Geotech Geol Eng (2018) 36:3683–3705

https://doi.org/10.1007/s10706-018-0565-0 (0123456789().,-volV)
(0123456789().,-volV)

ORIGINAL PAPER

Design Aspects of Under-Passing a City’s Heritage

Landmark with EPB Machines Under Low Overburden;
The Case of Chandpole Gate in Jaipur Metro, India
Spiros Massinas . George K. Prountzopoulos . Vijyant Bhardwaj .
Ashwini Saxena . Jim Clark . Michael G. Sakellariou

Received: 9 November 2017 / Accepted: 3 May 2018 / Published online: 10 May 2018
 Springer International Publishing AG, part of Springer Nature 2018

Abstract UG1B is the first underground Metro Line in Jaipur by two EPB TBM machines, and introduces
under construction in Jaipur which is the capital of the new method which is used for designing these
Rajastan (India). The twin bore tunnel route pass shallow tunnels, by combining the usage of the
through the historical centre of the old city of Jaipur analytical solution of Massinas and Sakellariou
called ‘‘Pink City’’. Significant historical monuments (Géotechnique 59:691–701, 2009) with detailed
of world’s heritage, such as Hawa Mahal, Jantar numerical analysis; the innovative closed form solu-
Mantar, Isarlat tower, Tripolia Gate (main entrance to tion has been applied to assess the required support
the city palace) and Chandpole Gate, are within the pressure, whereas its results have been used as input to
greater influence zone of the project. Specifically, the 3-D numerical analyses of tunnel excavation below
twin bore TBM tunnel passed 4.5 m under the the Gate to predict induced settlements. The route
foundation of Chandpole Gate which is one of the from the evaluation of the geotechnical model, to the
main significant historical entrances to the old city, study of the structural and foundation system of the
almost 300 years old. The paper addresses the main monument, the approach to the necessary tunnel face
design aspects of the under-passing of Chandpole Gate retaining pressure to be maintained by the EPB

S. Massinas (&) A. Saxena

ALYSJ JV (Aktor - Larsen & Toubro - Yapi Merkezi - Jaipur Metro Rail Corporation Ltd, Khanij Bhawan, Tilak
STFA - Al Jaber Engineering JV), Doha, Qatar Marg, C-Scheme, Jaipur, India
e-mail: [email protected] e-mail: [email protected]

S. Massinas J. Clark
Omikron kappa - Indus Consultrans JV, Spaze i-Tech The Robbins Company, Hyderabad, India
Park, Tower A, Sector 49, Sohna Road, Gurgaon, e-mail: [email protected]
Haryana 508-509, India
M. G. Sakellariou
G. K. Prountzopoulos Laboratory of Structural Mechanics and Engineering
Omikron Kappa Consulting SA, 4 Thaleias Str. & 109 Structures, School of Rural and Surveying Engineering,
Kimis Ave., Maroussi, Athens, Greece National Technical University of Athens (NTUA),
e-mail: [email protected] Athens, Greece
e-mail: [email protected]
V. Bhardwaj
Indus Consultans Pvt. Ltd, Spaze i-Tech Park, Tower A,
Sector 49, Sohna Road, Gurgaon, Haryana 508-509, India
e-mail: [email protected]

