Construction of Underground Works in Shanghai, P.R.C. Gao Quqing
AbatraceFor nearly thirty years, Shanghai has been constructing underground works designed to improve the city’s infrastrrccture. A number ofprojects were undertaken usingshield tunnelling method in order to determine whether this method could be used successfully in Shanghai’s geology, which is characterized by silt and silty orpuddly clay. These undergroundprojects have included subway tunnels, sewage and water supply tunnels, road tunnels, and offshore discharge tunnels. This paper describes eight subsurface projects in Shanghai that have used the shield tunnelling method successfully.
hanghai is the biggest city in China, with a population of 12 million people. Intheearly1960s, Shanghai began to construct a number of unhground works to improve the city’s ir&astru&m. The underground works completed thus far include two underwaterroadtaumelshnkingtheeastandwest bankaoftheHuangpuRiver,andof&hore &charge, intake and cooling water tunnels for steel and chemical complexes in the suburbs of Shanghai The Shanghai metro was under construction for many years. A section of metro tunnel was constructedin 196667. In 1981-84, experimental tunnels were constructed to determine the feasibility of shield use in Shanghai’s soil. Based on the successful construction of the experimental tunnels, the plan layout of the No. 1 metro line and the design of metro stations was approved by the government, and three stations are now under construction. Also under construction are sewage tunnel and water supply tunnel projects, which are being constructed by the pipe-jacking technique.
Geology of Shanghai The geology of Shanghai consists mainly of alluvial deposits, comprising four layers:
Presentaddress:Prof. Gao Quqing,Southwest Jiaotong University, Jiu Li Di, Chengdu, Sichuan, 610031, People’s Republic of China.
R&um6-Pendant environ trente ans, Shanghaii a construit de.9 ouvmges souterrains darts k but d’ameliorer l’infrastructum de la vilk. De nombmuxprojets firertent entrepris en utilisant la m&hook h &ran de fqon a &terminer si cette m&hoaizpouvait &re utilisee avec succes compte tenu de la geologk de Shanghai, qui est caracterisee par de la vase ou o!.elhrgik boueuse. Cesprojets en soutermins ont inclus des tunnel de metro, le tout h l’egout et des tunnels d’alimentation en eau, des tunnels routkr, et o!e o!&ersement au large. Cepapier d4cri.t huitprojets en soutermin hShangha~qui ont utilW avec succ&3la m&ho& a ecmn.
1. Yellowish-brown silty clay, 2 to 4 mthick, with groundwater 0.7 to 1.5 m below the ground surface. 2. Grey silt, up to 20 m thick, interbedded with layers of fine sand and sandy clay. It has a high water content, high void ratio, high compressibility, and a permeability of approximately 10%&ec-l. 3. Grey siltJpuddly clay, 20-24 m thick, interbedded with layers of silt and fine sand. This layer is characterized by a high void ratio, low shear strength, and high permeability. When the void ratio exceeds 1.3, the water content exceeds the liquid limit. This puddly clay forms the main layer of Shanghai ground that underground works and metro tunnels pass through. 4. Dark green silty clay, up to 8 m thick, characterized by a low void ratio, low water content, high shear strength, and low permeability.
Case Histories of Underground Works in Shanghai From the early 1960s to 1984,9 km of underground works were excavated by shield in Shanghai. Three km were excavated with the aid of compressed air; 1.5 km, with the aid of dewatering with eductor wells; and the remaining 4.5 km, by blind or semi-blind shields. Theremainderofthispaperpresents brief case histories of eight shielddriven tunnels in Shanghai. Table 1 lists statistics for the eight shield-driven tunnel constructed in Shanghai that are discussed in this
paper. Table 2 gives typical soil properties for these tunnels. I. Trial Tunnel (1963-64) The purpose of this tunnel was to determine the feasibility of shield use in the Shanghai soil. The 68-m-long tunnel (4.2.-m-outer-diameter) was constructed by the traditional shield method. The tunnel was excavated in two stages: the 6rst stage, in silt; and the second stage, in soft clay (see Fig. 1.). An intermediate shaft was provided for transferring the shield fivm one stage to the other. A conventional shield with face jacks (Greathead shield) was adopted. Dewatering and compressed air measures were adopted in different sections along the tunnel to ascertain the effectiveness of these measures in Shanghai ground. Conventional single-shell, bolted, precast concrete segments were used for the lining. For waterproofing, a mixture of coal tar and epoxy resin was used to filI the joints. The trial construction was carried out with good results. No serious problems occurred during construction.
