Environ Earth Sci (2016) 75:198 DOI 10.1007/s12665-015-4860-2 ORIGINAL ARTICLE Areal subsidence under pumping well–curtain interaction in subway foundation pit dewatering: conceptual model and numerical simulations Jianxiu Wang1,2,3, • Yuanbin Wu4 • Xiaotian Liu1 • Tianliang Yang5 Hanmei Wang5 • Yanfei Zhu6 • Received: 18 April 2014 / Accepted: 28 July 2015 / Published online: 25 January 2016 Ó Springer-Verlag Berlin Heidelberg 2015 Abstract Subway foundation pit dewatering may contribute to the regional land subsidence in a built-up area when aquifers are too thick to be cut off by a curtain The land subsidence induced by subway foundation pit dewatering was divided into local subsidence and areal subsidence The former was managed by construction organizations, while the latter was managed by a land resource, urban management, and hazard prevention department The areal subsidence could not be recovered after dewatering because of its vast influence area and contribution to regional land subsidence The boundary between local subsidence and areal subsidence was defined as the location of three times the foundation pit depth to the foundation pit boundary The subsidence within the boundary belonged to local subsidence, whereas the subsidence outside the boundary belonged to areal subsidence With Shanghai as background, the conceptual and mathematical models that considered hydrogeological conditions, & Jianxiu Wang wang_jianxiu@163.com College of Civil Engineering, Tongji University, Shanghai 200092, China Key Laboratory of Geotechnical and Underground Engineering of Ministry of Education, Tongji University, Shanghai 200092, China CCCC Key Lab of Environment Protection and Safety in Foundation Engineering of Transportation, Guangzhou 510230, China Institute of Karst Geology, CAGS, Guilin 541004, China Shanghai Institute of Geological Survey, Shanghai 200072, China Shanghai Tunnel Engineering Corporation Ltd., Shanghai 200082, China curtain depth, and pumping well screens were established, and numerical simulations were performed When the pumping well screens were enveloped by a diaphragm wall, the best vertical distance of about 1.0–4.0 m was obtained between the screen bottoms and curtain bottom The decreasing rate of areal subsidence was the highest under the interaction of diaphragm wall and pumping well screens Further increasing the penetration depth of the diaphragm wall and shortening the length of the pumping well screen were both less efficient Keywords Subway foundation pit Á Dewatering Á Local subsidence Á Areal subsidence Á Numerical simulation Introduction Land subsidence is a challenging issue addressed in many countries, such as Australia, China, Egypt, France, Germany, India, Iran, Israel, Italy, Japan, Mexico, Poland, Saudi Arabia, Sweden, Netherlands, the United Kingdom, and the United States (Hu et al 2004) Excessive groundwater withdrawal causes land subsidence in various regions worldwide (Shi et al 2012) In 1995, more than 150 areas were reported to experience subsidence, and as much as 10 m of subsidence was reported in Mexico, Japan, and the United States (Barends et al 1995) Land subsidence may disrupt surface drainage, reduce aquifer system storage, create ground fissures, and cause damage to properties, farmlands, and infrastructures that may be costly to repair (Sarychikhina et al 2011) A previous study has shown that land subsidence in China is primarily caused by the increasing withdrawal of groundwater at various depths (Hu et al 2004) In Shanghai, land subsidence has caused economic losses of more than US$ 13 billion (100 billion 123 198 Page of 13 RMB) and induced economic losses of approximately US$ 3.3 billion (24.57 billion RMB) during the first decade of the twenty-first century (Feng et al 2008) Although the extraction of groundwater for water supply can be restricted to control land subsidence, the deep foundation pit dewatering during excavation cannot be avoided because of the increasing exploitation of urban underground space Subway station foundation pits are often deep, constructed in a built-up area, and arranged close to one another Foundation pit dewatering may produce drawdown in a large area It also contributes to regional land subsidence when a large quantity of groundwater is extracted to rapidly lower the groundwater level for excavation safety The subsidence close to a foundation pit within the distance of three times the foundation pit depth (H) is defined as local subsidence This subsidence is believed to be caused by both excavation and dewatering and is usually managed by construction companies The subsidence far from a foundation pit (at a distance larger than three times the H) is defined as areal subsidence and is believed to be caused by dewatering alone This subsidence occurs in vast areas, and its accumulation contributes to the regional land subsidence in Shanghai Areal subsidence is