, while
they were 0.85, 1.14 and 0.62 C from 2008 to 2013. Of note is

3.3.2. Cooling effect for the permafrost layers

The main purpose of the TPCTs was to prevent the underlying
permafrost layers from thawing or warming. Permafrost thawing
can be directly reected by the raising/dropping of the permafrost
table, while mean annual ground temperature (MAGT) at the depth
of 20 m can be taken as an indicator for permafrost warming. Fig. 6
presents the variations of the depths of permafrost tables and MAGTs
below the left and right shoulders and in the middle of the road
embankment. At the left shoulder, the permafrost table rstly
dropped down from 2003 to 2008 without TPCTs, and then rose up
from 2008 to 2013 under the cooling inuence of the TPCTs. The
increasing rate of the depth of permafrost table was 0.30 m/a, while
the decreasing rate was 0.51 m/a. The MAGT kept rising during the
monitoring period, but different rising rates were observed. The temperature rising rate was 0.24 C/10a from 2003 to 2008, while it was
0.06 C/10a from 2008 to 2013. An obvious cooling effect was found
below the left shoulder, since the thermistor cable was so close to
the TPCTs. The permafrost table rose up faster in the rst two years
(20082010) than that in the latter years (20102013), while MAGT
kept increasing in the rst two years but then showed a tendency
to decrease in the latter years. A possible reason is that the bottom
of the evaporator section was just placed at the permafrost table
in 2008. Thus, the upper permafrost layers would be rstly cooled
with a result of the permafrost table rising fast, but the rising rate

Fig. 5. Ground temperatures at the depths of 0.5 m (a), 1.5 m (b), 2 m (c), 4 m (d), 6 m (e), and 8 m (f) below the left shoulder in 2009, 2010 and 2013.

F. Yu et al./Applied Thermal Engineering 98 (2016) 220227

225

Fig. 6. Depths of the permafrost tables and mean annual ground temperatures at the depth of 20 m below the left and right shoulders and in the middle of the road section.

would slow down with the heat transferring from the lower permafrost layers in the later years.
In the middle of the road section, a similar cooling effect as that
below the left shoulder was observed. Before installing the TPCTs,
the increasing rate of the depth of permafrost table was 0.18 m/a,
while the rising rate of MAGT was 0.13 oC/10a. After installing the
TPCTs, the decreasing rate of the depth of permafrost table was
0.11m/a, while the rising rate of MAGT was 0.02 C/10a. It is worth
mentioning that the permafrost table fell down from 2009 to 2011,
and then rose up from 2011 to 2014. This indicates that it would
take a few years for the TPCTs to case a cooling effect on the permafrost layers in the middle.
As for the permafrost layers below the right shoulder, it is impossible to determine the cooling effect due to the insucient
geothermal data. The depth of the permafrost table in 2013 was
about 3.5 m, which was much lower than that below the left shoulder or in the middle of the road section. This was because that the
right slope was on the shady side, which would be favorable for
cooling the underlying permafrost layers.

As for the different deformation points, the deformation rates

were 4.0 and 1.2 cm/a at the left shoulder before and after installing the TPCTs respectively, 4.4 and 0.2 cm/a in the middle, and 5.5
and 0.9 cm/a at the right shoulder. The deformation rate at left shoulder was 6 times faster than that in the middle after installing the
TPCTs. It seems that the cooling effect at the left shoulder was better
than that in the middle (Fig. 6), but a better remedying effect was
observed in the middle with respect to embankment deformation. Three possible reasons are 1) creep of the warm permafrost
layer was the main deformation source after installing the TPCTs,
since the geotemperatures of the permafrost layers below the left

3.4. Remedying effects of the TPCTs

To remedy the embankment with unacceptable deformations or
cracks was the purpose of installing the TPCTs. In another words,
the embankment stability can be taken as an indicator to examine
the cooling performance of the TPCTs. In this section, the cooling
performance will be discussed from the perspectives of embankment deformation and crack formation.
3.4.1. Remedying effect with respect to embankment deformation
Embankment deformation is closely related with the degradation of permafrost layers underneath the road embankment [42].
Thaw consolidation was associated with permafrost thawing, while
the creep of permafrost layers was mainly inuenced by permafrost warming [6,9]. Therefore, the installation of the TPCTs was
supposed to ease the embankment deformation. The accumulated
deformations before and after installing the TPCTs at the right and
left shoulders and in the middle are presented in Fig. 7. The overall
characteristic is that the deformation rates with TPCTs were obviously slower than that without TPCTs. The average deformation rate
before installing the TPCTs was about 6 times larger than that after
installing the TPCTs. From this point, the TPCTs may be a reliable
way to alleviate embankment deformation.