2. Subway Tunnel (1966-67) In the early 196Os, when Shanghai
wasSrstconsideringbuildinga subway system for city passenger transportation, two 600-m-long, single-track parallel tunnels were excavated. A mesh shield was used for the excavation. In this method, a steel mesh
OFS (=d!) Sand
Cohesion (kg cm )
Plastic Limit (%)
Liquid Limit (%)
Bulk Unit (wt. ton/
Table 1. Soil properties along several tunnel stretches in ShanghaL
Table 2. Comparative characteristics of shield-driven tunnels constructed in Shangshai since 1965.
Type of Erector
Outer Dia. of Shield
Type of Shield
Subway Tunnel (section 1)
2 x 599 m
Offshore Intake Tunnel
Cooling Water Discharge Tunnel
2 x 314
Name of Tunnel
Experimental Subway Tunnel
frame is mounted on the cutting edge ofthe shield;when the shieldisjacked forward, the ground oozes into the shield through the mesh opening. When the shield stops, it stabilizes the working face. The first length of the t - n n e l was aided by dewatering measures (see Fig. 2). Dewatering was very effective in the silt stratum, and the shield could even be used in the soft clay stratum without the need for compressed air. However, in order to gain more experience, most of the t - n n e l was driven with a bulkhead in front under compressed air. This t - -nelllng experience was very successful. The lining utilized five single-shell, precast concrete segments: two ordi-
nary segments, two top segments, and one key. For waterproofing, a mixture of coal tar and epoxy resin was used to fill the joints. In 1964, the construction was stopped due to a lack of funds.
3. Dapulu Road Tunnel (1966-67) The 2.7-ban-long ])apuln Road Tunnel was the first underwater tlmnel
across the Hwangpu River. The road is 7.07 m wide to accommodate doublelane traffic. A longitudinal profile is shown in Figure 3. The approaches at both ends of the tnnnel were built by the cut-and-cover or caisson method. The center portion of the t , nnel, under the river, was 1320 m long, with a 10-
m outer diameter; it was constructed by shield with an adjustable b-lkhead, so that it could be used as a blind, semiblind, or mesh shield. The shield diameter was 10.3 m. At the first stage of shield excavation, a mesh was installed. ARer excavating a few hundred meters, the t~mnel reached the soft clay strata, and the bnlkhead was totally closed (i.e., blind shield). The ground under the river bed heaved up to 3 m when the shield passed through it. During excavation, maint~inln g the shield in line and grade was difficult because it had a tendency to rise and to move backward when it stopped. To help adjust for this movement and to control the direction of the shield movement, the ground around the shield was allowed to ooze in. After the shield reached the other bank, the final deviation was estlmAted to be 2 cm. After the shield reached the east bank, it was driven with the b-llchead removed (i.e., mesh shield), aided by dewatering or compressed air.
segments, including five ordinary
segments, two top segments, and key Figure 1. Longitudinal profile and geological features of the trial tunnel built in Shanghai.
Volume 5, Number 4, 1990
segment for each ring. The width of ring was 900 ram; each segment weighed 5 T. and was 600 mm thick Epoxy resin mortar was used for waterproofing.
Heoder adopted, utilizing a 0.63-m-thick slurry wall.
'"'" I""': : ' ':""
. . . .
I Ejector wett
Type/ k I
~ @ ~ "
Figure 2. Dewateringarrangement for the Shanghai subway tunnel.
Because some water leakage occurred after the ~lnnel was completed, post-grouting was required.
5. Cooling Water Discharge Tunnels (1980-82) This scheme, built for Boo Shan Iron and Steel Complex, comprises two 314-m-long tunnels and one 135-mlong tunnel for cooling w a t e r and sewage discharge. The tunnels were excavated by a hydromechanical shield (see Fig. 5) assisted by dewatering, without pressurizing the chamber. A double lining was used to resist corrosion by chemicals in the cooling water. Bolted precast segments were used for the primary ]in'ing; the secondary lining was ofcast-in-situ concrete, for a total thickness of 450 ram.