managed by a land resource, urban management, and hazard prevention department Evaluating areal subsidence and its contribution to regional land subsidence is important Numerous studies based on adequate monitoring data have been performed on local subsidence, whereas knowledge on areal subsidence induced by foundation pit dewatering remains insufficient Previous studies refer to land subsidence as caused by the groundwater exploitation for water supply (e.g., Thomas and Johnson 1985; Dassargues and Zhang 1992; Larson et al 2001; Chen et al 2003, 2010; Li et al 2006; Shen and Xu 2011; Abidin et al 2011; Calderhead et al 2011; Yi et al 2011; Hung et al 2012) Most studies on foundation pit dewatering focused on the local subsidence (Xu et al 2012; Wang et al 2009a, 2012a, b, 2013a, b), while few studies are available on areal subsidence induced by subway foundation pit dewatering When the shape and scale of a subway foundation pit are determined and geological conditions are known, the areal subsidence is often influenced by the foundation pit curtain and pumping wells To control areal subsidence, curtain and pumping wells can be adopted Therefore, understanding the interaction between curtain and pumping wells and its effect on areal subsidence is important Shanghai was selected as the background of this study because it is located in a multi-aquitard multi-aquifer (MAMA) area (SGEAEB 2002), and its subway is the most developed in China The intensive dewatering of subway station foundation pits is one of the most important sources of regional land subsidence Extensive 123 Environ Earth Sci (2016) 75:198 information on regional geological conditions is available in open databases (SHGS 2014) A conceptual model that considers hydrogeological conditions, curtain depth, and pumping well screens was established herein Numerical simulations based on the finite difference method were performed to simulate the areal subsidence under pumping well and curtain interaction conditions in a subway foundation pit Background Subway station foundation pits Shanghai rail transit lines are mainly categorized into (1) domain-level fast subway (connects the suburbs with the central city), (2) municipal fast subway (connects the central city), and (3) light rail (Table 1) Subway stations in Shanghai are classified into underground, ground, and elevated ground stations (Table 2) According to depth, the underground stations are divided into two-, three-, four-, and five-floor underground stations Two- and three-floor underground subway stations are generally independent stations Four- and five-floor underground subway stations are generally transfer stations and large commuter hubs, respectively The shape and dimension of the station foundation pits are similar (Table 3) Engineering geological conditions In Shanghai, anthropogenetic activity influences the underground 75 m-deep soil layers (SGEAEB 2002) These layers (Fig 1; Table 4) include the following: Fill layer (cultivated soil and embankment of layer À) Cultivated soil is mainly plain fill clay, which is widely distributed and contains abundant plant roots and insect holes This layer is 0.2–0.5 m thick The 2.0–4.0 m-thick dredger of small old ponds belongs to plain fill clay Towns and settlements deposit construction waste, garbage soil, and local industrial waste soil with uneven thickness in this layer Topsoil layer (yellow brown silty clay and clay of layer `1) This layer belongs to coastal–estuarine sediments It is widely distributed in the region, plastic to soft Table Classification of Shanghai rail transit lines Classification Line Domain-level fast subway 1, 2, 5, 9, 11 Municipal fast subway 3, 4, 7, 8, 10, 12, 13, 14 Light rail 6, 15, 16, 17, 18 Environ Earth Sci (2016) 75:198 Page of 13 198 Table Subway lines 1–11, Shanghai, China Name Length (km) Number of stations Year of completion Subway no 1, the first phase 16.365 13 (two ground and 11 underground) 1994 Subway no 1, north extension 12.430 (three ground and six underground) 2004 Subway no 27.080 17 1999 Subway no 2, west extension 6.155 2006 Subway no 3, the first phase 24.975 19 (three ground and 16 underground) 2000 Subway no 22.002 17 (10 transfer stations) 2005 Subway no 17.160 11 (one ground and 10 underground) 2003 Subway no 33.130 28 (nine ground and 19 underground) 2007 Subway no 34.200 28 (two elevated and 15 transfer stations) 2009 Subway no 8, the first phase 23.218 22 (13 transfer stations) 2007 Subway no 9, the first phase Subway no 10 31.340 36.038 12 29 (13 transfer stations) 2007 2010 Subway no 11 66.90 34 (one ground, one underground, and 10 elevated) 2010 Table Typical subway station foundation pits Subway station Underground floors Scale (m m) Depth (m) Curtain depth (m) Water level (m) þ ¼2, ½ Qilianshan South Road (L13) 193.0 22.0 17.4–19.5 30.8–34.2 -3.7 -2.7 Xizang South Road (L3) 171.6 23.1 23.0–25.0 38.0–40.0 -3.7 -2.7 Hanzhong