Fig. 7. Embankment deformations at the left/right shoulders and in the middle of

the road section before and after installing the TPCTs.

226

F. Yu et al./Applied Thermal Engineering 98 (2016) 220227

Fig. 8. Two photos showing the cracks on the asphalt pavement of the road embankment.

shoulder were higher than that in the middle; 2) consolidation of

the thawed permafrost layers in the middle can be negligible, since
the ice content was very low; 3) permafrost conditions were different below the left shoulder and in the middle, leading to different
deformation rates.
3.4.2. Remedying effect with respect to crack formation
Fig. 8 shows the cracks on the road surface on July 2013 (a) and
April 2015 (b). The road surface was repaired by lling with asphalts before April of 2015. It can be seen that 1) the cracks were
mainly longitudinal; 2) the cracks formed earlier at the left shoulder than the right shoulder; 3) the cracks at the left shoulder were
presented more severe than that at the right shoulder. If taking the
embankment deformation in the middle as a reference, deformation differences between the middle and the left/right shoulders after
installing the TPCTs are shown in Fig. 9. The annual deformation difference was about 1 cm, which would be insucient to induce the
cracks. Therefore, the longitudinal cracks may not be caused by the
deformation differences. From Fig. 8a, it can be seen that a longitudinal crack between the curb and the asphalt pavement appeared
rstly with the maximum width of about 30 cm. Thus, the cracks
may be resulted from the different levels of permafrost degradation below the asphalt pavement and the left slope (or the
neighboring natural ground surface). An intense thermal effect at
the left slope (sunny slope) may be the cause of more and severe
cracks appearing at the left shoulder than the right shoulder.
4. Conclusions and suggestions
In order to examine the cooling performance of TPCTs, ground
temperatures, embankment deformations and some metrological

Fig. 9. Deformation differences between the middle and left/right shoulders after
installing the TPCTs.

factors have been monitored for a road section of the Qinghai

Tibet Highway before and after installing the TPCTs since 2003.
Frequently-concerned topics are analyzed including cooling scope,
cooling period, cooling effects and remedying effects with respect
to both embankment deformation and crack formation. The major
conclusions and some suggestions can be summarized as follows:
1. The cooling depth of the TPCTs reached the maximum depth of
12 m during the monitoring period, which may be reliable to
prevent the lower permafrost layers from warming. However,
small changes were observed for the ground temperatures of permafrost layers in the middle of the road section, indicating that
inclined TPCTs should be suggested. The cooling scope of the
TPCTs is mainly inuenced by climate warming, thermal effect
of road embankments, and upward geothermal ows. These three
main inuencing factors should be taken into consideration when
designing the layout of TPCTs in the construction of roadways.
2. The cooling period was eight months at 2 m, which was one
month longer than that at 4, 6, and 8 m. In previous literatures,
only one value of soil temperature was usually employed for determining the cooling period of the TPCTs, which may induce
some errors. Thus, the variation of soil temperatures along the
evaporator section should be taken into consideration for improving the calculation accuracy.
3. The best cooling effect was found around the middle of the evaporator section. This should be considered when designing the
embedding depth of the TPCTs. Different cooling effects were
found for the permafrost layers below the left/right shoulders
and in the middle of the road section. The permafrost tables below
the two shoulders were cooled fast with the permafrost tables
raising up and the MAGTs decreasing. However, it would be taken
a long period for the permafrost layers in the middle to cool down.
The large differences among the depths of the permafrost tables
below the left/right shoulders and in the middle may be unfavorable for the stability of the road embankment. Thus, different
embedding depths, inclining angles and longitudinal intervals
of the TPCTs are recommended for the sunny and shady
shoulders.
4. The embankment deformation slowed down very obviously after
installing the TPCTs. However, creep of the warm permafrost
layers may be a potential deformation source under the cooling
effect of the TPCTs. The deformation differences at different places
of the road embankment was not the cause of the longitudinal
cracks. Instead, different thermal states of the permafrost layers
were presented below the asphalt pavement, the embankment
slopes and the neighboring natural ground surfaces. This may
be the reason for the formation of the cracks. Thus, preventing
or controlling the longitudinal cracks would be an important topic
in both engineering construction and scientic research in the