4. Offshoredischargeand intake tunnels (1974-75) The discharge tunnel built for the Shanghai Petro Chemical Complex is 1 km long, with an outer diameter of 3.6 m (a longitudinal profile is shown in Fig. 4). The tunnel was excavated primarily by a blind shield in soft clay. However, during the first part of the tunnel excavation, the bulkhead was partly removed to allow the clay to ooze in so that the shield could pass under the sea wall without disturbing it. Most of the tunnel was lined with singleshell, unbolted precast segments with tongue-and-groove joints; six identical trapezoidal segments sealed with polyurethane foam strips were glued into the grooves. The last portion of the tunnel was lined with bolted, mild steel segments to facilitate the connection between the tunnel and shafts. The two intake tll n nels, each 1.6 k m long with an outer diameter of 4.2 m, were excavated by blind shield. The same type of lining and method of waterproofing used for the discharge tunnel were used for the intake tunnels.
No 6 Ventibltion stuart ~
HO2 VeMiL~tlo~ $•ft I
6. Experimentalmetro tunnel (1981-84) This tunnel, built in the Xuziahwy district, was built when a metro for Shanghai was again under consideration (tunnelling for a first metro line had been stopped in 1964 [see case history #2, above]). This experimenta. 1 tunnel, built to gain additional experience, consists of two sections: 1. From shaft 151 to shaft 101, shield excavation was used. 2. From shaft 101 to the Triangle Area, the slurry wall method was
No. 3 VentlLaton shaft A r r i v a l shaft
Soft cloy with ~hln
No S Vent~Lotion shaft
•j OrownishySlLl~v I liLtysorKI ¢/ .
T opL oye , .
The ground condition was puddley clay, with horizontal layers of fine, mildly sandy clay. A mesh shield with an outer diameter of 6.42 m w a s u s e d . Theembedded depth from the top of the shield was 6 m. Two 565-m-long, parallel tunnels with a c-c of 10.8 m were excavated. Two types of lining--one utilizing six segments; the other, four segm e n t s - - w e r e used. For waterproofing, the joints between segments were filled with chlorinated rubber. One hundred or more monitoring points were established at the site to measure surface subsidence. Table 3 gives the values and settlement ratios and ratios of total subsidences for different times elapsed and for different types of grouting. The results of the subsidence monitoring showed that: 1. I f no grouting follow alter the shield places the lining, the soil will collapse into the void, and collapse subsidence will be completed within two weeks. 2. Consolidation subsidence can last for a year or more. 3. The maximum width of the subsidence area was 20--25 m on each side. The subsidence area was divisible into two parts: Direct part: H = 100 ram, transverse gradient 0.6". Indirect part: 0.6", t r a n s v e r s e gradient. The following conclusions may be drawn from the experience gained from the construction and monitoring of this ~mnel:
1. Subsidence was caused mainly by the construction void. The grouting materials should have low shrinkage and high strength, and the tu n n el must be grouted in a timely manner. For unstabilized soil, the shield advance and grouting should be performed simultaneously. 2. Controlllngthe speed of the shield advance and the inlet amount of ground soil will control the upheaving and subsidence of the ground. 3. Sand wells dug from the surface could be used to release the super pore pressure of the soil in order to quicken soil consolidation and hardening of the grouting material. 4. A slurry wall could be used to intersect the water so t h a t subsidence due to water drainage could be controlled. 5. Consolidation grouting could be used to control the ground during later periods of subsidence.
Figure 3. Longitudinalprofile of the Dapulu Tunnel. 3 6 0 TUNNEJffJNG AND UNDERGROUND SPACE TECHNOLOGY
Volume 5, Number 4, 1 9 9 0
Onshoreshof~ ~ / /
HHWL * ~ § Se°vlevel ~l
___~__--_- v.,,,<<,,oo,,.,..<,,. e
Figure 4. Longitudinal profile of the outfaU tunnel built for the Shanghai Petro Chemical Complex.
L__,o,,..o° High pressurewoter
- - . . Ira'S,,_
." . . . . . . .
Figure 5. Hydromechanical shield with pressurized chamber.