Road (L13) 204.4 20.6 30.8–32.3 58.0–62.0 -3.7 -2.7 Fig Typical geological section for the central city of Shanghai (SHGS 2014) plastic, slightly damp to wet, has no bedding, highly compressed, and contains a large amount of brown ferromanganese dip spots or tuberculos The layer is generally 1.5–3.0 m thick, but its thickness is 0.2–1.0 m in areas with thick artificial filling First sand layer This layer belongs to coastal–estuarine sediments and is mostly distributed along the eastern coast It includes sub-layers `2 and `a, together with phase transition layer `2 First soft soil layer This layer belongs to coastal– neritic deposition and is widely distributed in the area It comprises mucky clay of layers ´ and ˆ, which are typical soft soils; the first sand layer is distributed in the eastern zone, 1.0–14.8 m deep and 2.6–19.4 m thick, and the top elevation in most of the other regions is 0–2.8 m, with a thickness of 12.4–21.7 m Second soft soil layer This layer belongs to coastal swamp and drowned valley sediments It is widely 123 198 Page of 13 Environ Earth Sci (2016) 75:198 Table Strata and aquifers in Shanghai, China (SHGS 2014) Geological layer Aquifer Top soil The first sand layer The first soft soil layer The second soft soil layer Phreatic aquifer Phreatic aquifer Aquitard Aquitard Shallow confined aquifer Depth (m) Thickness (m) Soil serial Sub soil serial Soil name 1.07–6.00 1.5–3.0 ` `1 Brown yellow silty clay 2.0–3.0 0.3–9.40 ` 0–2.80 -13.00 to -18.00 ´ 12.4–21.7 Several m to 30.00 ˆ ˜ Aquitard `2 Gray sandy silt `2a Gray clayey silt `3 Gray silt ´ Gray silty clay ´a Gray sandy silt, silt ˆ ˜1 Gray silty clay Gray clay, silty clay ˜2 Gray silty clay silty fine sand ˜2a Gray silt, sandy silt ˜3 Gray silty clay, clay ˜4 Gray-green silty clay, clay Hard soil layer Aquitard -18.5 to -27.8 4–8.0 Þ Þ Dark green–brownish yellow silty clay, clay The second sand layer First confined aquifer -24.0 to -36.0 Several m to 39.0 þ þ1 Yellowish gray silt, sandy silt þ1l Gray-green silt, sandy silt þ2 Gray silt, sandy silt The third soft layer The third sand layer Aquitard Second confined aquifer -22 to -69.0 -52 to -78.0 Several m to 35.00 ¼ ¼1 Gray-green silty clay silty fine sand 30.00–50.00 ½ ¼2 ½ Gray silty clay and silt interbedded Gray fine sand with gravel in fine sand distributed, with stable roof elevations, most of which ranges between 13.0 and 18.0 m; the thickness of this layer varies greatly Hard clay soil layer This layer belongs to river–lake sediments and consists of dark green-brownish yellow silty clay and clay of layer Þ It is categorized as plastic to hard plastic, slightly wet, without bedding, lowly compressed, contains Fe–Mn spots and stripes, and occasionally tuberculotic; from top to bottom, the soil color of this layer varies from dark green to brownish yellow, clay content decreases, silt content increases, and Fe–Mn quality spots decrease This layer is cut by a river in Late Pleistocene and has discontinuous distribution, mainly distributed in the middle and west zones Second sand layer This layer is the first confined aquifer in Shanghai and belongs to coastal–estuarine sediments In Late Pleistocene, this layer was cleaned up and depleted by the deep cut of a paleochannel, which resulted in its absence in several areas It is mainly silty fine sand and sandy silt distributed locally From top to bottom, the soil in this layer becomes crude and locally becomes fine sand Its color ranges from yellow to gray grass (brown), and based on lithological differences, the layer is divided into sublayers þ1 and þ2 and phase transition layer þ1 123 Third soft soil layer This layer belongs to coastal and shallow marine sediments The cut of a paleochannel in Late Middle Pleistocene resulted in the absence of this layer in the central and southern areas It is commonly known as ‘‘layer-cake’’ sticky sand, alternating from top to bottom Third sand layer (gray fine sand and medium sand of layer ½) This layer belongs to estuarine coastal–fluvial deposition and consists of middle dense to dense fine sand From top to bottom, particles in this layer become coarser and sorted local folder pebbly coarse sand, with quartz as its primary mineral composition, followed by dark mineral, and lowly compressed It is mainly distributed in the central and southern regions and is connected with the second sand layer without the third soft soil Hydrogeological conditions The quaternary layers in Shanghai region include a phreatic aquifer, a micro-confined aquifer, and five layers of confined aquifers that influence the underground engineering and underground space development Phreatic water is stored in shallow strata Phreatic water depth is normally 0.3–1.5 m and is influenced by rainfall, tide, surface water, and ground evaporation Environ Earth Sci (2016) 75:198 Soil serial Soil name KH KV `3 Silt clay ˆ1 Muddy clay ˜1-1 198 Hydraulic conductivity (cm/s) Clay 