Tunnelling and Underground Space Technology 81 (2018) 121–128

Contents lists available at ScienceDirect

Tunnelling and Underground Space Technology

journal homepage: www.elsevier.com/locate/tust

Behaviors of existing twin subway tunnels due to new subway station T

excavation below in close vicinity
⁎
Chengping Zhang, Xu Zhang, Qian Fang
Key Laboratory of Urban Underground Engineering of Ministry of Education, Beijing Jiaotong University, Beijing 100044, China

A R T I C LE I N FO A B S T R A C T

Keywords: This paper presents a case of Chongwenmen subway station of Beijing Subway Line 5 excavated below the
Adjacent excavation existing Line 2, which was the first project of new subway construction below existing subway in Beijing. The
Subway station construction minimum clearance between the new and the existing tunnels was 1.98 m. The layout, the geological and hy-
Shallow tunnelling method drogeological conditions, the construction process, and the monitoring measures are introduced. During con-
Compensation grouting
struction, the maximum settlement of the existing tunnels reached 31.3 mm and the speed of trains run inside the
existing tunnels was slowed down. In order to compensate the settlement of the existing tunnels and ensure the
operational safety of the trains running inside the existing tunnels, compensation grouting was adopted. The
pillar between the existing and the new tunnels was strengthened by grouting first. Then the pillar and the
existing tunnels were uplifted together by specific controlled grouting pressure. The compensation grouting
successfully uplift the existing tunnels. Moreover the subsequent settlements of the existing tunnels were small as
the pillar had been sufficiently reinforced.

1. Introduction 2. Project overview

As more subway tunnels are excavated in urban areas, a new tunnel
construction below an existing tunnel is commonly encountered. In the
cases that the two tunnels are closely spaced and the cross-sectional
area of the new tunnel is large, it is very challenging to ensure both the
serviceability of the existing tunnel and the construction safety of the
new tunnel. Therefore, it is important to investigate the mechanical
characteristics associated with new tunnel construction below the ex-
isting tunnel.
The cases of new tunnel construction below existing tunnel are
generally rare and are only reported by a few researchers (Kimmance
et al., 1996; Cooper, 2001; Standing and Selman, 2001; Cooper et al.,
2002). A well-documented case can undoubtedly offer valuable in-
formation for future research and project. In this research, we introduce
the project of the Chongwenmen subway station of Line 5 excavated
below the existing twin subway tunnels of Line 2 in Beijing. It was the
first project of new subway construction below existing subway in
Beijing (2003–2007). The construction measures, the monitoring re-
sults, and the lessons learned from this project are systematically illu-
strated.
The plan view of the existing twin subway tunnels of Line 2 and the
new subway station of Line 5 at the Chongwenmen subway station in
Beijing is shown in Fig. 1. The cross section and the longitudinal section
of the project are shown in Fig. 2. The existing twin tunnels, running
east–west parallelly, have the same square cross section. The external
length of each tunnel is 5.9 m. They are supported by reinforced con-
crete lining. The thickness of the roof, floor, and side walls of each
tunnel are 0.8 m, 0.7 m and 0.7 m respectively. Each tunnel is composed
of a series of 18 m long sections, which are isolated from each other by
means of movement joints. The overburden depth of the existing twin
tunnels is about 5.1 m. The existing twin tunnels were originally con-
structed from 1965 to 1968 using the open cut method.
The new Chongwenmen subway station of Line 5 is composed of
two double-deck parts on two sides of the station and one single-deck
part in the middle. The single-deck part, with single arch and triple
spans, was constructed below the existing twin tunnels. The minimum
vertical clearance between the existing tunnels and the new station is
1.98 m. The shallow tunnelling method, particularly designed for
shallowly-buried tunnels constructed in soft ground (Fang et al., 2012),
was adopted for the construction of the new station. The Drift Column
Approach (Fang et al., 2012), composed of eight stages, was adopted to

⁎
Corresponding author.
E-mail address: qfang@bjtu.edu.cn (Q. Fang).

https://doi.org/10.1016/j.tust.2018.07.020
Received 29 November 2017; Received in revised form 31 March 2018; Accepted 7 July 2018
0886-7798/ © 2018 Published by Elsevier Ltd.
C. Zhang et al. Tunnelling and Underground Space Technology 81 (2018) 121–128

Chong'wenmen station N
(Line 5)
Movement joint

Tunnelling
direction
Transfer
n
tunnel statio
nnel B e i j i n g 2 )
North tu (Line

Chong'wenmen station nnel

South tu
(Line 2)
24.2 m

18 m 18 m 18 m 18 m
Chainage K216+32 K216+50 K216+68 K216+86 K217+04

Fig. 1. Plan view.