123
3684 Geotech Geol Eng (2018) 36:3683–3705

machines during under passing and the prediction of and Corté 1994) and analytical approaches (Mu-
the surface and Gate settlement due to tunneling rayama et al. 1966; Leca and Dormieux 1990) of the
operations is introduced and analyzed. Monitoring of required face pressure to maintain tunnel face stability
TBM operation parameters and settlement measure- and minimize plasticization ahead of and around the
ments during construction are presented in the form of tunnel have been presented early at the beginning of
comparative diagrams and graphs, to illustrate the the worldwide need for metro systems due to intense
application of the main design recommendations and urbanization, while lately an innovative analytical
verify the design predictions, in terms of applied face solution for shallow tunneling problems in cohesive-
pressure and resulting settlements of the Chandpole frictional soil is introduced by Massinas and Sakel-
Gate. Comparison has shown that main recommenda- lariou (2009). A widely used methodology has been
tions, as derived from the new proposed method of presented by Anagnostou and Kovári (1994, 1996),
design and followed during the tunnelling operations, based on the silo theory introduced by Horn (1961)
resulted to a successful under-passing of Chandpole and concerns both, the cases of EPB and Slurry
Gate, with induced settlements within the predicted, machines, whereas it foresees the effect of the
permissible limits. groundwater table.
Numerical approaches to the problem of face
Keywords Shallow tunnelling  Analytic solutions  stability of shallow urban tunnels and tunnelling
Numerical analysis  Settlement prediction  Earth induced settlements in urban environments have been
pressure balance shield  Monument under-passing  extensively presented through 3D analyses of contin-
Support pressure design  Monitoring uous and discontinuous media, taking advantage of the
development of personal computers. Some of these
research studies focus more on the stability of the
tunnel face (Vermeer et al. 2002; Mollon et al. 2009;
1 Introduction Kim and Tonon 2010; Chen et al. 2011; Prount-
zopoulos 2012), whereas others address the tunnelling
Soft ground tunnelling in dense urban environments, induced settlements and aspects of tunnel–soil-surface
under remarkably strict limitations in terms of ground structures interaction (Potts and Addenbrooke 1997;
surface disturbance due to tunnelling activities and Liu 1997; Burd et al. 2000; Mirhabibi and Soroush
under very stringent time schedules for completion, 2012). Comparison between analytical, numerical and
have rendered Tunnel Boring Machines as the rule for experimental tools has also been used (Migliazza et al.
urban tunnel construction. Increased construction 2009; Giardina et al. 2015) to verify analytical and
speed and the ability to control ground deformation numerical methodologies that are simpler to apply,
both at the excavation face and at the tunnel cavity, are especially for the execution of parametric evaluation.
the most important amongst the advantages of TBMs Inevitably, protection of heritage buildings (mainly
when compared to conventional excavation methods. masonry) and monuments during construction of
During TBM tunnelling in soft ground, control of Metro systems in historic cities, as well as critical
surface settlement is mainly dependent on the control and difficult underpasses (Massinas et al. 2015)
of the face stability and extrusion, most commonly becomes a major task for both the tunnelling industry
through Earth Pressure Balance shield machines, and and the research community (Burghignoli et al. 2010).
the timely and effective application of grouting of the Recently, Rampello et al. (2012) used 3-D numerical
annular gap between the tunnel profile and the modeling, applying different approaches of soil–
segmental lining. Therefore, numerous research stud- tunnel-structure interaction, in order to evaluate the
ies have been focused on the estimation of surface effect of EPB tunnelling along Contract T2 of the new
settlements caused by shallow tunnels in urban line C of Rome Metro, which underpasses several
environment, starting from analytical solutions, such masonry buildings in the historical centre of the city.
as the ones presented by Atkinson and Potts (1977) for In the current study the combination of the analytical
granular soils and Davis et al. (1980) for purely solution with 3-D numerical modelling is proposed as
cohesive soil. Furthermore, experimental (Broms and a method, for solving shallow tunnelling problems.
Bennenmark 1967; Kimura and Mair 1981; Chambon