F. Yu et al./Applied Thermal Engineering 98 (2016) 220227

future. Under this circumstance, combining other cooling methods

with TPCTs may be a solution for solving this problem.
There are limitations for any approach; some improvements and
additional work should be done in the future. The cooling permafrost of TPCTs needs to be tested using more road sections with
different geological conditions. Also, some improvements should be
made for the eld experiment, such as direct measurement of pipewall temperatures of TPCTs, and increase in the thermistor cables
for measuring ground temperatures. At present, the major challenge is to establish a fully-coupled air-TPCT-soil model, which is
just our ongoing research. For this purpose, laboratory model tests
are planned to be conducted to obtain some key input parameters, such as the heat-convection coecient between air and the
radiating ns of TPCTs. These three kinds of research would benet
to an optimal design of TPCTs for the engineering constructions in
permafrost regions.
Acknowledgements
This study was supported by the National Natural Science Foundation of China (Grant No. 41572268), the National Key Basic Research
Program of China (973 Program No. 2012CB026102), the National
Natural Science Foundation of China (Grant Nos. 41230630 and
41471063), the 100-Talent Program of the Chinese Academy of Sciences (Granted to Dr. Mingyi Zhang), the National Natural Science
Foundation of China (Grant No. 41501084), and the Foundation for
Excellent Youth Scholars of CAREERI, CAS (Grant No. Y451091001).
The authors greatly appreciate the two anonymous reviewers for
their constructive comments and suggestions, and we would also
like to thank Dr. Wansheng Pei for providing some eld photos.
References
[1] F.E. Nelson, O.A. Anisimov, N.I. Shiklomanov, Subsidence risk from thawing
permafrost, Nature 410 (6831) (2001) 889890.
[2] Z.W. Wu, Y.Z. Liu, Frozen Subsoil and Engineering, Ocean Press, Beijing, 2005.
[3] G.D. Cheng, W. Ma, Frozen soil engineering problems in construction of the
Qinghai-Tibet Railway, Chin. J. Nat. 28 (6) (2006) 315320 (in Chinese).
[4] Q.B. Wu, Y.Z. Liu, J.M. Zhang, C.J. Tong, A review of recent frozen soil engineering
in permafrost regions along Qinghai-Tibet Highway, China, Permaf. Perig. 13
(3) (2002) 199205.
[5] Q.B. Wu, X.F. Dong, Y.Z. Liu, H.J. Jin, Responses of permafrost on the Qinghai-Tibet
Plateau, China, to climate change and engineering construction, Arct. Antarct.
Alp. Res. 39 (4) (2007) 682687.
[6] J.L. Qi, S. Yu, J.M. Zhang, Z. Wen, Settlement of embankments in permafrost
regions in the Qinghai-Tibet Plateau, Norw. J. Geogr. 61 (2) (2007) 4955.
[7] B. Zheng, J.M. Zhang, Y.H. Qin, Investigation for the deformation of embankment
underlain by warm and ice-rich permafrost, Cold Reg. Sci. Technol. 60 (2) (2010)
161168.
[8] S.J. Wang, M. Huo, W.J. Zhou, Subgrade failure of Qinghai-Tibet Highway in
permafrost area, Highway 5 (2004) 2226 (in Chinese).
[9] F. Yu, J.L. Qi, X.L. Yao, Y.Z. Liu, In-situ monitoring of settlement at different layers
under embankments in permafrost regions on the Qinghai-Tibet Plateau, Eng.
Geol. 160 (2013) 4453.
[10] G.D. Cheng, Research on engineering geology of the roadbed in permafrost
regions of Qinghai-Xizang Plateau, Qual. Sci. 23 (2) (2003) 134141 (in Chinese).
[11] C.E. Heuer, The application of heat pipes on the Trans-Alaska pipeline, United
States Army Crops of Engineers, Cold Regions Research and Engineering
Laboratory Special Report 79-26, 1979, 127.
[12] J. McKenna, K. Biggar, The rehabilitation of a passive-ventilated slab on grade
foundation using horizontal thermosyphons, Can. Geotech. J. 35 (4) (1998)
684691.
[13] Y. Song, L. Jin, J.Z. Zhang, In-situ study on cooling characteristics of two-phase
closed thermosyphon embankment of QinghaiTibet Highway in permafrost
regions, Cold Reg. Sci. Technol. 93 (2013) 1219.
[14] D. Wu, L. Jin, J.B. Peng, Y.H. Dong, Z.Y. Liu, The thermal budget evaluation of
the two-phase closed thermosyphon embankment of the QinghaiTibet
Highway in permafrost regions, Cold Reg. Sci. Technol. 103 (2014) 115122.
[15] W.D. Pan, S.C. Zhao, W.Z. Xu, S.S. Yu, W.D. Ma, Application of probe to enhance
thermal stability of roadbed in Plateau permafrost areas, J. Glaciol. Geocryol.
25 (4) (2003) 433438 (in Chinese).

227

[16] J.J. Wu, W. Ma, Z.Z. Sun, Z. Wen, In-situ study on cooling effect of the two-phase
closed thermosyphon and insulation combinational embankment of the
QinghaiTibet Railway, Cold Reg. Sci. Technol. 60 (3) (2010) 234244.
[17] B. Zhang, Y. Sheng, J. Chen, J. Li, In-situ test study on the cooling effect of
two-phase closed thermosyphon in marshy permafrost regions along the
ChaidaerMuli Railway, Qinghai Province, China, Cold Reg. Sci. Technol. 65 (3)
(2011) 456464.
[18] Y.M. Lai, H.X. Guo, Y.H. Dong, Laboratory investigation on the cooling effect of
the embankment with L-shaped thermosyphon and crushed-rock revetment
in permafrost regions, Cold Reg. Sci. Technol. 58 (3) (2009) 143150.
[19] W. Ma, Z. Wen, Y. Sheng, Q.B. Wu, D.Y. Wang, W.J. Feng, Remedying embankment
thaw settlement in a warm permafrost region with thermosyphons and crushed
rock revetment, Can. Geotech. J. 49 (9) (2012) 10051014.
[20] L. Smith, J. Graham, J. Nixon, A. Washuta, Thermal analysis of forced-air and
thermosyphon cooling systems for the Inuvik airport expansion, Can. Geotech.
J. 28 (3) (1991) 399409.
[21] Z. Wen, Y. Sheng, W. Ma, J.L. Qi, J.C. Wu, Analysis on effect of permafrost
protection by two-phase closed thermosyphon and insulation jointly in
permafrost regions, Cold Reg. Sci. Technol. 43 (3) (2005) 150163.
[22] M.Y. Zhang, Y.M. Lai, Y.H. Dong, L. Jin, W.S. Pei, J. Harbor, Laboratory investigation
of the heat transfer characteristics of a two-phase closed thermosyphon, Cold
Reg. Sci. Technol. 95 (2013) 6773.
[23] M.Y. Zhang, Y.M. Lai, W.S. Pei, L. Jin, Effect of inclination angle on the heat
transfer performance of a two-phase closed thermosyphon under lowtemperature conditions, J. Cold Reg. Eng. 28 (4) (2014) 04014007.
[24] Y.P. Yang, S.H. Zhou, Q.C. Wei, Numerical analysis of proper thermosyphon
inclination angle used in permafrost embankment, China Railway Sci. 27 (3)
(2006) 17 (in Chinese).
[25] T. Payakaruk, P. Terdtoon, S. Ritthidech, Correlations to predict heat transfer
characteristics of an inclined closed two-phase thermosyphon at normal
operating conditions, Appl. Therm. Eng. 20 (9) (2000) 781790.
[26] B. Jiao, L.M. Qiu, X.B. Zhang, Y. Zhang, Investigation on the effect of lling ratio
on the steady-state heat transfer performance of a vertical two-phase closed
thermosyphon, Appl. Therm. Eng. 28 (11) (2008) 14171426.
[27] B.M. Ziapour, H. Shaker, Heat transfer characteristics of a two-phase closed
thermosyphon using different working uids, Heat Mass Transf. 46 (3) (2010)
307314.
[28] B. Jiao, L.M. Qiu, Z.H. Gan, X.B. Zhang, Determination of the operation range
of a vertical two-phase closed thermosyphon, Heat Mass Transf. 48 (6) (2012)
10431055.