5.1 m
K216+50 K216+68 K216+86
MJ1 MJ2 MJ3

5.9 m
Existing Line 2

1.98 m

Chong'wenmen Station
(Line 5)
11.4 m

24.2 m

(a) Cross section

5.1 m

South tunnel North tunnel

4.1 m
5.9 m

1.98 m

5.9 m
11.4 m

Chong'wenmen Station
(Line 5)

(b) Longitudinal section

Fig. 2. Cross section and longitudinal section.

construct the single-deck part. The construction process is shown in
Fig. 3. The diameter of each steel pipe column is 1.0 m filling with C50
concrete. The distance of two columns in a cross section is 7.2 m, and
the longitudinal distance of two adjacent columns is 6.0 m. A typical
geological profile of the project is shown in Fig. 4. The profile reveals
that the existing twin tunnels were mainly located in silt and sand,
while the single-deck part of the new station is located in gravel, sand
and silty clay. The typical physical and mechanical properties of the
soils, which were obtained from the site investigation report, are shown
in Table 1. The properties of the materials of the support structures are
shown in Table 2. In order to safeguard both the existing twin tunnels
and the new station, pipe roof support was installed above the single-
deck station prior to drift excavation below. The pipe roof installation
adopted the hydraulic pipe jacking machine, as shown in Fig. 5. The
pipe roof support consisted of pipe jacking large diameter steel pipes
(600 mm in diameter, 16 mm thick and 36 m long) horizontally around

122
C. Zhang et al. Tunnelling and Underground Space Technology 81 (2018) 121–128

1
8
4 5 9 11

7 10

6
2 3

(a) Stage 1 (b) Stage 2 (c) Stage 3 (d) Stage 4

12 12

13 13
16 16
14 14

15 15

(e) Stage 5 (f) Stage 6

17
19
18

(g) Stage 7 (h) Stage 8

Fig. 3. Construction process.

Elevation (m) Table 1

44 Soil properties.
42 ķ1 Back fill
ID Soil type Specific Elastic Poisson’s ratio Cohesion Internal
40 ķ Silt weight modulus (kPa) friction
(kN/m3) (MPa) angle (°)
38 Ĺe Clay
36 1 Back fill 16.5 8.0 0.35 5.0 25
Existing Line 2 Ĺ Silt Silt 19.6 9.8 0.35 10.0 24
34 Ĺ1
Silt 19.9 15.1 0.30 24.1 31
32 ĺ3 3 Sand 20.0 19.0 0.27 0.1 25
Water level 4 Sand 20.5 35.0 0.25 0.1 40
30 Sand Gravel 21.0 65.0 0.18 0.1 45
Ļ1 ĺ4
28 1 Silty clay 18.9 12.5 0.45 41.6 17
Ļ 3 Sand 20.6 45.0 0.25 0.1 40
26 Ļ2 Gravel
Chong'wenmen station 1 Gravel 21.5 80.0 0.18 0.1 50
24
(Line 5)
22 ļ3 Sand
ļ Table 2
20 ļ1 Silty clay Support properties.
18 Ľ1 Sand
Gravel Specific weight Elastic modulus Poisson’s ratio
(kN/m3) (GPa)
Fig. 4. Typical geological profile.
Roof pipes 24.0 30 0.15
Existing tunnel 23.0 30 0.20
the periphery of the excavation. The roof pipes were interlocked with structure
each other by lipped channel steel beams and H-section steel beams Primary lining 24.0 21 0.30
welded on the pipes (The center-to-center spacing between two ad- Permanent lining 22.0 30 0.15
Steel pipe column 20.0 210 0.12
jacent pipes is about 0.725 m. Four 32 mm diameter grouting pipes
were installed along each large diameter pipe. The details of two ad-
jacent roof pipes are shown in Fig. 6. The small diameter pipes were
used to inject grout to fill the gaps between the roof pipes and the
ground during pipe jacking. After the installation of each large diameter
pipe, the roof pipes themselves were filled by a low shrinkage cement to

123
C. Zhang et al. Tunnelling and Underground Space Technology 81 (2018) 121–128

Hydraulic pipe jacking machine

Fig. 8. Hydrostatic level cells adjacent to a movement joint.
Fig. 5. Pipe roof installation at construction site.

Distance (m)
Lipped channel steel beams H-section steel beams
-40 -30 -20 -10 0 10 20 30
Grouting pipe 0

Settlement (mm)
-5
m
0 -10
30
0.
-15
-20
-25 North tunnel
-30 South tunnel
0.725 m -35
Existing tunnel MJ1 MJ2 MJ3
Fig. 6. Sketch of roof pipes.

improve its stiffness. The mix ratio of water and cement is 0.6:1 to Tunnel
0.8:1. The grouting pressure was controlled between 0.5 and 0.8 MPa.
Fig. 9. Settlements after pipe roof installation above the side drifts.

3. Monitoring
Table 3
3.1. Monitoring points layout Differential settlement values at movement joints of existing tunnels after pipe
roof installation above new station side drifts.
During the new subway station construction, the deformation of the Movement Movement Movement joint
existing twin tunnels was continuously monitored by an automatic and joint 1 joint 2 3
high precision hydrostatic level system. A capacitive hydrostatic level
Differential settlement 0.6 9.2 10.8
gauge consists of a container body, a capacitive sensor, a transmitter, values of north tunnel
and a communicating pipe. The liquid level in the container varies with (mm)
the elevation of the container body, and the relative position between Differential settlement 0.4 8.6 13.9
the floater of intermediate electrode and a pair of plate capacitors fixed values of south tunnel
(mm)
on the top of the container varies accordingly. The relative displace-
ments of monitoring points can be calculated via the capacitance ratio
attained from the capacitive sensor. The measurement range and the Fig. 7. The first letter “N” or “S” of the monitoring point label indicates
minimum resolution of the hydrostatic level gauge are 50 mm and the “north tunnel” or the “south tunnel”. At each monitoring point, a
0.01 mm, respectively. A benchmark, which was stable and remote hydrostatic level cell was installed on tunnel side wall (Fig. 8). The
from the possible vertical movements, is used as reference datum. The signals observed from the hydrostatic level cells were sent back to the
monitoring points layout along the existing twin tunnels is shown in

MJ1 MJ2 MJ3 MJ4

N1 Center line N2 N3 N4 N5 N6 N7 N8 N9 N10 N11 N12 N13 N14

(S1) (S2) (S3) (S4) (S5) (S6) (S8) (S10) (S11) (S12) (S13) (S14)
12.0 m 9.0 m 9.0 m 4.5 m 4.5 m 4.5 m 4.5 m 9.0 m 9.0 m 9.0 m

Monitoring point

Fig. 7. Layout of monitoring points.

124
C. Zhang et al. Tunnelling and Underground Space Technology 81 (2018) 121–128

Table 4 tunnel, were 31.3 mm and 27.1 mm respectively. The differential set-
Gap values between tunnel floor and track bed of existing tunnels after pipe tlements of the three movement joints of each tunnel are shown in
roof installation above new station side drifts. Table 3. The maximum differential settlement values of the existing
Movement joint Movement joint Movement joint 3 north and south tunnels, reported at movement joint 3 of each tunnel,
1 2 were 10.8 mm and 13.9 mm respectively. Due to the differential set-
tlements, some separations were produced between the tunnel floor and
Gap values in north 0 3.4 5.9
the track bed, which in turn led to the track irregularities. The thick-
tunnel (mm)
Gap values in south 0 3.6 12.7 nesses of separations of the movement joints are shown in Table 4. The
tunnel (mm) maximum separation thickness was 12.7 mm at movement joint 3 of the
south tunnel. In order to reduce the vibration of the trains running
inside the existing tunnels and fulfill the operational safety require-
Table 5 ments, the trains speed was significantly reduced from 80 km/h to
HSC grouting material properties. 40 km/h at this part of the project, which resulted in a remarkable
Water cement ratio 0.8 decrease of the transport capacity of Beijing Subway Line 2.
Relative viscosity (s) 23.8
Fluidity (mm) 350
Syneresis rate (%) 0.8
4. Compensation grouting for existing tunnels
Initial setting time (min) 75
Final setting time (min) 98
7 days compressive strength (MPa) 12.3 In order to restrain the settlement of the existing tunnels and drive
28 days compressive strength (MPa) 24.6 the trains inside the existing tunnels back to their normal speed, com-
pensation grouting was adopted to uplift the existing tunnels.
Compensation grouting is based on the idea of replacing the volume of
central monitoring system so that measurements could be taken on a ground loss with a volume of grout to compensate (reverse) the ground
24-hour basis. Readings were generally taken at 30-minute intervals but or underground structure settlement (Mair and Hight, 1994). It is ap-
with increased frequencies during construction under the existing plicable to various soil conditions, and the grouting parameters can be
tunnels. flexibly adjusted to cater to various ground conditions with reference to
The hydrostatic level system was very sensitive to the ambient in- the measured settlements of the existing structures (Kolymbas, 2005).
terferences, such as the passage of a train nearby a monitoring point A special grouting material HSC (High Strength Cement) that was
(Fang et al., 2015). Therefore, representative data relatively free of composed of a special sulphoaluminate cement clinker, gypsum, gan-
interferences, which were obtained during the non-service time of Line ister sand, and water reducing flocculating agent was selected for the
2 after midnight, were selected from the huge volume of automatically uplifting grouting. The properties of the special grouting material HSC
recorded data to represent the daily settlement magnitudes of the ex- used for the uplifting grouting, which are obtained from the laboratory
isting tunnels. tests, are shown in Table 5. The longitudinal and cross-sectional layouts
of the grouting pipes under the existing tunnels are show in Fig. 10 and
3.2. Monitoring results after pipe roof installation above new station side Fig. 11 respectively. The diameter of the grouting pipe was 32 mm. A
drifts total of 56 pipes, 2.5 m long each (Type I), were installed under the
existing tunnels. And totally 98 pipes, 3 m long each (Type II), were
The settlement curves of the existing twin tunnels after the pipe roof installed on two sides of each twin tunnel. A fast setting 2-component
installation of stage 5 (Fig. 3) are shown in Fig. 9. It is noted that, due to grout was adopted for injection. The grout was first injected through
the presence of the movement joints of existing tunnels, there are jumps the 3 m long pipes to reinforce the ground around the existing tunnels.
of the settlement curves of the existing tunnels. Moreover, the existing Some grouting curtains, with strong strength and low permeability,
tunnel sections adjacent to the new station mainly tilted as rigid bodies. along the twin tunnels were accordingly formed to reduce the outflow
After the pipe roof installation, the maximum settlement values of the of the subsequent injected grout, which also served as confinement to
existing north and south tunnels, reported at movement joint 2 of each facilitate the formation of the uplift pressure generated by the

South tunnel North tunnel

Tunnelling direction N

Grouting pipes

Type I pipes Type II pipes Type I pipes Type II pipes Type I pipes
12 m 6m 4.5 m 6m 6m

Chong'wenmen station
(Line 5)

Fig. 10. Longitudinal section of grouting pipes.

125
C. Zhang et al. Tunnelling and Underground Space Technology 81 (2018) 121–128

MJ2 MJ3

Grouting pipes Existing Line 2

Pipe roof

Fig. 11. Cross section of grouting pipes.

Distance (m) 20 3.0

-40 -30 -20 -10 0 10 20 30 2.5
Uplift of N6
0 15

Settlement rate (mm/d)

Uplift rate of N6 2.0
Settlement (mm)

-5
Settlement (mm) 1.5
-10
-15 10
1.0
-20 Before uplift grouting
-25 During uplift grouting 0.5
5
-30 After uplift grouting 0.0
-35
North tunnel MJ1 MJ2 MJ3 0 -0.5
-1.0
-5 -1.5
Tunnel 0 5 10 15 20 25 30
Time (d)
(a) North tunnel (a) North tunnel
Distance (m) 20 3.0
-40 -30 -20 -10 0 10 20 30 2.5
Uplift of S6
0 15 Uplift rate of S6

Settlement rate (mm/d)

2.0
Settlement (mm)

-5
Settlement (mm)

-10 1.5
-15 10
Before uplift grouting 1.0
-20
During uplift grouting 0.5
-25 5
After uplift grouting
-30 0.0
-35
South tunnel MJ1 MJ2 MJ3 0 -0.5
-1.0
-5 -1.5
Tunnel 0 5 10 15 20 25 30
Time (d)
(b) South tunnel (b) South tunnel
Fig. 12. Settlements associated with uplift grouting. Fig. 13. Settlement development and uplift rate of existing tunnels during
compensation grouting.

subsequent grouting. With the help of the grouting curtains, the existing
tunnel sections can be uplifted by grout injections through the 2.5 m
long pipes under the existing tunnels. The compensation grouting was
performed during the non-service time of the existing Line 2. The
grouting pressure and grouting rate were strictly controlled according
to the monitoring results obtained from the hydrostatic level system.
The compensation grouting of this project was started on December
28, 2004, which lasted for about 50 days. The settlement curves along
the existing north and south tunnels before and after compensation
grouting are shown in Fig. 12. The uplift values of the monitoring
points N5, N6, S5 and S6 were 16.11 mm, 16.00 mm, 14.11 mm and
16.07 mm respectively. The settlement development and daily uplift of
the points at movement joint 2 of the existing north and south tunnels

126
C. Zhang et al. Tunnelling and Underground Space Technology 81 (2018) 121–128

Date
02/01/2004 05/01/2004 08/01/2004 11/01/2004 02/01/2005 05/01/2005 08/01/2005 11/01/2005 02/01/2006
0

-5

Settlement (mm) -10 Step1

-15
Step15
-20 Step 5 Step 16 N6
-25 Step 8 Step 19 N5
Step 9 S6
-30
S5
Step 12
-35

Fig. 14. Settlement development of the existing twin subway tunnels.

Construction step
0 2 4 6 8 10 12 14 16 18 20
0
-5 Numerical predictions
Measured settlements
11.4 m

-10
Settlement (mm) -15
With grouting
-20
-25
24.2 m -30 Without grouting
-35
(a) Cross-sectional profile of Chongwenmen subway station of Line 5
-40
Fig. 16. Settlement development of the existing subway tunnel derived by the
Existing tunnels monitoring point N6.

to new subway station construction, three-dimensional (3D) numerical

Soils
New subway station
(b) Layout of existing twin subway tunnels and Chongwenmen subway station
Fig. 15. A typical finite element mesh used in the numerical simulation.
due to compensation grouting are shown in Fig. 13. The daily uplift
values of each monitoring point was controlled in 1.5 mm/day. Due to
the compensation grouting, the maximum differential settlement value,
reported at movement joint 3 of the south tunnel, was recovered from
13.9 mm to 7.2 mm. Meanwhile, the maximum separation thickness
between the tunnel floor and the track bed, reported at movement joint
3 of the south tunnel, was reduced from 12.7 mm to 7.0 mm, which had
no effect on fracture and damage to the existing tunnel structures. After
compensation grouting, the trains speed inside the existing tunnels was
resumed. Therefore, the safety of the operating twin subway tunnels
was guaranteed during the construction of the new subway station.
After compensation grouting, the subsequent settlements of the
existing tunnels were insignificant, which was mainly attributed to the
fact that the pillar had been sufficiently reinforced (Fig. 14).
To simulate the behaviors of the existing twin subway tunnels due
analysis was conducted using the finite difference program, FLAC3D. In
order to balance accuracy and computational efficiency, the dimensions
of the numerical model were 120 m (length) × 50 m (width) × 55 m
(height). A typical finite element mesh of the existing twin subway
tunnels and the new subway station is shown in Fig. 15. The model was
free at the top, and the displacements were zero in all directions at the
bottom. The normal displacements of the four vertical boundary sur-
faces were restrained. The soil layers distribution of the project was
considered in the numerical simulation. Each soil layer was assumed to
be a homogeneous, isotropic and elastic-perfectly plastic material
obeying the Mohr-Coulomb failure criterion. The tunnel support
structures were assumed to be linear elastic materials. The mechanical
parameters of the soils and the support structures were shown in Tables
1 and 2.
The calculated and the measured settlements of the monitoring
point N6 of the existing subway tunnels are shown in Fig. 16. The
calculated and the measured results are close before compensation
grouting (Before step 12), which verify the availability of the numerical
simulation. As the grouting process is difficult to be included in the
numerical simulation, the calculated and the measured settlement re-
sults diverge after step 12. But we can still use the numerical simulation
results to evaluate the effect of the compensation grouting measure.
5. Conclusions and discussions
This paper presents a case of the Chongwenmen subway station of
Beijing Subway Line 5 excavated below the existing Line 2 with a
minimum pillar width of 1.98 m. The key findings are as follows:

127
C. Zhang et al.
(1) The large-diameter pipe roof installed before excavation produced
much disturbance to the surrounding ground. It was the first and
last time that this kind of reinforcement measure was adopted in
such adjacent tunnelling project in Beijing.
(2) The compensation grouting was proved to be an effective measure
in compensating ground loss and uplifting existing tunnel. There
are two keypoints of compensation grouting. First, the pillar be-
tween the existing and new tunnel should be sufficiently reinforced
before uplifting. Second, the grouting pressure should be applied to
the rigid pillar instead of the bottom structure of the existing
tunnel.
(3) The settlement curves of the existing tunnels jumped at the loca-
tions of the movement joints. In addition, the existing tunnel sec-
tions affected by new station construction mainly tilted as rigid
bodies, due to its higher stiffness than the surrounding soils.
Acknowledgments
The authors acknowledge the financial support provided by the
Chinese National Natural Science Foundation (Grant No. 51678035)
128
Tunnelling and Underground Space Technology 81 (2018) 121–128
and Beijing Municipal Natural Science Foundation of China (Grant No.
8172037).
References
Cooper, M.L., 2001. Tunnel induced ground movements and their effects on existing
tunnels and tunnel linings. PhD Thesis, University of Birmingham, UK.
Cooper, M.L., Chapman, D.N., Rogers, C.D.F., Chan, A.H.C., 2002. Movements in the
Piccadilly line tunnels due to the Heathrow Express construction. Géotechnique 52
(4), 243–257.
Fang, Q., Zhang, D.L., Wong, L.N.Y., 2012. Shallow tunnelling method (STM) for subway
station construction in soft ground. Tunn. Undergr. Space Technol. 29, 10–30.
Fang, Q., Zhang, D.L., Wong, L.N.Y., 2015. Effects of twin tunnels construction beneath
existing shield–driven twin tunnels. Tunn. Undergr. Space Technol. 45, 128–137.
Kolymbas, D., 2005. Tunnelling and Tunnel Mechanics. Springer-Verlag, Berlin,
Heidelberg.
Kimmance, J.P., Lawrence, S., Hassan, O., Purchase, N.J., Tollinger, G., 1996. In:
Geotechnical Aspects of Underground Construction in Soft Ground: Proceedings of
the International Symposium. Taylor Francis Group, Rotterdam, pp. 707–712.
Mair, R.J., Hight, D.W., 1994. Compensation grouting. World Tunnel. 361–367.
Standing, J.R., Selman, R., 2001. The response to tunnelling of existing tunnels at
Waterloo and Westminster. In: Burland, J.B., Standing, J.R., Jardine, F.M. (Eds.),
Building response to tunnelling: case studies from construction of Jubilee Line
Extension, vol. 2, case studies. Thomas Telford, London, UK, pp. 509–546.