123
Geotech Geol Eng (2018) 36:3683–3705 3685

2 Project Description and an overburden increasing from 7 to 14 m towards

the end of the alignment.
2.1 General
2.3 The Chandpole Gate
Jaipur Metro Rail Project has been planned to stretch
along two corridors, the East–West Corridor from Jaipur is widely known as the ‘‘Pink City’’. This is due
Mansarovar to Badi Chaupar (Phase 1) and the North– to the Walls surrounding the city’s historic centre and
South Corridor from Ambabari to Sitapura (Phase 2). the majority of the buildings within these walls, which
Phase 1 is divided in two sub-phases, Phase 1A which are uniformly painted in terracotta pink colour since
is completed and under operation and Phase 1B, 1876. Pink City is the most important attraction in
currently under construction. Phase 1A (Mansarovar Jaipur with huge cultural, heritage and touristic
to Chandpole) has a total length of 9.72 km and 9 significance. Entrance to the historic centre is possible
stations and is an elevated line along the major part of through 8 gates around the city, with Chandpole Gate
its length (9.28 km and 8 stations). The last 0.44 km of being the one situated at Chandpol Bazaar area.
Phase 1A is underground, ending up to Chandpole Chandpole Gate is an impressive stone masonry
underground station. Phase 1B, underground along its structure, with approximate height of 10 m, con-
total length, starts from Chandpole Station and structed in 1727. The main load bearing elements are
includes two underground stations, Choti Chaupar made of stone and constitute the sides of the main arch
and Badi Chaupar, with the last being the end station (Figs. 3, 4).
of the Corridor. Underground part of the line involves
twin bore, single track tunnels, constructed with two
Earth Pressure Balance (EPB) Shield TBMs, launched 3 Geotechnical Model and Investigation of Gate
from Chandpole Shaft, under-passing the Chandpole Foundation
Gate heritage structure of the Jaipur Pink City’s walls
after approx. 100 m drive, towards Choti Chaupar and 3.1 Geotechnical Model
Badi Chaupar. Figure 1 illustrates the layout of Jaipur
Metro, while in Fig. 2 a more detail view of the Soil Alluvium is the main formation encountered
alignment at Chandpole gate is given. along the alignment of the project, in the form of sand
to silty sand with gravels. A total number of 29
2.2 Alignment and Tunnel Characteristics exploratory boreholes have been executed along the
alignment of Phase 1B, up to depths of 1–2 tunnel
The tunnel alignment of Phase 1B involves twin bore diameters below the invert level, revealing a generally
tunnels of approx. length 2.3 km per bore, excavated homogeneous geotechnical model, comprising of the
by two EPB TBMs, launched from Chandpole Shaft. following distinct entities:
Excavation diameter is 6.51 m, whereas the finished
• Made Ground/Fill Materials These are artificial
tunnel diameter, after placement of the segmental
deposits, consisting of coarse grained soil materi-
lining is 5.8 m. Segmental lining rings have a length of
als, with varying thickness and distribution
1.2 m and thickness of 0.275 m. At launching posi-
depending on the previous and current land uses.
tion, center to center distance of tunnels is 12 m and
The thickness of the formation is usually around
the tunnel overburden is 5.5 m. For the initial drive of
1.5 m and only locally exceeds 2 m.
approximately 100 m from launching, TBMs are
• Soil Alluvium The formation has been found in all
driven along a horizontal curve of approximately
boreholes, whereas its thickness along the align-
400 m radius, to arrive at the Chandpole Gate with a
ment is estimated to be larger than 30 m, since all
center to center distance of 13.6–13.9 m and an
boreholes encountered only this formation up to
overburden of 6.6–6.8 m. After under-passing the
their termination depth (* 24 m). Alluvium for-
Gate, tunnels are aligned below Chandpol Bazar Road,
mation is generally fine grained soil material in the
a main shopping avenue of the city of Jaipur, with a
form of silty sand, with interlayers and varying
center-to-center distance of 14.2 m kept constant
small percentages of gravels and clay. A distinct
along the remaining straight part of the tunnel stretch

123
3686 Geotech Geol Eng (2018) 36:3683–3705

Fig. 1 Indicative illustration of Jaipur Metro. (Source www.mapsofindia.com)

Fig. 2 Horizontal alignment of up and down tunnel lines at Chandpole Gate Area

123
Geotech Geol Eng (2018) 36:3683–3705 3687

Fig. 3 Typical view of Chandpole Gate (towards Chotti Chaupar)

Fig. 4 Typical view of 1st floor passage inside Chandpole Gate (left) and staircase going to terrace (right)

feature within the formation is the appearance of characteristics of the silty sand layer, since their
Kankar layers. Such layers are generally coarse compacted and hard nature increases the overall
sand to medium grained and occasionally coarse geotechnical characteristics of the alluvium
gravel size calcareous nodules. Such layers have soil.
been found in almost all boreholes and they exhibit
In general, ground water has not been encountered
a major impact on the geomechanical
in the boreholes along the alignment, or, it has been