[29] F.X. Liu, J.H. Liu, D.H. Guo, L.X. Zhang, Numerical analysis of thermosyphon in
protecting the roadbed in permafrost, China Civil Eng. J. 37 (9) (2005) 4147
(in Chinese).
[30] Y. Sheng, Z. Wen, W. Ma, J.C. Wu, Three dimensional nonlinear analysis of
thermal regime of the two phase closed thermosyphon embankment of
Qinghai-Tibetan Railway, J. China Railway Soc. 28 (1) (2006) 125130 (in
Chinese).
[31] J.L. Wang, S.M. Liao, Numerical heat transfer analysis and optimization for
thermosyphons coupled with natural ground in permafrost regions, China
Railway Sci. 27 (6) (2007) 16 (in Chinese).
[32] Y. Pan, C.Z. Wu, Numerical investigations and engineering applications on
freezing expansion of soil restrained two-phase closed thermosyphons, Int. J.
Therm. Sci. 41 (4) (2002) 341347.
[33] M.Y. Zhang, Y.M. Lai, J.M. Zhang, Z.Z. Sun, Numerical study on cooling
characteristics of two-phase closed thermosyphon embankment in permafrost
regions, Cold Reg. Sci. Technol. 65 (2) (2011) 203210.
[34] M.Y. Zhang, Y.M. Lai, Z.H. Gao, W.B. Yu, Inuence of boundary conditions on
the cooling effect of crushed-rock embankment in permafrost regions of
QinghaiTibetan Plateau, Cold Reg. Sci. Technol. 44 (3) (2006) 225239.
[35] Y.M. Lai, W.S. Pei, W.B. Yu, Calculation theories and analysis methods of
thermodynamic stability of embankment engineering in cold regions, Chinese
Sci. Bull. 59 (3) (2014) 261272.
[36] Z.J. Zuo, F.S. Gunnerson, Numerical modeling of the steady-state two-phase
closed thermosyphon, Int. J. Heat Mass Trans. 37 (17) (1994) 27152722.
[37] B. Fadhl, L.C. Wrobel, H. Jouhara, Numerical modelling of the temperature
distribution in a two-phase closed thermosyphon, Appl. Therm. Eng. 60 (1)
(2013) 122131.
[38] B. Fadhl, L.C. Wrobel, H. Jouhara, CFD modelling of a two-phase closed
thermosyphon charged with R134a and R404a, Appl. Therm. Eng. 78 (2015)
482490.
[39] V.E. Romanovsky, T.E. Osterkamp, N.S. Duxbury, An evaluation of three
numerical models used in simulations of the active layer and permafrost
temperature regimes, Cold Reg. Sci. Technol. 26 (3) (1997) 195203.
[40] Y.M. Lai, M.Y. Zhang, S.Y. Li, Theory and Application of Cold Regions Engineering,
Science Press, Beijing, 2009.
[41] M.M. Darrow, Thermal modeling of roadway embankments over permafrost,
Cold Reg. Sci. Technol. 65 (3) (2011) 474487.
[42] F. Yu, J.L. Qi, X.L. Yao, Y.Z. Liu, Degradation process of permafrost underneath
embankments along Qinghai-Tibet Highway: an engineering view, Cold Reg.
Sci. Technol. 85 (2013) 150156.
[43] X.Z. Xu, J.C. Wang, Y.S. Deng, Thermo-physic of Frozen Soil, Science Press, Beijing,
2010.
[44] Q.B. Wu, S.Y. Liang, X.W. Gao, Research of Convection heat transfer between
thermalpile and air, J. Glaciol. Geocryol. 18 (1996) 3742 (in Chinese).