2.10 10-5 4.53 10-5 1.06 10 -7 1.65 10-7 9.44 10 -8 1.11 10-7 -7 5.61 10-7 ˜1-2 Silt clay 3.23 10 Þ Silt clay 2.19 10-7 5.46 10-7 1.52 10 -4 1.64 10-4 3.79 10 -4 4.27 10-4 1.82 10 -7 2.22 10-7 2.74 10 -7 1.45 10-6 þ1 Sandy slit þ2 Silty fine sand ¼1 Clay ¼2 Interbed of silt clay and silty fine sand Fig Conceptual model of subway foundation pit and boundary between local subsidence and areal subsidence Boundary Depth 30.83 m Diaphragm wall Depth 58 m Diaphragm wall Depth 58 m Foundation pit Depth 30.83 m 3H Diaphragm wall Depth 58 m Boundary Depth 32.63 m Boundary Diaphragm wall Depth 62 m Depth 32.63 m Table Hydraulic conductivity of typical soil layers from laboratory experiments in Hanzhong Road Station, Subway line 12/13, Shanghai, China (SUCDRI 2008) Page of 13 Diaphragm wall Depth 62 m local subsidence Boundary Areal subsidence The micro-confined aquifer is silt sand or silt and local folder clay, and has uneven soil properties and good permeability It is distributed discontinuously in layers ˆ and ˜, and in sandwich layers of layer ˜, locally connecting with the first confined aquifer The water level in this aquifer is similar to that in the first confined aquifer The first confined aquifer (layer þ) is silt or fine sand, good in permeability, and rich in water It is continuously distributed spatially, locally connecting with the microconfined or second confined aquifer The groundwater depth in this aquifer is 3.0–11.0 m and varies in different regions The second confined aquifer (layer ½) is fine sand and coarse sand, good in permeability, and extremely rich in water It is distributed continuously, locally connecting with the first confined aquifer and forming thick aquifer groups The water level depth in this aquifer is generally 3.0–12.0 m and varies in different regions The hydraulic conductivities of the layers are listed in Table Conceptual models Land subsidence is a significant environmental problem in subway stations, especially in five-floor subway stations where the aquifers are too thick to be penetrated by curtains Dewatering inside the pits may result in drawdown and subsidence outside In Shanghai, the maximum subsidence caused by excavation was normally observed in the distance of 0.5 H to the diaphragm walls outside the pits The areas between and H to pit are significantly influenced by both excavation and dewatering The areas between and H to pit are significantly influenced by dewatering but insignificantly influenced by excavation The areas that exceed H to pit are usually influenced by dewatering alone Thus, the distance of H from the foundation pit is defined as the separation boundary between local and areal subsidence (Fig 2) Areal subsidence cones around the pits may appear during the dewatering process The development and superposition of the cones may lead to a larger cone contributing to the regional land subsidence Recovery of the local subsidence may be 123 198 Page of 13 possible through construction, but is difficult for the areal subsidence A subway station foundation pit normally has a standard shape and scale A diaphragm wall is often used as both retaining structure and curtain, and its depth varies under different geological conditions The conceptual model of a subway station foundation pit is shown in Fig In the conceptual model, the engineering geological and hydrogeological conditions are described through geological partitions Each partition is characterized by the formation and strata structure (SHGS 2014) The strata are divided into ancient river cutting and normal deposition areas according to the existence of layer Þ The central city is divided into two main geological partitions The area with layer Þ is categorized as Partition I (normal deposition zone), and that without layer Þ is categorized as Partition II (ancient river cutting area) The main partitions are further divided into sub partitions (Fig 3) The partitions include the combinations of aquitards and aquifers, and the different partitions have various MAMA Fig Geological partition for the central city of Shanghai (SHGS 2014) 123 Environ Earth Sci (2016) 75:198 combinations The areal subsidence is not only determined by the dewatered aquifer, but also by the nearby layers The mechanism of this subsidence is different in various partitions When the bottom of a foundation pit is in an impermeable layer underlying a confined aquifer (Fig 4), its stability during excavation is checked using the following equation (MOHURD 1999; GAQSIQ MOHURD 2000): X cs pwk c i hi ð1Þ cRY where cs is the partial factor of confined water (adopted as 1.1), pwk is the standard value of the water pressure of the confined aquifer (kPa), ci is the unit weight of the layer i (kN/m3), hi is the thickness of layer i (m), and cRY is the partial factor of resisting confined aquifer water pressure (adopted as 1.1, MOHURD 1999) The unit weights of soils are listed