Long-term settlement behavior of ground around shield tunnel due to leakage of water in soft deposit of Shanghai

Abstract

The lining of shield tunnel is usually composed of segments, in which the joints, cracks, and the grouting holes (hereafter called lining deficit) exist. During the long-term running, soils and groundwater may leak from these kinds of lining deficit. The leaking of soil and groundwater causes the long-term ground loss around tunnel and thus results in the settlement of ground surface. This paper aims to analyze the impact of the leakage of groundwater through segments on the long-term settlement of ground surface. The adopted analytical method is based on the theory of groundwater seepage by using numerical simulation. The analyzed results show that settlement of ground surface increases gradually with the increase of the leaked volume of tunnel segments. When the leaked volume was unevenly distributed, differential settlement occurred locally. Comparative analysis by changing the leaked volume was conducted. The results reveal that there is a linear relationship between settlement and leaked volume when the leaked volume was controlled within the allowable limit.

Keywords

lining deficit / leakage of water / shield tunnel / settlement / soft deposit

Introduction

Shanghai is located on the deltaic deposit of the Yangtze River [1]. The thickness of the soft deposit is generally from 100 m to 300 m with the maximum of 400 m in some locations. The quaternary soft deposit of Shanghai is composed of aquifers with high groundwater head pressure and aquitards with low permeability. This formation of the groundwater system in Shanghai is a so-called multi-aquifer-aquitard system (MAAS) [1]. Shanghai is suffered from land subsidence due to groundwater withdrawal [1-4]. The metro tunnels of Shanghai are buried in the depth from 10 m to 40 m. The corresponding soil layer is silty clay layer or sandy silt layer [1,5].

The long-term ground movement and their effect on tunnel lining behavior are becoming increasingly important. Some researches indicate that ground movement induced by tunneling is significant and can continue for many years [6,7]. Shirlaw illustrated that ground surface continued to subside after tunnel excavation in clay soils and the long-term settlement accounted for 30%~90% of the total settlement [8]. O’Reilly et al. reported monitoring of long-term settlements over a period of 11 years for a 3 m diameter tunnel constructed in normally consolidated silty clay in Grimsby and found a final equilibrium reached at the tenth year [9]. Mair [7] and Shin et al. [10] reported that leaking of groundwater through tunnel lining may cause serious additional inner stress in the lining and serious settlement of ground around tunnel in long term. The long-term settlement of tunnel in soft deposit is becoming a serious problem for cities in coastal region of China. Differential settlement at some locations of Shanghai metro line No. 1 have been measured after construction for five years and the accumulated tunnel settlement of metro line No.1 has reached 350 mm after it was used for 15 years [5,11]. This phenomenon also exists in other metro lines of Shanghai. Differential settlement of tunnel may result in the distortion of track and thereby impact the operating safety of train.

The causes of tunnel settlement and deformation in soft deposits have been discussed, including land subsidence, variation of ground water table, leakage of tunnel, dynamic response of train, etc. [11,12]. Tunnel leakage is an important factor that cannot be neglected. Shanghai metro is constructed by shield method, in which the lining is consisted of 6 segments. Fieldwork shows soil and ground water may leak from the joints, cracks, and the grouting holes of segments [13]. The leaked volume is initially small and the tunnel deformation is usually neglected. When the leakage develops, original differential settlement may aggravate and will lead to further leakage of water even soils. The leakage of soil and groundwater causes long-term ground loss around tunnel and thus results in the settlement of ground surface over the tunnel. Wu et al [13]. has discussed the influence of leakage on long-term settlement. However, the boundary effect is not considered felicitously and the behavior of uneven leakage is not considered [13].

In this paper, numerical simulation is adopted to analyze the impact of segmental leakage on long-term settlement behavior of ground surface. The analyses are undertaken in connection with the project of one sectional tunnel of Shanghai metro line No. 1. Based on the assumption of different leaked volumes, numerical simulations are conducted to summarize the law of the long-term ground settlement induced by the leakage of groundwater through segmental lining.

Soil properties in field

The sectional tunnel of Shanghai metro line No. 1 includes two parallel tunnels (the up line and the down line) with a length of 1250 m. The external and inner diameter of the tunnel is 6.2 m and 5.5 m respectively. The lining of the tunnel is assembled by 6 reinforced concrete segments with a width of 1 m. The buried depth of the tunnel is 7-15 m under the ground surface.