123
3688 Geotech Geol Eng (2018) 36:3683–3705

found to lie at increased depths below ground surface. through grouting below and around the Gate’s foun-
Therefore no groundwater was expected to be encoun- dation. The grain size distribution of the in situ soil
tered during the excavation works. practically excluded the successful and efficient
Evaluation of in situ and laboratory testing on the application of low-pressure cement grouting, even
silty sand alluvium formation, resulted in the strength, with use of micro-fine cement. Other grouting meth-
deformability and in situ stress parameters for the ods that would be more appropriate for this soil type,
design of the under-passing that are presented in e.g. jet grouting and compensation grouting were
Table 1. excluded, as the risk of soil disturbance, instability,
The coefficient of earth pressure at rest Ko is temporary liquefaction and settlement below the gate
determined equal to 0.6, based on OCR results from foundation was considered too high for the signifi-
the consolidation tests performed in samples from the cance and vulnerability of the structure. Additionally,
boreholes giving a value of 1.1. the loading of the structure itself and the long period of
its application was naturally assumed to have compact
3.2 Chandpole Gate Foundation Survey the soil below the Gate. This assumption could be
partly verified by the soil inspection in the trial pits. In
For the determination of the structure’s foundation the light of the above considerations, it was decided by
system, dimensions and condition, special survey was the designer and the contractor to under-pass the
carried out, consisting of nine (9) trial pits that were structure without any pre-treatment of the soil, relying
executed in certain locations near the gate, as on the appropriate TBM operation, mainly in terms of
illustrated in Fig. 5. Representative images from the applied face pressure and annular gap grouting
execution of the trial pits are given in Fig. 6. application.
According to the findings from the trial pits on both
sides of the structure, the Gate rests on stone masonry
foundation with depth of approx. 2.0–2.4 m, whereas 4 Design of Underpassing Scheme
below the main arch, a boulder layer was found with
depth * 1.5 m near the gate’s walls and * 1.0 m 4.1 EPB Support Pressure
below the mid part. Water pipes and other utilities
were found below the two passageways and the main Jeffery (1920) published a solution of the displace-
arch. A typical section of the Gate with its foundations ments at the surface caused by the stresses applied at
and the under-passing TBM tunnels is given in Fig. 7, the boundary of a shallow tunnel. Strack (2002) gave a
indicating the soil cover of approx. 4.5 m between the relationship between the converges of tunnel boundary
bottom of the Gate’s foundation and the tunnels’ with the surface settlements in the case of an elastic
crown. half space. The assumption that the induced surface
Examination and co-evaluation of the geotechnical settlements are related to the tunnel deformation is
investigation and Gate foundation survey results was realistic. Whereas there is a closed form solution of the
jointly performed by the designer and the contractor, plastic zone formation in a half space obeying Mohr–
aiming to decide on the necessity of soil strengthening Coulomb criterion (Massinas and Sakellariou 2009),

Table 1 Design Design parameter Adopted value

geotechnical parameters
considered for analytical Characteristic N60 value 3.2z ? 5, z = depth below ground surface (m)
computations and numerical
Modulus of Elasticity E (MPa) 1.0 9 N60
analyses of Chand Pole
Gate under-passing Poisson’s ratio (v) 0.4
0
Effective friction angle u () 27, for z = 0–10 m
30, for z [ 10 m
Effective cohesion, c0 (kPa) 5
Unit weight, c (kN/m3) 19
Coefficient of earth pressure at rest, K0 0.57

123
Geotech Geol Eng (2018) 36:3683–3705 3689

Fig. 5 Location of the trial pits (grey rectangle shapes), executed at Chand Pole Gate foundation (striped) area

there is not available so far a solution relating the More specifically, the specific closed form solution
induced surface settlements to the extent of the plastic calculates the plastic zone extent around the excava-
zone. So, a combined analytical-computational tion and consequently the extension of the plastic ring
method is proposed. At first, the critical internal into the ground around a circular tunnel in half-space
pressure to onset the plastic zone is calculated based obeying the Mohr–Coulomb failure criterion by
on the Massinas and Sakellariou solution (2009). considering the tunnel-surface interaction, as indi-
Then, based on the parametric solution for different cated in Figs. 8 and 9. The assumption of hydrostatic
values of internal pressure, with starting value the (or isotropic) in situ stress field has been made in order
critical one, a graph is obtained. The analytically to obtain a closed form solution. This assumption
obtained values are used as input to a parametric holds, also, for respective closed form solutions for
numerical analysis for a complete tunnel-surface deep tunnels, assuming an infinite space. In the case of
interaction to estimate the induced surface settle- deep tunnels, for k = 1 there are solutions only for
ments. Finally, the required value of internal pressure Tresca (Savin 1961) and Mohr–Coulomb (Detournay
to constrain the settlements to a tolerable magnitude is and St. John 1988) criteria, so far. In the shallow
obtained. tunnels case, the plastic zone is not a concentric circle

123
3690 Geotech Geol Eng (2018) 36:3683–3705

Trial Pit No.5

Trial Pit No.9

Fig. 6 Trial pit No. 5 (Left-Stone masonry foundation with lime mortar with depth * 2.4 m) and trial pit No. 9 (Right-Boulders’ layer
below the main arch with thickness * 1.0 m)

to the tunnel opening, as in the deep tunnels case (for depth of the center of the plastic zone from the surface,
k = 1), because of the influence of ground surface. j is the distance from the origin to a pole, in bipolar
The critical internal support pressure Pcr, where the coordinates, k = (1 ? sinu)/(1 - sinu),
initial yielding occurs at the tunnel periphery and the Y = 2ccosu/(1 - sinu), c, u are the soil cohesion
plastic zone formation around the cavity are given by and friction angle respectively, and a, b the bipolar
the following equations: coordinates.
The presence of the existing gate was considered by
2j2
Pcr ¼ assuming its foundation level as an upper boundary of
ðdi2 
2  ri2 cos2 bÞ þ j2 ðk 1Þ the semi-infinite space. Thus a total overburden of
d 2  ri2 cos2 b Y
 P0 i  ð1Þ 4.5 m has been considered to examine the interaction
j2 2 between the Gate and the tunnel, as presented in
 1k Fig. 10. Furthermore, the total earth pressure (due to
rc di  ri cos b ½2M0 þ j2 ðk  1Þ  ½ þ Pi ðk  1Þ
 ¼ gravity) at the real depth of the tunnel is also taken into
ri dc  rc cos b 2M0 ½ þ P0 ðk  1Þ
account, and, a total uniform load (P0) at the upper
ð2Þ boundary of the half-space is applied. The comparison
where ri is the tunnel radius, rc is the plastic zone between the physical model of the problem and its
radius for specific direction, di is the depth of the equivalent model for the analytical calculations is
center of the circular tunnel from the surface, dc is the illustrated in Fig. 10 (Massinas and Sakellariou 2010).

123
Geotech Geol Eng (2018) 36:3683–3705 3691

Fig. 7 Typical section of Chand Pole Gate with its foundations, as per the results from the executed trial pits and the under-passing
tunnels at depth of approx. 4.5 m

By using Eqs. (1) and (2) for the plastic zone pressure: at the tunnel crown, 45 from crown, at the
calculation and the soil’s properties according to sidewalls and at the invert.
Table 1, the graph illustrated in Fig. 11 was created As it is evident from the graph, a minimum plastic
showing the plastic zone width at four certain points zone width (\ 0.5 m) is calculated for a mean support
around the circular tunnel for different values of pressure within the range of 1.4–1.5 bar, while for

123
3692 Geotech Geol Eng (2018) 36:3683–3705

Fig. 8 Bipolar coordinate system for plastic zone calculation—shallow tunnel problem. Reproduced with permission from (Massinas
and Sakellariou 2009)

