Tunnelling and Underground Space Technology 61 (2017) 145–149

Contents lists available at ScienceDirect

Tunnelling and Underground Space Technology

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

Basal heave stability of supported circular excavations in clay

Anthony T.C. Goh
School of Civil and Environmental Engineering, Nanyang Technological University, Singapore 639798, Singapore

a r t i c l e i n f o a b s t r a c t

Article history: Deep circular shafts are commonly used for example to construct access shafts for transit tunnels, pump
Received 11 April 2016 stations for wastewater tunnels, and launch shafts for tunnel boring machines. Since the earth pressures
Received in revised form 1 September 2016 acting on a circular shaft subjects the shaft support to ring compression, the reinforcement in the struc-
Accepted 13 October 2016
tural elements can be reduced and the need for internal support is eliminated, thereby speeding up exca-
vation. The design of circular support systems for excavations in clays involves assessing the ground
stresses and ground movements, the capacity of the structural elements, and the basal heave stability.
Keywords:
This study focused on assessing the basal heave stability of diaphragm wall supported circular excava-
Basal heave
Clay
tions in clays using the finite element method. The analyses have shown that the basal heave factor of
Circular excavation safety is dependent on the undrained shear strength of the clay, the geometrical properties of the exca-
Diaphragm wall vation system and the thickness of the soil stratum. Based on these results, a simplified method is
Factor of safety proposed for assessing the basal heave factor of safety for axisymmetric supported excavations in clay.
Stability Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction Most of the studies on the basal heave stability of circular excava-
tions in clays have focused on open and unsupported excavations
Tunnelling projects for transit railway and sewage systems using similar numerical techniques (Salencon and Matar, 1982;
often require the construction of deep excavation shafts to facili- Britto and Kusakabe, 1983; Khatri and Kumar, 2010).
tate the launching of tunnel boring machines as well as for access In this study, extensive finite element analyses were carried out
and ventilation purposes. These shafts can be either rectangular or to assess the basal heave factor of safety for circular excavations
circular in shape. Circular shafts are becoming increasingly popular supported by diaphragm walls. The parametric study was per-
due to improvements in construction methods and equipment. formed using the finite element software Plaxis (Brinkgreve et al.,
Circular shafts are often preferred because their high structural 2011). Based on these results, a simplified method is proposed for
stiffness through hoop forces can lead to less reinforcement in assessing the basal heave factor of safety for axisymmetric sup-
the structural elements and also minimize the ground displace- ported excavations in clay.
ment during excavation. In addition, the shafts can be constructed
without the use of internal struts, hence providing a relatively
2. Numerical study
obstruction free area for excavation works.
As with rectangular-shaped excavations, the design of circular
Fig. 1 shows schematically the cross section of the excavation
excavations in clays involves the assessment of the ground stresses
system, with a simplified soil profile comprising of a thick nor-
and ground movements, the capacity of the structural elements,
mally consolidated clay deposit. The Mohr Coulomb constitutive
and the basal heave stability. The majority of the studies involving
relationship was used to model the undrained behavior of the clay.
circular excavations have focused on the performance of the wall
For this study, cases with a homogeneous clay layer with constant
system and associated ground movements (Parashar et al., 2007;
undrained shear strength cu and cases with cu linearly increasing
Arai et al., 2007; McNamara et al., 2008; Kim et al., 2013; Tan
with depth were considered. The soil is assumed to be subjected
and Wang, 2015; Schwamb et al., 2014). While a number of
to undrained shearing during excavation. The thickness of the clay
researchers have examined the basal heave stability of rectangular
below the final excavation level is denoted as T in Fig. 1.
braced excavation systems (Bjerrum and Eide, 1956; Goh, 1994;
The analyses considered an axisymmetric excavation of diame-
Faheem et al., 2003, 2004), only limited studies (e.g., Cai et al.,
ter B supported by a stiff diaphragm wall system. Because of sym-
2002) have been carried out for circular supported excavations.
metry, only half the cross-section was considered. The soil was
modeled by 15-noded triangular elements. The wall structural
E-mail address: ctcgoh@ntu.edu.sg elements were assumed to be linear elastic and were modeled by

http://dx.doi.org/10.1016/j.tust.2016.10.005
0886-7798/Ó 2016 Elsevier Ltd. All rights reserved.
146 A.T.C. Goh / Tunnelling and Underground Space Technology 61 (2017) 145–149

Table 1
Summary of soil, wall and geometrical properties.

Parameter Symbol Range of values

Wall diameter (m) B 10–100
Final excavation depth (m) H 16–24
Wall penetration depth (m) D 0–24
Thickness to hard stratum (m) T 7–120
Wall stiffness (kN m2) EI 5  106
Soil unit weight (kN/m3) c 16
Soil undrained shear strength (kPa) cu 50–100
Soil friction angle (degrees) /u 0
Soil undrained stiffness ratio Eu/cu 300
Soil Poisson’s ratio t 0.495

failure. The reciprocal of this reduction factor F is identified as

the factor of safety. The iterative procedure to determine factor
of safety starts with assigning Fi = F0 = 1. Subsequently, Fi is
increased by DF to Fi+1 = Fi + DF where typically DF is set to 0.1.
With the updated F, for a Mohr-Coulomb material, the cohesion c
and tangent of the friction angle (tan /) are reduced as follows:
c
Fig. 1. Cross-sectional soil and wall profile.
ciþ1 ¼ ð1Þ
Fiþ1

tan /
5-noded beam elements. The nodes along the side boundaries of tan /iþ1 ¼ ð2Þ
Fiþ1
the mesh were constrained from displacing horizontally while
the nodes along the bottom boundary were constrained from mov- Using the reduced strength parameters, a new equilibrium is
ing horizontally and vertically. A typical very refined mesh com- sought by carrying out an elastic-plastic finite element (FE) analy-
prising 3834 elements and 31,337 nodes is shown in Fig. 2. The sis. If a new equilibrium is found, i.e. the analysis converges, then F
right vertical boundary which extends far from the excavation is increased and a new equilibrium is sought again with reduced
(
5B) to minimize the effects of the boundary restraints, is not strength parameters. The process is repeated until the analysis
shown in Fig. 2. The range of geometrical properties of the excava- does not converge or F remains constant with continuing deforma-
tion that were considered, and the assumed wall and soil proper- tion, within a certain specified number of iterations. In this study,
ties are shown in Table 1. this was set to the default of 100 iterations. This is described in
The construction sequence comprised the following steps: (1) detail in Brinkgreve and Bakker (1991), Matsui and San (1992),
the wall is installed (‘‘wished into place”) without any disturbance and Brinkgreve et al. (2011). This critical strength reduction value
in the surrounding soil; (2) the soil is excavated uniformly in 2 m is taken to be the vertical shaft basal heave factor of safety FSFE.
intervals until the final depth H is reached. The stability of the ver- Fig. 3 shows the results of a typical finite element (FE) shear
tical shaft was then determined using the shear strength reduction strength reduction analysis to compute the FSFE for a circular
technique. This technique has been used by various authors includ- excavation.
ing Matsui and San (1992), Griffiths and Lane (1999), Hammah
et al. (2007), Zhang and Goh (2012), and Do et al. (2013). The 3. Results
method is now available in many commercial finite element and
finite difference programs. For brevity, only some general trends are highlighted. The influ-
The shear strength reduction method involves progressively ence of the diameter of the supported shaft B is shown in Fig. 4 for
reducing the shear strength of the soil until the geotechnical struc- two cases with H = 16 m, T = 60 m and cu = 50 kPa. The basal heave
ture such as a slope or retaining wall is on the verge of global factor of safety FSFE decreases with the increase of the shaft

Fig. 2. Partial finite element mesh.

 A.T.C. Goh / Tunnelling and Underground Space Technology 61 (2017) 145–149 147
Strength Reduction Factor F

2.0
H = 16 m, B = 40 m, cu = 50 kPa
3.5
1.8
3
1.6
2.5
1.4 D=4m
FS FE
D = 10 m
2
1.2

1.0 1.5
0 20 40 60 80 100 120
Iteration step 1
0 20 40 60 80
Fig. 3. Typical plot of shear strength reduction factor F versus iterative step for
B = 40 m, H = 24 m, D = 4 m and cu = 80 kPa.
T (m)

Fig. 5. Effects of clay thickness T on FSFE.

H = 16 m, T = 60 m, cu = 50 kPa
Table 2
2.5
Comparison of stability number.

H/B Ns NcL NcU

0.2 – 6.897 7.523
2 0.24 7.162 – –
D=4m 0.27 7.183 – –
FSFE 0.4 7.615 7.303 8.104
D=0 0.6 8.136 7.866 8.608
1.5 0.8 8.463 8.37 9.034
1 8.905 8.771 9.429
1.33 9.509 – –
1.6 9.789 – –
2 10.464 9.973 11.008
1
0 1 2 3 4
B/H

Fig. 4. Effects of shaft diameter B on FSFE. 12

10
diameter. In addition, increasing the depth of penetration of the
wall below the final excavation depth D resulted in an increase
8
in the factor of safety. The increase in FSFE was more significant
for B/H < 1.
The influence of the thickness of the clay below the final exca- Ns 6
vation level T for two cases with H = 16 m, B = 40 m and cu = 50 kPa
is presented in Fig. 5. As expected, the presence of the hard stratum
N s = -0.4187(H/B) 2 + 2.769(H/B) + 6.5443
4
close to the excavation (reduced T) significantly increases the FSFE. R² = 0.9969
As in Fig. 4, the general trend was for FSFE to increase with increas-
2
ing D. These trends are similar to those reported by Cai et al. (2002)
for circular excavations and by Goh (1994) for two-dimensional
plane strain excavations. 0
For the cases with D = 0 (zero wall embedment depth), the 0 0.5 1 1.5 2 2.5
dimensionless stability number Ns was determined from the H/B
expression
Fig. 6. Stability number Ns.
FSFE cH
Ns ¼ ð3Þ
cu
where c is the soil unit weight. The values of Ns obtained from Eq. Based on regression analysis of all the cases analyzed, the basal
(3) are shown in Table 2. Also shown in Table 2 are the correspond- heave stability for supported vertical shafts can be determined as
ing lower and upper bound bearing capacity factors (NcL and NcU) follows:
for circular footings determined from finite element limit analysis Ns c u
(Salgado et al., 2004). The Ns values from this study were found FSshaft ¼ l l ð5Þ
cH D T
to fall between NcL and NcU.
The values of Ns for D = 0 obtained from Eq. (3) are also plotted The dimensionless correction factor lD which takes into consid-
in Fig. 6. A polynomial best fit results in the following: eration the depth of penetration of the wall D below the final exca-
vation depth is shown in Eq. (6) and Fig. 7. Since for stiff walls, the
Ns ¼ 0:4187ðH=BÞ2 þ 2:769ðH=BÞ þ 6:5443 ð4Þ basal heave mechanism occurs below the base of the wall (Wong
148 A.T.C. Goh / Tunnelling and Underground Space Technology 61 (2017) 145–149

1.5 4

uniform cu

linearly increasing cu
1 3

µ D = 0.6727(D/B) + 1.0367 Series1

µD
R² = 0.8867 FS Shaft

0.5
2

0
0 0.1 0.2 0.3 0.4 0.5 0.6 1
D/B 1 2 3 4
FSFE
Fig. 7. Wall depth correction factor lD.
Fig. 9. Plot of FSshaft versus FSFE.

2 Fig. 9 shows the plot of FSshaft obtained using Eq. (5) versus the FSFE
µT = 0.7753[(T-D)/B]-0.252 values for all the cases is reasonably accurate.
In Eq. (5), it has been assumed that the clay is homogeneous
R² = 0.952
1.5 with constant undrained shear strength cu. Table 3 shows addi-
tional finite element analyses that were carried out for cases in
which the undrained shear strength increases linearly with depth
µT 1 such that at any depth z below the ground surface
cu ¼ c0 þ mz ð8Þ
where c0 is the cohesion at ground surface and m is a dimensionless
0.5
factor that defines the rate at which the cohesion increases linearly
with depth.
For these cases, by using the average cu between a depth of (H
0 + D) and a depth of (H + D + 0.15B), Eq. (5) was found to be valid
0 0.1 0.2 0.3 0.4 0.5
and yielded satisfactory results as shown in Table 3 and Fig. 9.
(T-D)/B
Fig. 8. Clay thickness correction factor lT. 4. Conclusions

and Goh, 2002), the dimensionless correction factor lT takes into This paper has demonstrated that the finite element method
consideration the depth from the base of the wall to the hard stra- can be used to assess the basal heave stability of circular supported
tum (T – D). The value of lT is shown in Eq. (7) and Fig. 8. excavations. For supported diaphragm wall excavations in clays,
this study has shown that the basal heave factor of safety is depen-
lD ¼ 0:6727ðD=BÞ þ 1:0367 ð6Þ dent on the undrained shear strength of the clay, the geometrical
properties of the excavation system and the thickness of the soil
lT ¼ 0:7753½ðT  DÞ=B0:252 forðT  DÞ=B < 0:4 ð7aÞ stratum. Based on these results, a simplified method is proposed
for assessing the basal heave factor of safety for axisymmetric sup-
lT ¼ 1:0 forðT  DÞ=B P 0:4 ð7bÞ ported excavations in clay.

Table 3
Comparison of FS for cu linearly increasing with depth.

c (kN/m3) B (m) T (m) H (m) D (m) c0 (kPa) m Ave cu (kPa) FSFE FSshaft
16 40 60 16 4 5 1.5 39.5 1.288 1.287
16 40 60 16 4 10 1.5 44.5 1.473 1.450
16 40 60 16 4 20 1.5 54.5 1.813 1.776
16 40 60 16 10 10 1.5 53.5 1.923 1.952
16 40 72 24 4 10 1.5 56.5 1.285 1.291
16 40 80 24 12 10 1.5 68.5 1.748 1.785
16 100 120 24 24 10 1.5 93.25 2.192 2.180
16 20 60 16 10 10 1.5 51.25 2.177 2.300
16 20 60 16 4 10 1.5 42.25 1.544 1.610
16 30 60 16 4 10 1.5 43.375 1.479 1.498
16 40 60 16 4 20 1.2 47.6 1.584 1.551
16 40 60 16 4 10 1.2 37.6 1.246 1.225
16 40 60 16 10 10 1.2 44.8 1.617 1.634
 A.T.C. Goh / Tunnelling and Underground Space Technology 61 (2017) 145–149 149

References Century of Rock Mechanics 11th Congress of the International Society for Rock
Mechanics. Taylor & Francis, pp. 1137–1140.
Khatri, V.N., Kumar, J., 2010. Stability of an unsupported vertical circular excavation
Arai, Y., Kusakabe, O., Osamu Murata, O., Konishi, S., 2007. A numerical study on
in clays under undrained condition. Comput. Geotech. 37 (3), 419–424.
ground displacement and stress during and after the installation of
Kim, K.Y., Lee, D.S., Cho, J., Jeong, S.S., Lee, S., 2013. The effect of arching pressure on
deep circular diaphragm walls and soil excavation. Comp. Geotech. 35 (5),
a vertical circular shaft. Tunnell. Undergr. Space Technol. 37, 10–21.
791–807.
Matsui, T., San, K.C., 1992. Finite element slope stability analysis by shear strength
Bjerrum, L., Eide, O., 1956. Stability of strutted excavations in clay. Géotechnique 6
reduction technique. Soils Found. 32 (1), 59–70.
(1), 32–47.
McNamara, A.M., Roberts, T.O.L., Morrison, P.R.J., Holmes, G., 2008. Construction of a
Brinkgreve, R.B.J., Bakker, H.L., 1991. Non-linear finite element analysis of safety
deep shaft for Crossrail. Proc. Instn. Civ. Eng. – Geotech. Eng. 161 (6), 299–309.
factors. In: Proc. 7th Int. Conf. Comp. Methods and Advances in Geomech. A.A.
Parashar, S., Mitchell, R., Hee, M.W., Sanmuganathan, D., Nicholson, G., 2007.
Balkema, pp. 1117–1122.
Performance monitoring of deep shafts at Changi WRP project, Singapore. In:
Brinkgreve, R.B.J., Swolfs, W.M., Engin, E., 2011. PLAXIS 2D User Manual. Plaxis BV,
7th Int. Symp. on Field Measurements in Geomechanics (FMGM 2007), Geotech.
Delft, The Netherlands.
Spec. Publ. 175 (GSP175), ASCE, Reston, VA, pp. 1–12.
Britto, A., Kusakabe, O., 1983. Stability of axisymmetric excavations in clays. J.
Salençon, J., Matar, M., 1982. Capacité portante des fondations circulaires. Journal
Geotech. Eng. ASCE 109 (5), 666–681.
de Mécanique Théorique et Appliquée 1 (2), 237–267.
Cai, F., Ugai, K., Hagiwara, T., 2002. Base stability of circular excavations in soft clay.
Salgado, R., Lyamin, A.V., Sloan, S.W., Yu, H.S., 2004. Two- and three-dimensional
J. Geotech. Eng. ASCE 128 (8), 702–706.
bearing capacity of foundations in clay. Géotechnique 54 (5), 297–306.
Do, T., Ou, C.Y., Lim, A., 2013. Evaluation of factors of safety against basal heave for
Schwamb, T., Soga, K., Mair, R.J., Elshafie, M.Z.E.B., Sutherden, R., Boquet, C.,
deep excavations in soft clay using the finite element method. J. Geotech.
Greenwood, J., 2014. Fibre optic monitoring of a deep circular excavation. Proc.
Geoenviron. Eng. ASCE 139 (12), 2125–2135.
Instn. Civ. Engrs. – Geotech. Eng. 167 (2), 144–154.
Faheem, H., Cai, F., Ugai, K., Hagiwara, T., 2003. Two-dimensional base stability of
Tan, Y., Wang, D., 2015. Structural behaviours of large underground earth-retaining
excavations in soft soils using FEM. Comput. Geotech. 30 (2), 141–163.
systems in Shanghia. I: Unpropped circular diaphragm wall. J. Perform. Constr.
Faheem, H., Cai, F., Ugai, K., 2004. Three-dimensional base stability of rectangular
Facil. ASCE 29 (2), 04014059.
excavations in soft soils using FEM. Comput. Geotech. 31 (2), 67–74.
Wong, K.S., Goh, A.T.C., 2002. Basal heave stability for wide excavations. In: Proc.
Goh, A.T.C., 1994. Estimating basal heave stability for braced excavations in soft
3rd Int. Symp. Geotech. Aspects of Underground Construction in Soft Ground,
clay. J. Geotech. Eng. ASCE 120 (8), 1430–1436.
Toulouse, pp. 699–704.
Griffiths, D.V., Lane, P.A., 1999. Slope stability analysis by finite elements.
Zhang, W.G., Goh, A.T.C., 2012. Reliability assessment on ultimate and serviceability
Géotechnique 49 (3), 387–403.
limit states and determination of critical factor of safety for underground rock
Hammah, R.E., Yacoub, T., Curran, J.H., 2007. Serviceability-based slope factor of
caverns. Tunnell. Undergr. Space Technol. 32, 221–230.
safety using the shear strength reduction (SSR) method. In: The Second Half