An illustrative soil profile with some available physical and mechanical properties is given in Fig. 1. Generally shield tunnels in Shanghai are constructed at the depth from several meters to 40 m under the ground surface. The corresponding soil layers are the upper mucky clay layer and the lower mucky silty clay layer with characteristics of saturated, flow to soft plastic clay with high compressibility and sensitivity and low strength, long stabilizing time and big settlement after being disturbed [15]. According to the Local Standard of Engineering Construction in Shanghai [16], the groundwater level is relatively shallow and the average level is about -0.5 to -0.7 m.

Analytical method

Basic equation for 3D groundwater flowed in the saturated porous medium is expressed as the following equation [17]:

(1){\displaystyle \frac{\partial }{\partial x_i}}\left(K_{ij}{\displaystyle \frac{\partial n}{\partial x_j}}\right)-q={\displaystyle \frac{\partial }{\partial t}}(S_s),

i,j=\mathrm{1},\mathrm{2},\mathrm{3}(\mathrm{1}:x,\mathrm{2}:y,\mathrm{3}:z)

where K_ij = hydraulic conductivity, h = hydraulic head, q = volume of seepage, S_s = water storage ratio, t = time.

According to the theory of consolidation, the compression of soil layer can be calculated by the coefficient of specific storage and the effective stress:

(2)S_t=m_v\mathrm{\Delta }{\sigma }_{z}H={\displaystyle \frac{S_s}{y_w}}\mathrm{\Delta }{\sigma }_{z}H=S_s\mathrm{\Delta }hH,

where ▵h is the variation value of hydraulic head, H is the thickness of soil layer.

Based on this formula, compression of soil layer is available and the value of land settlement can be calculated accumulating the compression of soil layers. The consolidation compression by using the proposed approach was compared with the results from Terzaghi’s 1D consolidation theory. It was confirmed that the proposed approach is accurate enough to solve engineering problem [18].

Calculation model and settlement prediction

A 3D finite element model was used to predicted tunnel long-term settlement, which considered groundwater seepage and land subsidence in one process. Figure 2 illustrates the calculation model which is 1750 m in length, 900 m in width and 85 m in height. The tunnel is 1250 m in length. The distance between the two parallel tunnels was 11.15 m and the buried depth of the tunnels was 13 m. It is difficult to confirm the leaked volume in the field because monitor data was always quite limited. Even though the leaking location is monitored, the leaked volume of leakage water or seepage water is uncertainty. In this model the leaked volume was assumed to be 0.1 L/(m²·d^-1), which is the allowable leaking value in Shanghai. The leaked water was averagely assigned on 24 places along the tunnel. To simulate leaking, pumping ground water at leaking location is adopted accordingly. The volume pumped was equal to that of leaking.

Figure 3 plots the 3D finite element model. The distribution of soil layers and the soil properties were based on the data in Fig.1. The variation of the thickness of each soil layer was ignored. The phreatic water level and confined water level were -0.5 m and -3.0 m respectively. Since the scale of the model established is large enough, the boundary effect is extremely small and therefore the boundary water head is assumed to be constant.

The long-term settlement contour at the ground surface after 10 years is depicted in Fig. 4. If the leakage speed keeps 0.1 L/(m²·d^-1) (case 1), a maximum settlement of more than 90 mm will occurred at the ground surface 10 years later. There will be 1–2 mm after 10 years at the place 200 meters away. As shown in Fig. 2, section

I is longitudinal section through the axis of the downward line. The corresponding long-term settlement profile after 10 years is shown in Fig .5.

Long-term settlement pattern

Comparative analysis has been conducted by varying the leaked volume at each location. The leaked volume in case 1 was 0.1 L/(m²·d^-1), corresponding to 0.09 m³/d (hereinafter labeled as L) at each place. Cases from 2-1 to 2-6 changed the leaked volume gradually, as detail shown in Table 1.

The corresponding long-term settlement profiles at the ground surface for seven cases (including case 1) are plot in Fig. 6. It is not difficult to find that with the total volume of leakage increases, the magnitude of long-term settlement rises gradually. If the velocity of leakage keeps 5 L, large settlement occurs with a value of about 600 mm.

The relationships between magnitude of ground surface squat at P₁, P₂, and P₃ (see Fig. 4) and total leaked volume after 10 years are depicted in Fig. 7. All of them nearly present a linear relationship.

Base on case 1, case 3 includes 2 cases to simulate the different extents of uneven leak of the downward line by varying the leaked volume of the middle two locations of the downward line Q₁ and the leaked volume of other location Q₂ (Q₁, Q₂ refer to Fig.2). Case 3-1 assumed the leaked volume in the middle two location Q₁ was 0.18 m³/d respectively, and the leaked volume of other location Q₂ was 0.0818 m³/d, while in case 3-2 the value of Q₁ was 0.27 m³/d and the value of Q₂ was 0.0758 m³/d respectively.