7. Subaqueous pipeline (1981) This pipeline was built by the pipejacking technique for an oil refinery in Zejian Province near Shanghai. Ninem-long sections of 24-ram-thick steel pipe (2.6-m inner diameter) were used for the 581.9-m-long pipeline. A 19-m-dia. starting shaft was built on one bRnk of the Yong River, using
six 300-T. jacks. An ll-m-dia, arrival shaft was excavated on the other bank of the river. A profile of the project is shown in Figure 6. The pipeline was laid at a 17-m embedded depth along the river bank and at 8.3 m along the river bed on a 0.1% upward slope from the starting shaft. Silty sand was encountered for
the first 180 m; the next 180400 m consisted of mucky loam; and the remainder was of mucky clay. The shield was a double-hinged, triple-section installed in front of pipe. Bentonite lubricant was injected at the rear of the shield. The soil in the working face was washed out by a water gun as slurry, which was then pumped up to the surface. Compressed air was used in the quicksand section; six hours were required for each working cycle to jack a 9-m section forward from the shaft. Five intermediate jacking stations, at 60-mto 120-mintervals, were established during the later stages of jacking; 40 jacks, each weighing 25 T. and having a 10-cm stroke, were adopted for each station. Rotation of pipe is unavoidable during jacking, but was corrected through the use of counterweights. To break through the shaft wall without deviation is important because any initial deviation will have a negative effect on the later jacking process. This process also could be done with jack pressure monitoring. For plastic soils, a mesh shield was used. For quicksand and seft ground, a shield with a pressurized flushing cabin was used.
8. Yenan East Road Subaqueous Tunnel (1988-89) This t - n n e l is the second tunnel to be built under the Hwangpu river. Alongthe bAnkR of the river, thetllnnel passes through mildly clayey soft; un-
Table 3. Values and settlement ratios for different grouts and different times elapsed after grouting.
Ratio of Total Subsidences
Total Subsidence for Different T y p e s of G r o u t i n g
380% Grouting Three Times (H3)
No Grouting (H0)
Ratio after 50 Days
230% Grouting One Time (H1)
Volume 5, Number 4, 1990
Ratio after 50 Days
Ratio after 50 Days
Figure 6. Longitudinal profile of the pipe-jacking project under the }Tong River.
der the river, the soil is puddley clay. Because the 500-m-wide river is the return point of 10,000-T. ships, this section of the tunnel was constructed by the shield method. Three shafts were provided: one at west H w a n g p u a n d two at east Hwangpu. From shaft No. 1 to shaft No. 3 is a circular tunnel 1416 m long. From shaft No. 1 to the west and from shaft No. 3 to the east are rectangular tunnels with open-cut approaches. The East Hwangpu tnnnel was open cut with steel sheetpiles or
reinforced concrete piles. The West Hwangpu tunnel was cut-and-cover, using a slurry wall and piles. The grade was 2.4%-3.5%; road clearance was 7.5 m x 4.5 m. The outer diameter of the tnnnels was 11 m; the inner diameters were 9.9 m. Figure 7 shows the longitudinal profile and crosssections of the tunnel. The shield had an outer diameter of 11.20 m, an inner diameter of 11.1 m, and was 8.4 m long. Forty-eight hydraulic jacks, with a total force of 105,949 KN, were used. The construction started from east Hwangpu, and
No. 1 Shaft
, I il
traffic,and the automatic alarmsystem.
Conclusion These case histories demonstrate that different methods of shield excavation and different types of lining waterproofing can be used successfully in underground projects in Shanghai. 
No.2 Shaft " ~
progressed as follows: 1. For the No. 3 to No. 2 shaft (437 m), excavated in mildly sandy silt, a hydromechanical shield was used, and the water level was lowered. 2. For the under-river section (519 m) in puddley clay, the shield was l~shed and squeezed forward as a controlled an*aunt of clay was allowed im 3. The west Hwangpu section (520 m) was excavated in puddley clay under compressed air. The clay allowedin was strictly controlled to prevent subsidence. The tnnnel lining comprised eight reinforced concrete segments. Each segment is 1 m wide, with a rib thickness of 0.55 m, and a weight of 5.3 T. The segments are connected longitudinally by three bolts, and circumferentially by 32 bolts. For waterproofing, rubber gaskets were inserted and then injected with a waterproofing agent. For operating purposes, a computerized control room handles electricity, lighting,ventilation,
No. 3 Shaft III I
"l 200., I 20 ., I
i7~o0~~1 i ~
il - II
III - III
Figure 7. Longitudinal profile and cross-section of the }renan East Road Tunnel.
362 TUNNELLINGAND UNDERGROUND SPACE TECHNOLOGY
Volume 5, Number 4, 1990