in Table For a standard foundation pit, the safe drawdown with respect to the partial factors is listed in Table Environ Earth Sci (2016) 75:198 Ground 198 Ground Groundwater level D Fig Dewatering calculation model Page of 13 Foundation pit γ2 Layer h1 γ Layer h0 γ Aquiclude Water pressure pwk = γw H γ H h2 h1 γ h0 γw is the unit weight of groundwater pwk Confined aquifer Table Average unit weight of soil layers h2 γ Soil serial ˜1 ˜2 ˜3 ˜4 Þ1Þ2 þ1 þ2 ¼1 ¼2 ½1 ½2 Unit weight (kN/m3) 18.3 18.6 18.5 19.9 19.8 19.3 19.4 18.5 19.1 19.7 20.2 Table Safe water level of confined aquifer Ground elevation (m) 4.4 Scale (m m) 204.4 20.6 Depth (m) Pit bottom elevation (m) Curtain depth (m) Safe water level (m) Standard part End well Standard part End well Standard part End well Standard part End well 30.8 32.3 -26.4 -27.9 58.0 62.0 -13.3 -15.8 The elevation of the top of the first confined aquifer is -15.8 m, whereas the second is -45.6 m The top of the second confined aquifer is the top of layer ¼2 Numerical simulations Mathematical model The study was conducted to predict the areal subsidence of foundation pit dewatering under different geological conditions in a large scale In theory, the prediction of the areal subsidence is more precise when a three-dimensional (3D) hydro-mechanically coupled model is used As the total isotropic stress is considered to be locally constant with time, the decoupled model can be adopted Therefore, a two-step method was suggested In the first step, the drawdown during foundation pit dewatering was calculated using a 3D numerical method In the second step, areal subsidence was calculated using consolidation theory In the calculation, the arrangement and structure of pumping wells were difficult to determine because different arrangements of pumping wells may be adopted in various foundation pits Hence, the foundation pit dewatering was simplified as a constant head boundary in the mathematical model, i.e., the pumping of wells was simulated by the lowered water head in the foundation pit bottom with a specific thickness (Fig 5) According to the conceptual model (Fig 6), the groundwater mathematical model of foundation pit dewatering is as follows:       o o/ o o/ o o/ o/ > > kxx kyy kzz þ þ À W ¼ Ss > > ox ox oy oy oz oz ot > > > > > > < / ¼ /1 ðx; y; z; tÞ / ¼ /2 ðx; y; z; tÞ > > > o/ o/ oF o/ oF o/ oF > > > ¼ Kxx þ Kyy þ Kzz ¼0 > > on ox ox oy oy oz oz > > : /ðx; y; z; tÞjt¼t0 ¼ /0 ðx; y; zÞ ðx; y; zÞ ðx; y; zÞ C1 ðx; y; zÞ C2 ðx; y; zÞ C3 ðx; y; zÞ X ð2Þ where Kxx ,Kyy , and Kzz are the hydraulic conductivities along the x, y, and z directions (m/d), respectively; / is the water head (m); /1 is a constant water level on C1 (m); /2 is a constant water level on C2 (m); W is the source and sink (1/day); Ss is the specific storage (1/m); t is time (day); X is the computational domain; C1 is the first boundary (constant head boundary) at the bottom of the pit; C2 is the first boundary (constant head boundary) on the side of the 123 198 Page of 13 Fig Simulation of pumping well with an equal constant water level boundary Environ Earth Sci (2016) 75:198 FoundaƟon pit Pumping well Q Q Q Q IniƟal water level Water level Equal water level Aquitard Confined aquifer Pumping well screen (a) Pumping model FoundaƟon pit IniƟal water level Constant water level boun Aquitard Confined aquifer Equal pumping well screen (b) Equal constant water level boundary model computational domain far from the pit; C3 is the second boundary (impermeable boundary) at the bottom of the computational domain; n is the normal direction of the bottom of the computational domain; and F is the surface equation of the bottom of the computational domain The subsidence caused by dewatering was calculated using the following equation (MOHURD 1999): n X Db ¼ b0i mvi si cw F ð3Þ i¼1 where Db is the subsidence (mm); b0i is the initial thickness of layer i(m); mvi is the coefficient of volumetric compressibility of layer i (MPa-1), mvi ¼ 1=Ei ; si is the drawdown of layer i (m); cw is the unit weight of groundwater (kN/m3); F is the subsidence empirical coefficient; and n is the number of layers influenced by dewatering The recommended values of the compression modulus are shown in Table 123 Numerical model Numerical models based on the geological partitions and characteristics of a standard subway station foundation pit were established using MODFLOW software (USGS 2003) Pumping wells were simulated with a constant head boundary (safe water level) condition Three observation points, located in the middle of layers ¼2, ½1, and ½2, were arranged in each observation well The parameters of the dewatered aquifer were inversed by pumping tests performed in selected deep foundation pits (Table 9; Fig 7) The working conditions for the calculation are listed in Table 10 The pumping duration was fixed to 90 days Results and discussion Numerical simulations were performed for the working conditions in the partitions Environ Earth Sci (2016) 75:198 Page of 13 198 Foundation pit Ground Ground Groundwater level Diaphragm wall Secondary boundary First boundary Confined aquifer First boundary Secondary boundary First boundary Secondary boundary 3 Computational domain Fig Conceptual model of foundation pit dewatering Table Recommended compression modulus Soil serial Soil name Laboratory test E (MPa) Standard penetration test E (MPa) CPT E (MPa) Recommended E (MPa) þ1 Sandy slit 25.0 34.0 36.6 30.0 þ2 Silty fine sand 35.0 66.9 67.8 60.0 ¼2 Interbed of silt clay and silty fine sand 14.5 ½1 Silty fine sand the older silty clay 59.0 31.5 35.0 ½2 Fine sand 83.1 73.4 65.0 Table Inversed hydraulic conductivities Confined aquifer Kxx ,Kyy (m/s) Kzz (m/s) ¼2 1.16 10-6 1.27 10-6 4.21 10-5 ½1 -6 -6 6.32 10-5 9.05 10 2.12 10 Ss (1/m) As an example, the drawdown for Partition I1-1 is shown in Fig For the areal subsidence that resulted in the seepage field driving the stress field, the magnitude of subsidence influenced the drawdown The subsidence corresponding to the penetration depth of the diaphragm wall is shown in Fig The calculated subsidence is corrected according to local experience because it is often larger than the measured subsidence (Zhang et al 2001; Xu and Li 2004; Wang et al 2006, 2009b) The corrected subsidence corresponding to an empirical coefficient for Partition I1-1 is shown in Fig 10 The empirical coefficient in Shanghai is often adopted as 0.25–0.30 The areal subsidence in Partition I1-1 decreases with the increase in penetration depth of the diaphragm wall However, the decreasing rate is very limited when the pumping well screens in the pit are deeper than the diaphragm wall The screens stretch out 13.6 from the bottom of diaphragm wall and are not entirely enveloped in the horizontal direction Increasing the wall penetration depth cannot efficiently decrease the drawdown or areal subsidence If the lengths of the pumping wells are constant and the wall penetration depth is increased, then the areal subsidence will decrease When the screens are approximately 1.0 m above or below the elevation of the diaphragm wall bottom, its decreasing rate (Figs 8, 9) is the highest However, the decreasing rate lessens with further increases in penetration depth The subsidence corresponding to the penetration length in the other partitions without correction is shown in Fig 11 A maximum interaction location is found between diaphragm wall and pumping well screens The maximum vertical distance between diaphragm wall bottom and enveloped screen bottoms of pumping wells is approximately 1.0–4.0 m When the screen bottoms are approximately 1.0–4.0 m above that of the diaphragm wall, the decreasing rate of areal subsidence is the highest When the vertical length exceeds the value, the decreasing rate of areal subsidence decreases Therefore, further increasing the penetration depth and shortening the length of the pumping well screen are both less efficient The best vertical distance 123 198 Page 10 of 13 Environ Earth Sci (2016) 75:198 Fig Error in parameter inverse 0.0 Table 10 Designed working conditions Curtain depth (m) Elevation of curtain bottom (m) 51 -46.6 52 -47.6 53 -48.6 54 -49.6 58 -53.6 63 -58.6 68 -63.6 69 -64.6 70 -65.6 10 71 -66.6 11 76 -71.6 -20.0 20m 30m Subsidence (mm) Calculation case Screen of pumping well -40.0 50m 70m -60.0 100m 150m -80.0 200m 300m -100.0 Diaphragm wall -120.0 10 500m 15 20 25 30 Curtain penetrating depth into confined aquifer (m) Fig Subsidence corresponding to curtain penetrating depth for different distance to foundation pit (Partition I1-1) 0.0 F=1.0 -20.0 Best distance to diaphragm wall bottom Screen of pumping well F=0.9 Drawndown (m) 5.0 4.0 3.0 ⑧2 ⑨1 2.0 Subsidence (mm) 6.0 F=0.8 -40.0 F=0.7 F=0.6 -60.0 F=0.5 F=0.4 -80.0 F=0.3 ⑨2 -100.0 1.0 0.0 10 F=0.1 15 20 25 30 Curtain penetrating depth into confined aquifer (m) Fig Drawdown corresponding to curtain penetrating depth in different layers (Partition I1-1) 123 F=0.2 Diaphragm wall bottom Diaphragm wall -120.0 0.0 100.0 200.0 300.0 400.0 500.0 600.0 DIstance to foundaƟon pit boundary (m) Fig 10 Corrected subsidence based on different distances to the foundation pit boundary (Partition I1-1) Environ Earth Sci (2016) 75:198 (a) 0.0 Page 11 of 13 (b) Screen of pumping well Screen of pumping well -10.0 -2.0 30m 50m -6.0 70m -8.0 100m 150m -10.0 Subsidence (mm) 20m -4.0 Subsidence (mm) 0.0 -12.0 20m -30.0 30m 50m 70m -50.0 300m Diaphragm wall -20.0 -40.0 200m 100m -60.0 150m -70.0 200m 300m -80.0 Diaphragm wall 500m 500m -90.0 -14.0 10 15 20 25 0.0 10 15 20 25 Curtain penetraƟng depth into confined aquifer (m) Curtain penetraƟng depth into confined aquifer (m) (c) (d) Screen of pumping well Screen of pumping well 2.00 0.00 -10.0 20m 20m -2.00 30m -20.0 50m -30.0 70m 100m -40.0 150m Subsidence (mm) Subsidence (mm) 198 200m -50.0 Diaphragm wall 300m 30m -4.00 50m -6.00 70m -8.00 100m 150m -10.00 200m Diaphragm wall -12.00 300m 500m -60.0 500m -14.00 10 15 20 10 15 20 25 Curtain penetraƟng depth into confined aquifer (m) Curtain penetraƟng depth into confined aquifer (m) (e) 0.00 Screen of pumping well -5.00 20m Subsidence (mm) -10.00 30m -15.00 50m 70m -20.00 100m -25.00 150m -30.00 200m 300m -35.00 Diaphragm wall 500m -40.00 10 15 20 25 30 Curtain penetraƟng depth into confined aquifer (m) Fig 11 Subsidence corresponding to curtain penetrating depth for different distance to foundation pit between screen bottom and diaphragm wall bottoms is a crucial parameter for areal subsidence control Diaphragm wall can be adopted as one of the measures to control subsidence However, if the pumping well screens stretch out from the diaphragm wall, even if the walls are very deep, then the subsidence cannot be controlled efficiently When the screens are installed in the elevation 1.0–4.0 m above the diaphragm wall bottom, the decreasing efficiency of the areal subsidence is the highest Conclusions The following conclusions may be drawn: Contrary to the extraction of groundwater for water supply, subway foundation pit dewatering cannot be avoided and is one of the most important causes of regional land subsidence in a built-up area when aquifers are too thick to be cut off by curtains 123 198 Page 12 of 13 The land subsidence induced by subway foundation pit dewatering can be distinguished into local subsidence and areal subsidence The former is managed by companies, while the latter is managed by a land resource, urban management, and hazard prevention department The boundary between local subsidence and areal subsidence is fixed to a distance of H from the foundation pit boundary Shanghai was selected as a study background A conceptual model of subway foundation pit dewatering that considers hydrogeological conditions, diaphragm walls, and pumping wells was established Numerical simulations of areal subsidence were performed based on the finite difference method When the pumping well screens are installed in a pit and horizontally enveloped by a diaphragm wall, the decreasing rate of areal subsidence is the highest A maximum interaction location is found between diaphragm wall and screen bottoms of pumping wells for subway foundation pit dewatering The best vertical distance between diaphragm wall bottom and enveloped screens is 1.0–4.0 m When the screen bottom is 1.0–4.0 m higher than that of the diaphragm wall, the decreasing rate of areal subsidence is the highest The decreasing rate of areal subsidence decreases when the vertical distance exceeds the value Increasing the penetration depth and shortening the length of the pumping well screen are both less efficient Acknowledgments I must show my respect to the editors and anonymous reviewers for their detail and patient work to my manuscript This work is supported by the research grant (No 201311045-04) from the Special Fund for Land and Resources-scientific Research in the Public Interest of China, CCCC Key Lab of Environment Protection & Safety in Foundation Engineering of Transportation, GDUE Open Funding (SKLGDUEK1417), LSMP Open Funding (KLLSMP201403, KLLSMP201404), the National Natural Science Foundation of China (No 41072205), the Key Discipline Construction Program of Shanghai (Geological Engineering, No B308) and the Foundation of China Railway No Engineering Group Co., Ltd (No 201218) References Abidin HZ, Andreas H, Gumilar I, Fukuda Y, Pohan YE, Deguchi T (2011) Land subsidence of Jakarta (Indonesia) and its relation with urban development Nat Hazards 59:1753–1771 Barends FBJ, Brouwer FJJ, Schroder FH (1995) Land subsidence: natural causes, measuring techniques, the Groningen gasfields IAHS Press, Wallingford Calderhead AI, Therrien R, Rivera A, Martela R, Garfiasd J (2011) Simulating pumping-induced regional land subsidence with the use of InSAR and field data in the Toluca Valley, Mexico Adv Water Resour 34:83–97 123 Environ Earth Sci (2016) 75:198 Chen CX, Pei SP, Jiao JJ (2003) Land subsidence caused by groundwater exploitation in Suzhou City, China Hydrogeol J 11:275–287 Chen CH, Wang CH, Hsu YJ, Yu SB, Kuo LC (2010) Correlation between groundwater level and altitude variations in land subsidence area of the Choshuichi Alluvial Fan, Taiwan Eng Geol 115:122–131 Dassargues A, Zhang J (1992) Land subsidence in Shanghai: hydrogeological conditions and subsidence measurements Bull Eng Geol Environ 46:27–36 Feng QY, Liu GJ, Meng L, Fu EJ, Zhang HR, Zhang KF (2008) Land subsidence induced by groundwater extraction and building damage level assessment—a case study of Datun, China China Univ Mining Technol 18:556–560 General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China (GAQSIQ), Ministry of Housing and Urban-Rural Development of the People’s Republic of China (MOHURD) (2000) Code on geotechnical investigations for metro and light rail transit (GB 50307-1999) China Planning Press, Beijing Hu R, Yue ZQ, Wang LC, Wang SJ (2004) Review on current status and challenging issues of land subsidence in China Eng Geol 76:65–77 Hung WC, Hwang C, Liou JC, Lin YS, Yang HL (2012) Modeling aquifer-system compaction and predicting land subsidence in central Taiwan Eng Geol 147–148:78–90 Larson KJ, Bas¸ agˇaogˇlub H, Marino MA (2001) Prediction of optimal safe ground water yield and land subsidence in the Los BanosKettleman City area, California, using a calibrated numerical simulation model J Hydrol 242:79–102 Li CJ, Tang XM, Ma TH (2006) Land subsidence caused by groundwater exploitation in the Hangzhou-Jiaxing-Huzhou Plain, China Hydrogeol J 14:1652–1665 Ministry of Housing and Urban-Rural Development of the People’s Republic of China (MOHURD) (1999) Technical code for groundwater lower in engineering in building and municipal (TCGLEBM) (JGJ/T111-98) (1999) China Architecture & Building Press, Beijing Sarychikhina O, Glowacka E, Mellors R, Vidala FS (2011) Land subsidence in the Cerro Prieto Geothermal Field, Baja California, Mexico, from 1994 to 2005 J Volcanol Geotherm Res 204:76–90 Shanghai Geological Environmental Atlas Editorial Board (SGEAEB) (2002) Shanghai geological environmental atlas Geological Publishing House, Beijing (in Chinese) Shanghai Institute of geological survey (SHGS) (2014) http://www sigs.com.cn/sigsonlines/ Accessed 22 Sept 2012 Shanghai Urban Construction Design & Research Institute (SUCDRI) (2008) Geotechnical Engineering investigation report of Hanzhong Road Station, Shanghai Subway line 12 (detailed investigation) Shen SL, Xu YS (2011) Numerical evaluation of land subsidence induced by groundwater pumping in Shanghai Can Geotech J 48:1378–1392 Shi XQ, Xu HH, Sun YY, Yu J (2012) Sustainable development and utilization of groundwater resources considering land subsidence in Suzhou, China Eng Geol 124:77–89 Thomas LH, Johnson AI (1985) Land Subsidence caused by ground water withdrawal in urban areas GeoJournal 11(3):245–255 US Geological Survey (USGS) (2003) MODFLOW-2000, the US Geological Survey Modular Ground-Water Model—Documentation of the SEAWAT-2000 Version with the Variable-Density Flow Process (VDF) and the Integrated MT3DMS Transport Process (IMT) Tallahassee, Open-File Report 03-426, Florida 2003 Wang CY, Huang LX, Duan WC, Yu YF (2006) Determination of coefficient Ms during calculation of surrounding ground Environ Earth Sci (2016) 75:198 settlement due to foundation pit dewatering Rock Soil Mech 27:1011–1016 (in Chinese) Wang JX, Hu LS, Wu LG, Tang YQ, Zhu YF, Yang P (2009a) Hydraulic barrier function of the underground continuous concrete wall in the pit of subway station and its optimization Environ Geol 57:447–453 Wang JX, Wu LG, Zhu YF, Tang YQ, Yang P, Lou RX (2009b) Mechanism of dewatering-induced ground subsidence in deep subway station pit and calculation method Chin J Rock Mech Eng 28:1010–1019 (in Chinese) Wang JX, Feng B, Liu Y, Wu LG, Zhu YF, Zhang XS, Tang YQ, Yang P (2012a) Controlling subsidence caused by de-watering in a deep foundation pit B Eng Geol Environ 71:545–555 Wang JX, Wu YB, Zhang XS, Liu Y, Yang TL, Feng B (2012b) Field experiments and numerical simulations of confined aquifer response to multi-cycle recharge-recovery process through a well J Hydrol 464–465:328–343 Page 13 of 13 198 Wang JX, Feng B, Guo TP, Wu LG, Lou RX, Zhou Z (2013a) Using partial penetrating wells and curtains to lower the water level of confined aquifer of gravels Eng Geol 161:16–25 Wang JX, Feng B, Yu HP, Guo TP, Yang GY, Tang JW (2013b) Numerical study of dewatering in a large deep foundation pit Environ Earth Sci 69:863–872 Xu XJ, Li DX (2004) Study on calculating method of ground subsidence caused by pit-dewatering Geotech Eng Tech 18:194–198 (in Chinese) Xu YS, Ma L, Shen SL, Sun WJ (2012) Evaluation of land subsidence by considering underground structures penetrated into aquifers in Shanghai Hydogeol J doi:10.1007/s10040-012-0892-9 Yi LX, Zhang F, Xu H, Chen SJ, Wang W, Yu Q (2011) Land subsidence in Tianjin, China Environ Earth Sci 62:1151–1161 Zhang LH, Li RQ, Liu DF (2001) The calculation of effective stress and settlement based on the variable total stress in pit dewatering Chin J Geol Hazard Control 12(66–67):77 (in Chinese) 123 ... 10-5 4. 53 10-5 1.06 10 -7 1.65 10-7 9 .44 10 -8 1.11 10-7 -7 5.61 10-7 ˜1-2 Silt clay 3.23 10 Þ Silt clay 2.19 10-7 5 .46 10-7 1.52 10 -4 1. 64 10 -4 3.79 10 -4 4.27 10 -4 1.82 10 -7 2.22 10-7 2. 74 10... bottom (m) 51 -46 .6 52 -47 .6 53 -48 .6 54 -49 .6 58 -53.6 63 -58.6 68 -63.6 69 - 64. 6 70 -65.6 10 71 -66.6 11 76 -71.6 -20.0 20m 30m Subsidence (mm) Calculation case Screen of pumping well -40 .0 50m 70m... the areal subsidence is often influenced by the foundation pit curtain and pumping wells To control areal subsidence, curtain and pumping wells can be adopted Therefore, understanding the interaction