There is differential settlement at the ground surface for both case 3-1 and case 3-2, as shown in Fig. 8. The settlement at section I of case 1 and case 3 are shown in Fig. 9. With the extent of uneven leak rises, the differential settlement increases.

Conclusions

The relationship among long-term settlement, leaked volume of groundwater, and distribution pattern of leakage are investigated, the following conclusions can be drawn:

1). If the leakage in Shanghai tunnel distributes uniformly with an invariable velocity of 0.1 L/(m²·d^-1), the settlement of ground surface may reach to 90 mm after 10 years.

2). There is a linear relationship between settlement and leaked volume when the leaked volume is within the allowable limit. The settlement increases proportionally to the increase of leaked volume.

3). Large differential settlement may lead to the distortion of the rail track. Uneven leak will cause differential settlement which should be avoided.

4). The leaked volume will increase with the elapse of operating time. Measures to control leakage should be taken in time in case the groundwater level descends continuously and a larger long-time settlement occurs eventually.

5). Investigation on the in-situ leakage of Shanghai subway tunnel is recommended so that the real loss of water and soil is available. Both monitoring and controlling on leakage of tunnel segments should be strengthened.

References

[1] XUY S, SHENS L, DuY J. Geological and hydrogeological environment in Shanghai with geohazards to construction and maintenance of infrastructures.Engineering Geology, 2009, 109(3-4): 241-254

[2] ChaiJ C, ShenS L, ZhuH H, Land subsidence due to groundwater drawdown in Shanghai.Geotechnique, 2004, 54(3): 143-148

[3] ChaiJ C, ShenS L, ZhuH H, . 1D analysis of land subsidence in Shanghai.Lowland Technology International, 2005, 7(1): 33-41

[4] XuY S, ShenS L, CaiZ Y, et al. The state of land subsidence and prediction activities due to groundwater withdrawal in China.Natural Hazards, 2008, 45(1): 123-135

[5] Shanghai Metro Running Management Co Ltd (SMRM). Report of Monitoring on Shanghai Metro Tunnels in 2007. 2007 (in Chinese)

[6] MairR J, TaylorR N. Bored tunnel in the urban environment. In: Proceedings of the 14th International Conference on Soil Mechanics and Foundation Engineering. Hamburg, 1997, 2353-2385

[7] MairR J. Tunnelling and geotechnics: new horizons-46th rankine lecture. Geotechnique, 2008, 58(9): 695-736

[8] ShirlawJ N. Subsidence owing to tunneling II, evaluation of a prediction techniques. Canadian Geotechnical Journal, 1994, 31(3): 463-466 (discussion)

[9] O’ReillyM P, MairR J, AldermanG H. Long-term settlements over tunnels: an eleven-year study at Grimsby. In: Proceedings of Conference Tunnelling. London: Institution of Mining and Metallurgy, 1991, 55-64

[10] ShinJ H, AddenbrookeT I, PottsD M. A numerical study of the effect of groundwater movement on long-term tunnel behavior.Geotechnique, 2002, 52(6): 391-403

[11] ChenJ W and ZhanL X. Deformation measuring of the metro tunnel and deformation data analysis of Shanghai metro line No. 1.Shanghai Geology, 2000, (2): 51-56 (in Chinese)

[12] WangR L, ZhouX H, YuY L. Problem and its treatment during Shanghai metro monitor. In: Proceedings of the 14th Conference of Metro Tunnel Professional Committee. Tunnel and Underground Works Branch of China Civil Engineering Society. 2001, 240-242 (in Chinese)

[13] WuH N, HuM D, XuY S, et al. Law of influence of segment leakage on long-term tunnel settlement.Chiese Journal of Underground Space and Engineering, 2009, 5(z2): 1608-1611 (in Chinese)

[14] XIAW P. Underground engineering construction in soft clay — civil engineering of Shanghai metro Xujiahui station. Master thesis. Xi an: Xi’an University of Architecture and Technology, 2003 (in Chinese)

[15] LIAOS M, LIUJ H, WangR L. Shield tunneling and environment protection in Shanghai soft ground.Tunnelling and Underground Space Technology, 2009, 24(4): 454-465

[16] Shanghai Urban Construction and Communications Commission. Code for investigation for geotechnical engineering (DGJ08-37-2002), 2002 (in Chinese)

[17] BearJ. Hydraulics of Groundwater. New York: McGraw-Hill, 1979

[18] SHENS L, XUY S, HONGZ S. Estimation of land subsidence based on groundwater flow model. Marine Georesources and Geotechnology, 2006, 24(2): 149-1

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg