Available online at www.sciencedirect.

com

ScienceDirect
Underground Space 6 (2021) 35–45
www.elsevier.com/locate/undsp

Groundwater-related aspects during the development of

deep excavations below the water table: A short review
Estanislao Pujades a,⇑, Anna Jurado b
a
Department of Computational Hydrosystems, UFZ – Helmholtz Centre for Environmental Research, Permoserstr. 15, 04318 Leipzig, Germany
b
Institute for Groundwater Management, Technische Universität Dresden, Dresden, Germany

Received 28 April 2019; accepted 17 June 2019

Available online 15 October 2019

Abstract

The expected growth of urban areas requires the construction of underground infrastructures, especially for improving transportation
and mobility. These new infrastructures are commonly constructed below the water table because they must be deeper than previous
underground structures. In addition, many urban areas are located near surface water bodies (i.e., seas, rivers and/or lakes) that induce
a shallow water table. The ‘‘cut and cover” method combined with dewatering wells is a frequently used procedure under these conditions
for the construction of underground infrastructures. The main concerns of this procedure are related to the stability of the excavation
bottom, the behavior of the retaining walls and/or the impacts induced by the dewatering actions. These issues must be evaluated, and
new solutions must be proposed to improve its weak points. This paper compiles the main publications focused on the construction of
underground infrastructures below the water table and in urban areas by using the cut and cover method combined with pumping. This
review reveals that, despite many published studies, more research is needed to address some relevant issues such as: (1) the mitigation of
dewatering impacts by using artificial recharge, (2) the adaptation of methods for characterizing retaining walls to be used under hetero-
geneous conditions, (3) the improvement of the prediction of surface settlements produced by pumping, and (4) the consideration of non-
Darcy flows in the design of dewatering systems.
Ó 2019 Tongji University. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Dewatering; Excavation; Cut and cover; Retaining walls; Settlements

1 Introduction tools can be adopted depending on the characteristics of

the infrastructure and the area in which the construction
Recent data from the United Nations show that half of is developed. For example, tunnels can be drilled with tun-
the world’s population lived in urban areas in 2014, and nel boring machines (TBMs), excavated by the ‘‘cut and
this proportion is projected to reach 66% by 2050 (UN, cover” method or constructed using traditional (Belgian,
2014). Cities around the world face the challenge of con- Madrid, German) or sequential (new Austrian) tunneling
structing new infrastructures to facilitate mobility, but they methods (Pulido, Darido, Munoz-Raskin, & Moody,
cannot be constructed on the surface due to space limita- 2018). Among the available techniques, the cut and cover
tions. This situation leads to the vertical development of method is one of the most commonly used to carry out
urban areas where new infrastructures must be deeper than excavations in urban areas (Konda, 2001). In fact, the
the previous ones. Different construction methods and cut and cover method is frequently adopted for the con-
struction of stations in railway projects (Di, Zhou, Xiao,
Gong, & Luo, 2016; Zhou, Wang, & Hu, 2013); the assem-
⇑ Corresponding author. Fax: +49 3412351939. bly, disassembly and maintenance shafts required when
E-mail address: estanislao.pujades-garnes@ufz.de (E. Pujades).

https://doi.org/10.1016/j.undsp.2019.10.002
2467-9674/Ó 2019 Tongji University. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
36 E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45

drilling tunnels by TBMs (Culı́, Pujades, Vázquez-Suñé, & tering system is properly designed, instability events may
Jurado, 2016) or other kind of underground infrastructures occur if the retaining walls have unidentified defects, which
such as parking lots (Pappa & Benardos, 2007). This promote the water pressure to be higher than expected.
method consists of excavating under the protection of ver- Defects in the retaining walls are relatively common and
tical retaining walls (Gulhati & Datta, 2005) that can be can produce a wide range of negative consequences (flood-
constructed of steel, concrete, or jet-grouting piles or dia- ing, sinkholes, bottom instability and large soil deforma-
phragm walls, among other materials. The main advantage tions) that depend on their relative location with respect
of this method is that the excavation walls are vertical, to the excavation bottom. In addition, independent of the
reducing the space occupied by the construction, which is location of the defects, they contribute to propagate the
meaningful in urban environments. The main steps of the effects of pumping outside the excavation enclosure. Thus,
cut and cover method are (1) to construct the retaining groundwater drawdown and its associated impacts, includ-
walls, (2) to excavate to the desired depth, (3) to construct ing surface settlements, increase when defects exist in the
the infrastructure, and (4) to fill the space between the top retaining walls. Dewatering impacts are of paramount
of the constructed structure and the surface (Gulhati & importance, especially in urban areas, and they must be
Datta, 2005; Mouratidis, 2008) (steps 1, 3, 4 and 5 in predicted in advance. The main impacts produced by dewa-
Fig. 1). tering are the drawdown of the water table and the surface
The construction of new urban underground infrastruc- settlements, which are very concerning since they can be
tures is usually performed below the water table because large and damage urban structures (e.g., buildings)
they must be deeper than existing underground structures. (Forth, 2004). Consequently, dewatering systems are
In addition, many cities are commonly located near rivers designed to minimize them. The easiest way to mitigate sur-
or in coastal areas where the water table is usually near face settlements associated with the dewatering of deep
the surface. The presence of groundwater and its interac- excavations consists of elongating the retaining walls more
tion with the construction process must be properly consid- than is structurally required until reaching deep impervious
ered to avoid accidents and difficulties (Pujades et al., formations, which allows reducing or eliminating the
2012), especially during deep excavations. When excavat- pumping. However, this decision is controversial since
ing below the water table, the retaining walls are usually the excessive elongation of the retaining walls may reduce
combined with dewatering wells (step 2 in Fig. 1). Retain- the global efficiency of the construction. In addition, when
ing walls prevent lateral groundwater inflow, whereas pumping is eliminated by elongating the retaining walls,
dewatering wells allow excavating in dry and stable condi- the safety exclusively depends on their watertightness
tions. Pumping reduces the water table inside the enclosure capacity, and as has been reported by numerous authors,
and the water pressure in the bottom of the excavation, they commonly have defects (Bruce, DePaoli, Mascardi,
which prevents liquefaction and bottom uplift (Calin, & Mongilardi, 1989). Concurrently, some authors affirm
Cristian, & Ioan, 2017). Liquefaction or bottom uplift that pumping settlements are smaller and less dangerous
occurs when the water pressure exceeds the soil weight. (i.e., no differential) than expected, especially when the soil
These phenomena have negative implications for the devel- is overconsolidated (Pujades, De Simone, Carrera,
opment of the excavation and the safety of nearby build- Vázquez-Suñé, & Jurado, 2017). Regardless, corrective
ings and workers. For this reason, the dewatering wells measures to mitigate dewatering impacts already exist.
and the retaining walls must be conscientiously designed These measures basically consist of returning the pumped
to reach the required water pressure at the excavation bot- groundwater to the aquifer by artificial recharge. Artificial
tom. Numerical models are a powerful tool to design dewa- recharge is commonly performed through deep wells
tering systems and to ensure that water pressure does not located outside the excavation (Zeng, Zheng, Xue, &
exceed the desired threshold. However, although a dewa- Mei, 2019). The efficiency of artificial recharge systems to

Fig. 1. Cut and cover method steps for the construction of an infrastructure below the water table (modified from Pujades et al., 2012).
 E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45 37

eliminate surface settlements depends on the spatial distri- duce, among others, structural instability and subsidence
bution of the recharge wells and the methods used to solve related to soil migration (Xu et al., 2014). Defects also
clogging-related issues. Finally, another concern during increase the hydraulic connection between the inside and
deep excavations is the deflection of the retaining walls, outside of the excavation enclosure. Consequently, the
which also produces soil deformations. Most published excavation may be flooded, and the effects of the dewater-
studies barely consider the influence of the groundwater ing (i.e., drawdown and settlements) are more easily prop-
fluctuations on the displacements of the retaining walls. agated outside. Defects are relatively easy to repair by
Only some authors have probed whether dewatering tasks injection grouting, but the repair must be undertaken
may produce important retaining wall deflections. before the excavation stage. If the repair is carried out dur-
Studies focused on investigating groundwater-related ing the excavation stage, all injected grout flows toward the
aspects during deep excavations have been published dewatering wells. As a result, defects remain open, and the
beginning many years ago. Science related to this topic dewatering wells are damaged. If defects are abundant and
has been developed and has evolved over time through they cannot be repaired, the alternative option is to rede-
numerous publications. The objective of this publication sign the dewatering system, considering the existence of
is to show the most recent innovative progress in the field. the defects and assuming a high value of hydraulic conduc-
For this reason, the main findings over the last five years tivity for the retaining walls. In both situations, isolated or
are compiled in this paper. Investigations regarding the sta- numerous defects, it is always mandatory to characterize
bility of the excavation bottom, the behavior of the retain- the retaining walls before the excavation stage.
ing walls, the impacts induced by the dewatering and Accidents such as those reported by Feng and Lu (2015)
corrective measures to avoid them are considered and and Tan and Lu (2017) may occur if defects are not identi-
discussed. fied before the excavation stage. Both studies reported acci-
dents during excavations associated with the presence of
2 Methodology and considered literature defects in the retaining walls. Feng and Lu (2015) presented
the case of a failure during the construction of a subway
This review includes published studies during the last station in Nanchang, China. The main consequences of
five years focused on excavations developed in urban areas. the faulty retaining walls were leaks, bottom liquefaction
In total, Tens of studies are considered. Most of them are and sinkholes outside the enclosure. Tan and Lu (2017)
focused on underground constructions developed in China reported groundwater inflow, dragged sediments, increased
but also in Spain, Italy and Singapore. Table 1 summarizes settlements, sinkholes outside the excavation and cracks
the main characteristics of the constructions. This manu- affecting nearby buildings.
script is structured considering relevant issues related to Given the potential risks associated with faulty retaining
deep excavations in urban areas (state and behavior of walls, some authors have proposed methods to assess their
the retaining walls, design of dewatering systems, impact state before the excavation stage by means of the interpre-
mitigation). Studies containing information about more tation of pumping tests. The pumping tests, which are
than one issue are cited and discussed in different sections. undertaken after the construction of the retaining wall,
are called ‘‘watertightness assessment tests” (WTATs).
3 State and behavior of the retaining walls Pujades, Vázquez-Suñé, Carrera, Vilarrasa, et al. (2014)
and Pujades, Vázquez-Suñé, Carrera, and Jurado (2014)
3.1 Watertightness and defects showed the results of three WTATs performed before the
excavation of three TBM maintenance shafts required for
One of the roles of the retaining walls used in deep exca- the construction of the high speed train (HST) tunnel in
vations is to prevent groundwater lateral inflow, which Barcelona. The shafts were located near the Sagrada
facilitates dewatering, and for this reason, they must have Familia Basilica at the sites of Padilla (Pujades, Vázquez-
a very low hydraulic conductivity. However, retaining Suñé, Carrera, Vilarrasa, et al., 2014), Bruc (Pujades,
walls commonly have defects (Bruce et al., 1989) that Vázquez-Suñé, Carrera, Vilarrasa, et al., 2014) and Trinx-
increase their effective hydraulic conductivity and allow ant (Pujades, Vázquez-Suñé, Carrera, & Jurado, 2014).
the groundwater to flow through them. Defects produce Pujades, Vázquez-Suñé, Carrera, Vilarrasa, et al. (2014)
different kinds of difficulties depending on their nature performed two WTATs. The retaining walls of one of the
and relative location with respect to the excavation bottom. excavations were made up of the combination of dia-
If defects are located above the excavation depth, ground- phragm walls and jet-grouting piles (shaft of Padilla), while
water inflows may drag sediments from behind the retain- only diaphragm walls were used for the enclosure of the
ing walls, which leads to the formation of sinkholes second shaft (Bruc shaft). As expected, the WTAT per-
outside the excavation (Pujades et al., 2012; Vilarrasa, formed in the Padilla shaft revealed that the hydraulic con-
Carrera, Jurado, Pujades, & Vázquez-Suné, 2011). If ductivity of the jet-grouting enclosure was relatively high
defects are located below the excavation bottom, the water (0.6 m/d), and groundwater could inflow easily inside the
pressure at the bottom is higher than expected, leading to excavation enclosure. The dewatering system was
liquefaction or bottom uplift phenomena, which can pro- redesigned considering the hydraulic conductivity of the
Table 1

38
Main characteristics of the considered studies. The column ‘‘Topic” refers to the addressed issue(s) and it is also related with the section(s) of this paper in which each study is classified. D is retaining
walls deflections (Section 3.2); I is dewatering impacts (Section 4.2); F is faulty retaining walls (Section 3.1); SF is design of dewatering systems and stability of the excavation bottom (Section 4.1); R is
impact mitigation by artificial recharge (Section 5).
Reference Location Topic Max. Max. Depth (m)
deflection settelment
Excavation Pumping wells Recharge wells Walls
(mm) (mm) (screen) (screen)
Feng and Lu (2015) Nanchang, China F/D 20
12 15.5–17.8 19.5–21.8 20.5
Pujades, Vázquez-Suñé, Carrera, Vilarrasa, Barcelona, Spain (Padilla F/SF/I <5 41 61.5 61.5
et al. (2014) and Pujades et al. (2016) shaft)
Barcelona, Spain (Bruc F/SF/I 2.4 40 60 60
shaft)
Pujades, Vázquez-Suñé, Carrera and Jurado Barcelona, Spain F/SF/I 2 33 55 50
(2014) (Trinxant shaft)
Russo et al. (2016) Napoli, Italy (San D 40 24 27 — 50

E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45

Pasquale station)
Serrano-Juan et al. (2016) Barcelona, Spain (Sagrera D 6 20 35 35
station)
Shen et al. (2017) Shanghai, China (East I 24.24–26 52 46–55
Nanjing Road)
Tan and Lu (2017) Shanghai, China (Central F/D 40 100 31.4–33.1 53 (26.5–30, 40–52) 55
business district)
Wang et al. (2016) Shanghai, China SF/I 17.4–19.5 30.8–
(Qilianshan South Road) 34.2
Shanghai, China (Xinzang SF 23–25 38–40
South Road)
Shanghai, China SF 30.8–32.3 59–62
(Hanzhong Road)
Wang et al. (2017) Hangzhou, China SF 26 58 (37–41/49–58); 45–49 52
(Qianjiang Century City) (37–41; 37–43)
Wang et al. (2018) Shanghai, China (Yishan I
Road)
Wu, Shen, Xu et al. (2015) and Wu, Shen, Yin, Hangzhou, China F/SF/I <5 24 43–50 43–54
et al. (2015)
Xing et al. (2016) Shanghai, China D <10 <10 16.3 — 33.5
You et al. (2018) Nanjing, China SF/R <95 9.5–27 51 (6–22, 28–50) 40–52
Zeng et al. (2018) Tianjin, China (Beizhan D 10 <5 17.6–23 2–20.6 33
station)
Zeng et al. (2019) Tianjin, China (Test 1) R < 50 21 28–31.5 46 (26,29,38– 37.9
44)
Tianjin, China (Test 2) R 28–31.5 51 (40–50)
Zhang, Li, et al. (2017) and Zhang, Wang, Shanghai, China R <17 33.1 73 (63–72) 77 (63–75) / 80 59–62
et al. (2017) (Hanzhong Road) (63–78)
Zhang et al. (2018) Singapore (Cashew D
15
120 20 — 29
station)
Zheng et al. (2014) Tianjin, China (Beizhan D 10 17.6–23 2–20.6 33–42
station)
Zheng et al. (2018) Tianjin, China R <1.8 21 28–31.5 51 (40–50) 37.9
Zhu et al. (2014) Shanghai, China (Yishan I 30 30.6 59/61 48–62
Road)
 E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45 39

jet-grouting. Conversely, no defects were observed during based synthetic aperture radar (Gb-SAR) during a pre-
the WTAT in the Bruc shaft. Pujades, Vázquez-Suñé, excavation dewatering (PED) test for the construction of
Carrera and Jurado (2014) also showed the results of a an intermodal station (railway, bus, subway, HST) in Bar-
WTAT performed before the excavation of the Trinxant celona, Spain. The Gb-SAR sensor was located on one side
shaft. In this case, discrepancies between expected and of the excavation, and the maximum measured displace-
observed drawdown indicated that the retaining walls ment was approximately 6 mm (Table 1). The advantage
had defects. Wu, Shen, Xu, and Yin (2015) showed the of using a Gb-SAR sensor is that it allows observation of
results of a WTAT that was performed to characterize the behavior of all the retaining walls located inside its field
the retaining walls of a deep excavation in Hangzhou, of vision, while inclinometers only provide punctual infor-
China. The excavation was required for the construction mation. Zheng, Zeng, Diao, and Xue (2014) observed and
of an interchange subway station. The results indicated modeled the retaining wall deflection that occurred during
the existence of a leak through a defect. The authors asso- a PED test for the construction of a subway station in
ciated the location of the faulty zone with difficulties that Tianjin, China. The numerical model was used to study
occurred during the construction of the retaining walls the mechanism that produced the deflections. Zheng et al.
through a gravel layer. Afterwards, Wu, Shen, Yin, and (2014) did not pay attention to outside settlements; in con-
Xu (2015) developed a numerical model to simulate the trast, they only considered the soil deformations occurring
WTAT. They varied the hydraulic conductivity at different inside the excavation, which reflected the influence of the
areas of the retaining walls to simulate the faulty zone. soil-wall friction. Xing, Xiong, and Wu (2016) developed
Finally, Pujades, Jurado, Carrera, Vázquez-Suñé, and a similar investigation as they observed and modeled the
Dassargues (2016) proposed a method for the characteriza- wall deflections and the surface settlements that occurred
tion of circular enclosures. This method is based on the during the construction of a station in Shanghai, China.
interpretation of a WTAT by using diagnostic plots and The main difficulty in this construction was that the new
allows determining the general state of the retaining walls station was constructed below a pre-existing subway sta-
and locating defects. The method was tested with data from tion. Consequently, it was of paramount importance to
the WTAT performed at the Padilla shaft (Pujades, minimize the retaining wall deflections and, in general, soil
Vázquez-Suñé, Carrera, Vilarrasa, et al., 2014). deformations. Xing et al. (2016) used a numerical model to
compute them under different scenarios. The numerical
3.2 Deflections results allowed proposing an alternative construction
method for mitigating deformations. This method con-
Some authors have focused their attention on the deflec- sisted of reinforcing the retaining walls by using a square
tion of retaining walls associated with groundwater varia- box formed by jet grouting soil piles. The adopted method
tions and its relation with soil settlements. Deflections proved to be useful, and the maximum observed retaining
occur when the balance of the forces affecting the walls is wall deflections and surface settlements were less than
modified because of excavation or dewatering. There are 10 mm (Table 1). The importance of predicting and moni-
different types of deflections, as well as, associated surface toring the retaining wall deflections was highlighted by
settlements (Fig. 2). Russo, Nicotera, and Autuori (2016), who focused their
In this review, only studies in which groundwater behav- investigation on the horizontal displacements of the retain-
ior plays a special role have been included. Wall deflections ing walls that occurred during the construction of the San
are commonly measured by using inclinometers embedded Pasquale Station in Napoli, Italy. This station was located
inside the retaining walls (Zhang et al., 2018). However, near historical buildings, which could be damaged by
some authors have proposed alternative methods to excessive soil displacements. In this context, Russo et al.
improve the measurements. Specifically, Serrano-Juan (2016) monitored and modeled the horizontal displace-
et al. (2016) observed wall deflections by using a ground- ments that occurred during the construction. The maxi-
mum displacements, which were measured with
inclinometers, reached up to 40 mm (Table 1). Similarly,
Zeng, Xue, Zheng, Xue, and Mei (2018) studied the retain-
ing wall deflections induced by PED tasks and their rela-
tionship with soil vertical deformations (i.e., settlements)
that occurred outside the excavation. This study was
focused on the construction of a subway station in Tianjin,
China. Initially, they developed a numerical model for
computing the retaining wall deflections, which was vali-
dated with real measurements (maximum retaining wall
deflection reached 10 mm). They also predicted the surface
settlements assuming that they were consequence of
Fig. 2. Different types of deflections and surface settlements (Chheng & ground loss behind the wall due to wall deflection and
Likitlersuang, 2018). using the empirical solution proposed by Hsieh and Ou
40 E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45

(1998). They also assessed the role of the soil hydraulic Carrera, Vilarrasa, et al., 2014). Fluidization might take
conductivity on the retaining wall deflections and surface place when the excavation is undertaken in unconfined
settlements. The system behavior was also analyzed by con- aquifers, and bottom uplift might occur when the excava-
sidering the presence of intercalated aquitards between tion bottom is located above a confined aquifer
aquifers. The main findings were that (1) the PED can (Cashman & Preene, 2001). An excavation can be per-
induce centimetric wall deflections and surface settlements, formed safely as long as the vertical total stress in the
and (2) the wall displacements and soil settlements occur ground (rV ) exceeds the pore water pressure (PW). This
during PED increase when the hydraulic conductivity of gives rise to a ‘‘safety factor” (SF),
the soil is high. Consequently, in the case of a multilayered rV
aquifer system, wall deflections and surface settlements SF ¼ ; ð1Þ
PW
increase as more aquifers are affected by pumping.
Zhang et al. (2018) developed a numerical model to which defines the excavation to be safe if SF = 1, but it is
compute the retaining wall deflections and settlements dur- common to require safety factors greater than 1 (i.e.,
ing the construction of an underground station in Singa- Pujades, Vázquez-Suñé, Carrera, & Jurado, 2014; Wu
pore. The numerical model was validated with real et al., 2015) to account for the insufficient characterization
measurements and used to predict the behavior of the sys- of the system. In practice, positive effective stresses are
tem. The maximum retaining wall deflection and surface essential at all stages of the construction process to work
0
settlement during the excavation were 15 and 120 mm in safe conditions. The vertical effective stress (rV ) can be
(Table 1), respectively. The authors also assessed the influ- calculated using Terzaghi’s equation in the vertical direc-
ence of groundwater drawdown outside the excavation. tion (Terzaghi & Peck, 1948) as follows:
They showed that the deflection of the retaining walls 0
rV ¼ rV  P W : ð2Þ
increases and settlements decrease as the groundwater
drawdown outside the excavation is reduced. In addition, The total rV is given by the following:
they observed that (1) the depth at which the maximum
rV ¼ zcS ; ð3Þ
wall deflection occurs does not depend on the groundwater
drawdown and (2) the magnitudes of the retaining wall and the water pressure as:
deflection and surface settlements are not only linearly pro-
P W ¼ hcW ; ð4Þ
portional to the drawdown but also depending on the
thickness of compressible layers. where z is the depth of the observation point, cS is the
Finally, it is interesting to note the cases reported by specific weight of the soil, h is the piezometric head at the
Feng and Lu (2015) and Tan and Lu (2017). They observed observation point, and cW is the specific weight of the
wall deflections and surface settlements during deep exca- water.
vations in which failures occurred. Feng and Lu (2015) Some authors have shown how to compute the SF to
observed wall deflections up to 20 mm (Table 1). The max- ensure the hydraulic stability of the excavation bottom.
imum retaining wall deflection occurred during the failure. Pujades, Vázquez-Suñé, Carrera, Vilarrasa, et al. (2014)
However, given that the maximum displacement was less analyzed different dewatering schemes for the excavation
than the critical action value, retaining wall displacements of two TBM shafts in Barcelona, Spain. The authors con-
were attributed to the normal behavior of the excavation. sidered three different scenarios: (1) short retaining walls
Surface settlements, which were nearly constant, increased that were constructed at the minimum depth required
up to 10 mm after the failure. The maximum wall deflection structurally, (2) long retaining walls that were deeper than
and surface settlement observed by Tan and Lu (2017) the minimum depth required structurally, and (3) an enclo-
reached 40 and 100 mm (Table 1), respectively. The retain- sure made up by the combination of diaphragm walls and
ing wall deflections were lower than the maximum critical jet-grouting piles. In the last scenario, the depth of the dia-
value, and the failure did not promote their increase. In phragm walls was the minimum required structurally, while
fact, the measured retaining wall displacements were simi- jet-grouting piles were used to increase the depth of the
lar at observation points located at different distances from enclosure up to an impervious layer. In all the scenarios,
the failure. In contrast, surface settlements increased the minimum imposed SF was ¥ 1.2, and it was computed
abruptly near the failure (up to 50 mm at some observation for each meter below the bottom of the excavation because
points). the soil was heterogeneous and the drop of pressure was
different in each geological layer. Similarly, Pujades,
4 Design of dewatering systems Vázquez-Suñé, Carrera and Jurado (2014) proposed a
methodology to improve the construction of deep excava-
4.1 (Hydraulic) stability of the excavation bottom tions in urban areas. This methodology was applied to
the Trinxant shaft of the HST tunnel in Barcelona, and
Excavations beneath the water table can suffer from the SF was calculated at different depths for four different
potential problems caused by groundwater, including flu- scenarios. The retaining walls (50 m depth) and the dewa-
idization and bottom uplift (Pujades, Vázquez-Suñé, tering well (55 m depth) have the same characteristics in
 E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45 41

all the scenarios. The difference between them was the Surface settlements are truly concerning because they
imposed drawdown in the dewatering well, which was 25, can damage nearby structures (Forth, 2004) (i.e., buildings
30, 35 and 40 m for scenarios 1 to 4, respectively. Stable or other infrastructures), while groundwater drawdown
conditions (SF ¥ 1.2) were achieved with 40 m of draw- affects the available water resources, which is especially rel-
down in the dewatering well, which allowed for draining evant in arid and semiarid regions. Both impacts are
materials with low values of hydraulic conductivity (alter- related since surface settlements are consequence of
nating layers of clay-silt and sands). Similarly, Wu, Shen, groundwater drawdown. In fact, surface settlements are
Xu, et al. (2015) and Wu, Shen, Yin, et al. (2015) evaluated proportional to groundwater drawdown (Bear &
the SF for the construction of a metro station in Corapcioglu, 1981a, 1981b; Cashman & Preene, 2001),
Hangzhou, China. The station was excavated in a gravel except in the vicinity of dewatering wells (Pujades et al.,
aquifer between retaining walls, and the maximum excava- 2017). Assuming only vertical deformations, settlement at
tion depth was 24 m (below the ground surface). Several the surface can be computed as follows:
pumping tests were conducted to determine whether the
q ¼ cW sDa; ð5Þ
original depth of the retaining walls (43 m) was sufficient
to maintain safety at the base of the excavation. Finally, where q is the surface settlement, s is the head drop, D is
the retaining walls were elongated to a depth of 54 m to the aquifer thickness and a is the soil compressibility. a
effectively protect the base of the excavation from the can be easily derived from the storage coefficient (S) if
upward seepage and to satisfy a SF ¥ 1.1 set up by the the aquifer is overconsolidated and behaves elastically.
Chinese Design Code guidelines. The same SF was consid- Considering the equation proposed by Jacob (1950) and
ered by Wang et al. (2016), who aimed to understand the cited by Ferris, Knowles, Brown, and Stallman (1962),
interaction between the retaining walls and dewatering  a
wells and their effect on the subsidence associated with S ¼ cW hD b þ ; ð6Þ
h
intensive dewatering of station foundation pits in the sub-
way system in Shanghai (China). Finally, You, Yan, Xu, where h is the porosity and b is the water compressibility, it
Liu, and Che (2018) proposed different dewatering schemes is possible to assume that S S ¼ a since b is very small com-
for the construction of a deep foundation pit in Nanjing, pared with a, where S S is the specific storage coefficient,
China near the Yangtze River. The foundation pit con- which can be obtained from pumping tests.
sisted of three ramps (A, B and C) and a main road con- The design of the dewatering system (i.e., retaining walls
necting a cross-river tunnel. The shape of the foundation and dewatering wells) plays a special role in mitigating the
pit was irregular, and the depth was not uniform. The max- potential impacts, and it must be conscientiously chosen.
imum depth of the foundation pit was 27 m. To guarantee This section is focused on studies that report dewatering
the safety of the pit excavation (SF ¥ 1.1) and ensure the impacts and assess different designs of dewatering systems
stability of the Yangtze River levee, different dewatering to mitigate them. Note that surface settlements can also
requirements were proposed. To conclude, it is necessary be produced by the deflection of retaining walls, but litera-
to mention the study developed by Wang et al., 2017. This ture focused on this kind of settlement is included in
study considered the effects of non-Darcy flow caused by Section 3.2.
the combination of dewatering wells and retaining walls. Zhu, Huang, Tan, and Chen (2014) analyzed the soil
They analyzed the area affected by non-Darcy flow under displacements induced by the dewatering undertaken for
different scenarios in which the geometry of the dewatering the construction of the Yishan Road station of the Shangai
wells and the retaining walls was modified. The objective Metro Line 9, which is a four-story underground station.
was to improve the efficiency of dewatering systems taking The maximum measured surface settlements reached
into consideration that drawdown under Darcy conditions 30 mm (Table 1). Zhu et al. (2014) monitored the settle-
is smaller than that produced under non-Darcy flow. Con- ments at different depths by using a high-precision
sequently, if the area affected by non-Darcy flow is multiple-position extensometer. Measurements revealed
increased, a lower pumping rate is required to produce that settlements in some geological layers were higher than
the same drawdown. From this study, it is possible to con- those observed at the surface. This fact would indicate that
clude that the drawdown necessary to reach a required SF some layers expanded during the dewatering, which miti-
is overestimated when non-Darcy flow is not considered. gated the surface settlements. This behavior was subse-
quently corroborated by Wang et al. (2018) by using a
4.2 Impacts physical model test and a numerical model. Similarly,
Xu, Shen, Ma, Sun, and Yin (2014) studied the influence
Impacts associated with dewatering of deep excavations of the depth of the retaining walls on outside drawdown.
are a relevant issue that is regarded in most studies. The This study arose from experiences during underground
two most meaningful impacts produced by dewatering constructions in Shanghai, China. They undertook labora-
actions are groundwater drawdown (Fig. 3) and soil defor- tory tests and developed numerical models to analyze the
mations (i.e., surface settlements), both of which occur out- blocking effect of retaining walls and how the outside
side the excavation (You et al., 2018). groundwater drawdown decreases when the depth of the
42 E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45

Fig. 3. Example of drawdown produced by the dewatering of a confined aquifer (Wang et al., 2017).

retaining walls is increased. They proposed a concept called water drawdown and surface settlements that occurred out-
‘‘effective depth ratio” that relates the aquifer thickness and side the excavation enclosure during the construction of a
the depth of the retaining walls. The ‘‘effective depth ratio” subway station in Hangzhou, China. They monitored the
allows estimating the magnitude of the groundwater draw- groundwater and soil behaviors during two pumping tests.
down occurring outside the excavation enclosure during The first test was performed with a 43 m retaining wall
the dewatering. Pujades, Vázquez-Suñé, Carrera, depth, while the retaining wall reached a 53 m depth during
Vilarrasa, et al. (2014) and Pujades, Vázquez-Suñé, the second test. Observations corroborated that groundwa-
Carrera and Jurado (2014) showed the procedure to design ter drawdown outside the excavation enclosure decreased
three dewatering systems for excavating maintenance when the retaining walls were elongated. As a result, a
shafts required for the construction of the HST tunnel in reduction in the surface settlements was also expected.
Barcelona, Spain. They used numerical models to compute However, monitoring data indicated that the surface settle-
groundwater drawdown and surface settlements under dif- ments during the two tests were similar. In addition, no
ferent scenarios in which the depth of the retaining walls pattern relating the evolution of the surface settlements
was modified. The optimum dewatering system was pro- and the pumping tests was observed, which would indicate
posed considering the groundwater drawdown, the surface that the surface settlements were influenced by other fac-
settlements and the SF in the excavation bottom. The most tors or activities. Anyway, the magnitude of the measured
relevant information is related to the monitored surface surface settlements during both pumping tests was very low
settlements during the excavation of the three shafts. Mea- (ffi3 mm), which would indicate that the soil was overcon-
surements indicated that the soil was overconsolidated solidated. Wu, Shen, Yin, et al. (2015) developed a numer-
because it behaved elastically before groundwater draw- ical model considering the measured data during these two
down. The maximum measured surface settlement was pumping tests. The numerical model was used to analyze
lower than 5 mm (Table 1), which is a relatively low mag- the blocking effect of the retaining walls. Groundwater
nitude considering that groundwater drawdown inside the drawdown and surface settlements during a dewatering
excavation was approximately 50 m. Measurements also period of 180 days were estimated considering the two
showed that surface settlements were produced not only depths for the retaining walls (43 and 53 m). The results
by dewatering tasks but also by other actions, namely, showed that the volume of pumped groundwater, the
the construction of the retaining walls, which produced groundwater drawdown and the surface settlements that
surface settlements even higher than those induced by occurred outside the excavation increased dramatically
pumping. Similarly, Wu, Shen, Xu et al. (2015) analyzed when the depth of the retaining walls was reduced. Wang
the influence of the depth of the retaining walls on ground- et al. (2016) investigated the influence of the depth of the
 E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45 43

retaining walls and dewatering wells on the impacts that

occurred outside the excavation. They developed research Head with recharge
in the context of the construction of the Shanghai rail tran- Building Head without recharge
sit lines and used a standard geometry of a railway station
Q
as a reference. The results showed how surface settlements Q
vary considerably depending on the configuration of the
dewatering system (i.e., the depth of the retaining walls
and the dewatering wells). An ideal distance between the
bottom of the pumping wells and the retaining walls was Bottom
proposed to minimize the surface settlements in an efficient Aquitard
manner. Most of the existing studies are based on the mod-
eling of real sites and the prediction of the system behavior Confined aquifer
under different circumstances. For this reason, the study
presented by Shen, Wu, and Misra (2017) is especially rel-
evant. Contrary to previous studies, Shen et al. (2017) pro- Recharge Pumping
posed an analytical equation to compute the groundwater well well
drawdown difference between the inside and outside of
the excavation enclosure. This equation is useful during
the design stage of dewatering systems because it allows Fig. 4. Example of artificial recharge by recharge wells (modified from
Zhang, Li, et al., 2017).
easy prediction of the expected drawdown outside the exca-
vation and, therefore, of the surface settlements. The pro-
posed solution was validated by comparing its results vation phase, which agreed with the results obtained in the
with numerical and real data. previous test, were also analyzed. In these two studies
(Zhang, Li, et al., 2017; Zhang, Wang, et al., 2017), artifi-
5 Adopted measures for impact mitigation cial recharge proved to be a useful method to mitigate
dewatering impacts. Similarly, Zheng, Cao, Cheng, Ha,
Artificial recharge outside the excavation is sometimes and Wang (2018) proposed a numerical study to estimate
considered to reintroduce the pumped water into the aqui- the effects of artificial recharge by deep wells prior to the
fer and mitigate impacts (i.e., settlements and groundwater construction of a subway station in Tianjin, China. In this
drawdown) induced by the dewatering tasks. The artificial study, some tests were performed near the excavation site
recharge system usually consists of wells located outside of before the construction of the walls. Zheng et al. (2018)
the excavation enclosure (Fig. 4). undertook and modeled three types of tests: single-well
Some studies reporting and studying numerically the tests (pumping or recharge), multiwell tests (pumping and
efficiency of artificial recharge systems have been published recharge using different wells) and combined-recharge tests
in recent years. For instance, Zhang, Li, Wang, Chen, and from two nearby wells. The single-well tests showed how
Zhu (2017) monitored and modeled a pumping-recharge the soil responded to pumping or recharge events, whereas
test before the construction of the Hanzhong Road station the results of the multiwell tests probed whether artificial
of subway line 13 at the Jing’an District in Shanghai, recharge was a useful method to control dewatering settle-
China. The planned excavation was surrounded by multi- ments. The objective of the combined-recharge tests was to
ple structures, including four roads, four high-rise build- ascertain whether settlements induced during redevelop-
ings, one subway station and the Suzhou River, which ment phases (to eliminate clogging) could be mitigated by
has a flood control wall. The objective was to ascertain recharging water at nearby wells. The results showed that
the effects of future dewatering on the buildings and on redevelopment settlements can be controlled. You et al.
the flood control wall. The results showed that artificial (2018) also showed the usefulness of recharge wells to mit-
recharge was useful to mitigate dewatering settlements. In igate surface settlements produced by dewatering tasks.
fact, settlements were eliminated on the side where the They controlled surface settlements by installing 27
recharge wells were installed since recharged water pro- recharge wells. Finally, Zeng et al. (2019) reported a case
duced the vertical expansion of the aquifer. In the same in which artificial recharge was used to minimize pumping
way, artificial recharge also mitigated horizontal displace- settlements during a deep excavation near a religious build-
ments induced by pumping. Later, Zhang, Wang, Chen, ing. Five recharge wells located between the excavation and
and Li (2017) developed a deeper study focused on artificial the building were used to mitigate settlements during the
recharge at the same location. The authors studied the sys- excavation phase. Measurements showed that settlements
tem behavior during a more realistic test and the excava- were stabilized when the recharge wells were started. How-
tion phase. In the preliminary test, whose results were ever, settlements increased during seven redevelopment
analogous to those obtained in the previous study (i.e., periods, which were undertaken to avoid clogging effects.
Zhang, Li, et al., 2017), the pumping rate was increased This observation motivated a test performed in similar
to simulate real conditions. Measurements during the exca- materials, whose objective was to ascertain if it was possi-
44 E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45

ble to mitigate redevelopment settlements. The test, which manner, more research is needed to better establish the sys-
was modeled, showed that redevelopment settlements can tem behavior during the redevelopment of recharge wells.
be eliminated by recharging from a nearby well during
the redevelopment phases. Acknowledgements

A. Jurado gratefully acknowledges the financial support

6 Discussion and conclusions from the Excellence Initiative of the German Federal and
State Governments through the ‘‘Open Topic Postdoc
Numerous studies in the literature have focused on Position” Fellow at the Technische Universität Dresden.
topics related to underground excavations below the water
table. Most of these studies are based on real construction
Declaration of Competing Interest
projects developed in Asia, especially in China, which
would be related to the increase in population in this area
No potential conflict of interest is reported by the
and therefore to the need to construct new underground
authors.
infrastructures. However, this fact also indicates the
increasing economic conditions in Asia that allow for
investments in new infrastructures. References
Defects in retaining walls are relatively common and
may affect the safety of construction and nearby structures. Bear, J., & Corapcioglu, M. Y. (1981a). Mathematical model for regional
land subsidence due to pumping 2. Integrated aquifer subsidence
In addition to the studies that report cases of faulty retain- equations for vertical and horizontal displacements. Water Resources
ing walls, some authors proposed methods to characterize Research, 17(4), 947–958.
them. However, these methods were developed by assum- Bear, J., & Corapcioglu, M. Y. (1981b). Mathematical model for regional
land subsidence due to pumping 1. Integrated aquifer subsidence
ing homogenous media. It would be interesting to go fur- equations based on vertical displacements only. Water Resources
ther and adapt the existing methods to heterogeneous Research, 17(4), 937–946.
media. Bruce, D. A., DePaoli, B., Mascardi, C., & Mongilardi, E. (1989).
Monitoring and quality control of a 100 meter deep diaphragm wall. In
Most of the literature focused on the deflections of DFI international conference on piling and deep foundations, London
retaining walls does not consider the influence of ground- (pp. 15–18).
water. However, the studies collected in this review show Calin, N., Cristian, R., & Ioan, B. (2017). Dewatering system of a deep of
excavation in urban area–Bucharest case study. Procedia Engineering,
that groundwater plays an important role. In fact, impor- 209, 210–215.
tant deflections can be produced only by dewatering tasks Cashman, P. M., & Preene, M. (2001). Groundwater lowering in construc-
(e.g., during pre-excavation dewatering or WTAT). Future tion—A practical guide. London: Spon Press.
Chheng, C., & Likitlersuang, S. (2018). Underground excavation
research must be devoted to clarify the real influence of behaviour in Bangkok using three-dimensional finite element method.
groundwater. Computers and Geotechnics, 95, 68–81.
The topic to which authors pay more attention is the Culı́, L., Pujades, E., Vázquez-Suñé, E., & Jurado, A. (2016). Modelling of
the EPB TBM shield tunnelling advance as a tool for geological
impacts produced by dewatering tasks, among which dewa- characterization. Tunnelling and Underground Space Technology, 56,
tering settlements stand out. This fact is understandable 12–21.
since some authors have attributed large soil deformations Di, H., Zhou, S., Xiao, J., Gong, Q., & Luo, Z. (2016). Investigation of the
long-term settlement of a cut-and-cover metro tunnel in a soft deposit.
to dewatering. However, recent studies have demonstrated Engineering Geology, 204, 33–40.
that surface dewatering settlements could be smaller than Feng, S. J., & Lu, S. F. (2015). Failure of a retaining structure in a metro
expected. This fact may be related to processes not consid- station excavation in Nanchang City, China. Journal of Performance of
Constructed Facilities, 30(4), 04015097.
ered by commonly used analytical equations, such as the Ferris, J., Knowles, D., Brown, R., & Stallman, R. (1962). Theory of
vertical expansion of layers or the nonproportionality aquifer tests. U.S. Geological Survey, Water Supply Paper 1536-E.
between drawdown and surface settlements. This issue Forth, R. A. (2004). Groundwater and geotechnical aspects of deep
excavations in Hong Kong. Engineering Geology, 72(3/4), 253–260.
requires more numerical and analytical investigation. Gulhati, S. K., & Datta, M. (2005). Geotechnical engineering. New Delhi:
Methods adopted by most researchers to design dewa- McGraw-Hill.
tering systems, in which the stability of the excavation bot- Hsieh, P. G., & Ou, C. Y. (1998). Shape of ground surface settlement
profiles caused by excavation. Canadian Geotechnical Journal, 35(6),
tom is guaranteed, are similar. It would be interesting to 1004–1017.
pay more attention to the influence of non-Darcy flows, Jacob, C. E. (1950). Flow of groundwater. In H. Rousep (Ed.),
as suggested by Wang et al. (2017), which might be relevant Engineering hydraulics (pp. 321–386). New York: John Wiley.
Konda, T. (2001). Shield tunneling method. Civil Engineering, Japan
for improving the efficiency of dewatering systems. Society of Civil Engineers, 39, 23–27.
Artificial recharge as a tool for mitigating dewatering Mouratidis, A. (2008). The ‘‘cut-and-cover” and ‘‘cover-and-cut” tech-
impacts is a relatively new investigation field, which is niques in highway engineering. Electronic Journal of Geotechnical
Engineering (EJGE), 13(Bundle F), 1–15.
reflected in the low number of scientific studies published Pappa, M., & Benardos, A. (2007). Comparative evaluation of alternative
to date. Consequently, more investigation is needed. methods for the construction of underground car parks. In Proceedings
Specifically, it would be interesting to propose analytical of the 11th ACUUS international conference underground space:
Expanding the frontiers (pp. 10–13).
solutions to design artificial recharge systems in an easy Pulido, D., Darido, G., Munoz-Raskin, R., & Moody, J. (Eds.) (2018).
way as a previous step to numerical modeling. In the same The urban rail development handbook. The World Bank.
 E. Pujades, A. Jurado / Underground Space 6 (2021) 35–45 45

Pujades, E., Carrera, J., Vázquez-Suñé, E., Jurado, A., Vilarrasa, V., & Xing, H., Xiong, F., & Wu, J. (2016). Effects of pit excavation on an
Mascuñano-Salvador, E. (2012). Hydraulic characterization of dia- existing subway station and preventive measures. Journal of Perfor-
phragm walls for cut and cover tunnelling. Engineering Geology, 125, mance of Constructed Facilities, 30(6), 04016063.
1–10. Xu, Y. S., Shen, S. L., Ma, L., Sun, W. J., & Yin, Z. Y. (2014). Evaluation
Pujades, E., Vázquez-Suñé, E., Carrera, J., Vilarrasa, V., De Simone, S., of the blocking effect of retaining walls on groundwater seepage in
Jurado, A., ... Lloret, A. (2014). Deep enclosures versus pumping to aquifers with different insertion depths. Engineering Geology, 183,
reduce settlements during shaft excavations. Engineering Geology, 169, 254–264.
100–111. You, Y., Yan, C., Xu, B., Liu, S., & Che, C. (2018). Optimization of
Pujades, E., Vázquez-Suñé, E., Carrera, J., & Jurado, A. (2014). dewatering schemes for a deep foundation pit near the Yangtze River,
Dewatering of a deep excavation undertaken in a layered soil. China. Journal of Rock Mechanics and Geotechnical Engineering, 10(3),
Engineering Geology, 178, 15–27. 555–566.
Pujades, E., Jurado, A., Carrera, J., Vázquez-Suñé, E., & Dassargues, A. Zeng, C. F., Xue, X. L., Zheng, G., Xue, T. Y., & Mei, G. X. (2018).
(2016). Hydrogeological assessment of non-linear underground enclo- Responses of retaining wall and surrounding ground to pre-excavation
sures. Engineering Geology, 207, 91–102. dewatering in an alternated multi-aquifer-aquitard system. Journal of
Pujades, E., De Simone, S., Carrera, J., Vázquez-Suñé, E., & Jurado, A. Hydrology, 559, 609–626.
(2017). Settlements around pumping wells: Analysis of influential Zeng, C. F., Zheng, G., Xue, X. L., & Mei, G. X. (2019). Combined
factors and a simple calculation procedure. Journal of Hydrology, 548, recharge: A method to prevent ground settlement induced by redevel-
225–236. opment of recharge wells. Journal of Hydrology, 568, 1–11.
Russo, G., Nicotera, M. V., & Autuori, S. (2016). San Pasquale Station of Zhang, Y. Q., Li, M. G., Wang, J. H., Chen, J. J., & Zhu, Y. F. (2017).
Line 6 in Naples: Measurements and numerical analyses. Procedia Field tests of pumping-recharge technology for deep confined aquifers
engineering, 143, 1503–1510. and its application to a deep excavation. Engineering Geology, 228,
Serrano-Juan, A., Vázquez-Suñè, E., Monserrat, O., Crosetto, M., 249–259.
Hoffmann, C., Ledesma, A., ... Alcaraz, M. (2016). Gb-SAR interfer- Zhang, Y. Q., Wang, J. H., Chen, J. J., & Li, M. G. (2017). Numerical
ometry displacement measurements during dewatering in construction study on the responses of groundwater and strata to pumping and
works. Case of La Sagrera railway station in Barcelona, Spain. recharge in a deep confined aquifer. Journal of Hydrology, 548,
Engineering Geology, 205, 104–115. 342–352.
Shen, S. L., Wu, Y. X., & Misra, A. (2017). Calculation of head difference Zhang, W., Wang, W., Zhou, D., Zhang, R., Goh, A. T. C., & Hou, Z.
at two sides of a cut-off barrier during excavation dewatering. (2018). Influence of groundwater drawdown on excavation responses–
Computers and Geotechnics, 91, 192–202. A case history in Bukit Timah granitic residual soils. Journal of Rock
Tan, Y., & Lu, Y. (2017). Forensic diagnosis of a leaking accident during Mechanics and Geotechnical Engineering, 10(5), 856–864.
excavation. Journal of Performance of Constructed Facilities, 31(5), Zheng, G., Zeng, C. F., Diao, Y., & Xue, X. L. (2014). Test and numerical
04017061. research on wall deflections induced by pre-excavation dewatering.
Terzaghi, K., & Peck, R. B. (1948). Soil mechanics in engineering practice Computers and Geotechnics, 62, 244–256.
(2nd ed.). USA: John Wiley and Sons. Zheng, G., Cao, J. R., Cheng, X. S., Ha, D., & Wang, F. J. (2018).
UN - United Nations (2014). World Urbanization Prospects: The 2014 Experimental study on the artificial recharge of semiconfined aquifers
Revision. Department of Economic and Social Affairs, United involved in deep excavation engineering. Journal of Hydrology, 557,
Nations, New York (ISBN 978-92-1-151517-6). 868–877.
Vilarrasa, V., Carrera, J., Jurado, A., Pujades, E., & Vázquez-Suné, E. Zhou, X. J., Wang, J. H., & Hu, H. Y. (2013). Design of enclosure for
(2011). A methodology for characterizing the hydraulic effectiveness subway station foundation pit in southward extension line of Chengdu
of an annular low-permeability barrier. Engineering Geology, metro line one. Advanced Materials Research, 671, 1122–1125.
120(1/2/3/4), 68–80. Zhu, Y. F., Huang, Y. Z., Tan, Y. P., & Chen, J. J. (2014). Stratified
Wang, J., Wu, Y., Liu, X., Yang, T., Wang, H., & Zhu, Y. (2016). Areal settlement characteristics of the soil strata in Shanghai due to
subsidence under pumping well–curtain interaction in subway foun- dewatering. Journal of Aerospace Engineering, 28(6), A4014005.
dation pit dewatering: Conceptual model and numerical simulations.
Environmental Earth Sciences, 75(3), 198.
Wang, J., Liu, X., Wu, Y., Liu, S., Wu, L., Lou, R., ... Yin, Y. (2017). Further Reading
Field experiment and numerical simulation of coupling non-Darcy
flow caused by curtain and pumping well in foundation pit dewatering. Angel, M. L., Johnston, S., O’Steen, K., Brown, C., & Sando, T. (2015).
Journal of Hydrology, 549, 277–293. Groundwater control issues in transportation engineering: A short
Wang, J., Deng, Y., Ma, R., Liu, X., Guo, Q., Liu, S., ... Huang, X. review of dewatering methods and lessons learned. Journal of
(2018). Model test on partial expansion in stratified subsidence Engineering Science & Technology Review, 8(3), 8–13.
during foundation pit dewatering. Journal of Hydrology, 557, Hsiung, B. C. B., Yang, K. H., Aila, W., & Ge, L. (2018). Evaluation of
489–508. the wall deflections of a deep excavation in Central Jakarta using three-
Wu, Y. X., Shen, S. L., Xu, Y. S., & Yin, Z. Y. (2015). Characteristics of dimensional modeling. Tunnelling and Underground Space Technology,
groundwater seepage with cut-off wall in gravel aquifer. I: Field 72, 84–96.
observations. Canadian Geotechnical Journal, 52(10), 1526–1538. Serrano-Juan, A., Pujades, E., Vázquez-Suñè, E., Crosetto, M., & Cuevas-
Wu, Y. X., Shen, S. L., Yin, Z. Y., & Xu, Y. S. (2015). Characteristics of González, M. (2017). Leveling vs. InSAR in urban underground
groundwater seepage with cut-off wall in gravel aquifer. II: Numerical construction monitoring: Pros and cons. Case of la sagrera railway
analysis. Canadian Geotechnical Journal, 52(10), 1539–1549. station (Barcelona, Spain). Engineering Geology, 218, 1–11.