Fig. 9 Principle of bipolar

coordinate system.
Reproduced with
permission from (Massinas
and Sakellariou 2009)

1.6 bar the soil around the excavation remains within the powerful software FLAC3D, in order to investi-
the elastic domain. The value of 1.5 bar was selected gate the interaction of the tunnels with the Chandpole
for the execution of 3-D numerical analyses for Gate and specifically to:
settlement prediction.
• Evaluate the internal support pressure values
resulting from the application of the methodology
4.2 Settlement Analysis
of Massinas and Sakellariou (2009, 2010) and
propose the range of the support pressure to be
Taking into account all the above information for the
adopted during the operation of the TBMs.
Chandpole Gate and the face pressure design consid-
erations, special 3-D analysis was performed by using

123
Geotech Geol Eng (2018) 36:3683–3705 3693

Fig. 10 Physical and

equivalent model for
analytical calculations

Plastic Zone Width (m) - Under Chandpole Gate 4.2.1 Presentation of 3-D Models
6 5,5 5 4,5 4 3,5 3 2,5 2 1,5 1 0,5 0
0,8 Assuming stress induced failure mode of ground mass,
0,9 the computational three dimensional analysis is per-
formed with a continuum model approach, using the
1
Finite Difference Software FLAC3D. In order to
Pressure (bar)

1,1 ensure the accuracy of the results and to avoid any

1,2 ‘‘boundary effects’’, certain parametric 3-D analyses
Distance from tunnel axis (m)
15 10 5 0 were performed (Benchmark tests, see also Massinas
1,3
0 and Sakellariou 2006) with different 3-D meshes in
Tunnel 1,4
dimension and density. The conclusions derived from
Plastic Zone for Pi=1.1 bar
1,5 these Benchmark analyses led to large model with
Depth (m)

5
1,6
approximately 70,000 octahedral elements and more
than 210,000 unknowns which consequently increased
1,7
10
significantly the computational time. Typical views of
1,8 the final 3-D mesh are presented in Figs. 12 and 13.
1,9 Regarding boundary conditions, typical constraints
15 were used. The top surface was left free to displace, the
2
side surfaces had roller boundaries (horizontal con-
Curves calculated for c=5kPa, φ=27o
straint) and the bottom face of the 3-D Model was
Crown Sidewall Invert 45 deg. from crown
fixed. Figure 13 above illustrates the typical view of
Fig. 11 Analytical calculation of plastic zone width below the the EPB shield for both bores. The Gate foundation
Gate was simulated as homogeneous elastic material, with
properties typical for a medium strength rock. A check
of the stress condition of the foundation was per-
• Calculate the induced surface settlements and formed when processing the analysis results, regard-
angular distortions below, in the vicinity of the ing the possibility of tension development in the
Gate structure. foundation, which would render this simulation
• Verify the assumption for the necessity of appli- approach problematic. Load of superstructure was
cation of additional protection/strengthening calculated and applied on the foundations as a
measures. distributed load.
The exact geometry of the segments and the annular
gap for the tail shield grouting were simulated as
presented in Fig. 14.

123
3694 Geotech Geol Eng (2018) 36:3683–3705

Chandpole Gate
foundation

Down line
tunnel

Up line
tunnel

Fig. 12 Typical view of 3D model (Tunnels and Gate foundation)

EPB shield
area for
Down line

EPB shield
area for Up
line

EPB shield
area for 2nd
bore

2nd bore

EPB shield
area for 1st
bore
1st bore

Fig. 13 Typical view of 3D model (EPB shield area)

123
Geotech Geol Eng (2018) 36:3683–3705 3695

The tunnels’ construction was modeled with 64 lining. Typical illustration of the extent of the plastic
simulation steps, 32 for each bore. The excavation step zone around the tunnels is given in Fig. 15.
(each load case) was considered equal to 1.2 m which Calculated surface settlement troughs in 3-D are
is the length of each segmental lining ring. In each shown in Figs. 16 and 17, where vertical displace-
excavation step, a support pressure of 1.5 bar was ments after excavation of the Up Line tunnel and after
considered and the erection of the ‘‘new’’ ring and the excavation of both the Up Line and the Down Line
grouting were simulated. tunnels are illustrated. The maximum calculated
surface settlements are approximately 6 mm while at
4.2.1.1 3-D Analysis Results The critical parameter the foundation area of Chandpole Gate the maximum
that affects the development, magnitude and extent of settlements are calculated approximately as 5 mm.
displacements is the width of the plastic zone that is During the simulation of excavation of the first (Up
formed ahead of the tunnel face and consequently Line) tunnel, the maximum calculated deflection of
around the cavity, along the excavated tunnel. The the gate’s foundation (above the tunnel) was 1/1300,
greater the extent of the plastic zone width ahead of while the maximum calculated differential settlement
and above the face area the greater the between the two foundations of the Gate was 5 mm
‘‘plasticization’’ of the soil around the cavity. with an angular distortion of 1/1200. After simulation
Analyses results showed that the applied support of the second TBM under-passing, below the respec-
pressure of 1.5 bar, produces minimum disturbance of tive foundation of the Gate, the maximum calculated
the surrounding soil with a plastic zone width of less settlement was 4 mm. Thus, at the final stage of the
than 0.5 m only in the invert area and the soil around tunnel’s construction simulation, the maximum cal-
the excavation remaining within the elastic domain. culated differential settlement between the two foun-
Conducted analyses have of course simulated a proper, dations of the Gate was practically zero. Calculated
timely and continuous grouting process of the annular settlements of Chandpole Gate’s foundations, after
gap between the excavation profile and the segmental

Grout for
Down line

Segmental Segmental
Grout for Up lining for Up lining for
line line Down line

Fig. 14 Typical 3D view of segmental lining and grout for backfilling the annular gap

123
3696 Geotech Geol Eng (2018) 36:3683–3705

Fig. 15 Calculated yield

FLAC3D 2.10
zones along the Up Line Step 51906 Model Perspective
tunnel 20:46:21 Fri Mar 07 2014

Center: Rotation:
X: 2.250e+001 X: 30.000
Y: 4.368e+001 Y: 0.000
Z: -5.236e+000 Z: 60.000
Dist: 4.500e+002 Mag.: 2.05
Ang.: 22.500

None
shear-p

Itasca Consulting Group, Inc.

Minneapolis, MN USA

Fig. 16 Calculated surface

settlements after total
excavation of the Up Line
tunnel (values in m)

completion of under-passing by both tunnels, are always under compression. An indicative relevant
illustrated in Fig. 18. result after completion of under-passing by both
Calculation results showed that no tensile stresses TBMs, is given in Fig. 19.
are developed in the Gate’s pad foundations, which are

123
Geotech Geol Eng (2018) 36:3683–3705 3697

Fig. 17 Calculated surface

settlements after total
excavation of both Up and
Down Line tunnels (values
in m)

FLAC3D 2.10
Step 51906 Model Perspective
17:26:29 Sat Mar 08 2014

Center: Rotation:
X: 1.963e+001 X: 40.000
Y: 3.618e+001 Y: 0.000
Z: -1.562e+001 Z: 40.000
Dist: 4.50 0e+002 Ma g.: 3.9 1
Ang.: 22.500

Sketch
Magfac = 0.000e+000
Null zones only
Linestyle
Contour of Z-Displacement
Magfac = 0.000e+000
-5.1562e-003 to -5.0000e-003
-5.0000e-003 to -4.5000e-003
-4.5000e-003 to -4.0000e-003
-4.0000e-003 to -3.5000e-003
-3.5000e-003 to -3.0000e-003
-3.0000e-003 to -2.5000e-003
-2.5000e-003 to -2.0000e-003
-2.0000e-003 to -1.6881e-003
Interval = 5.0e-004

Itasca Consulting Group, Inc.

Minneapolis, MN USA

Fig. 18 Calculated settlements of Chandpole Gate’s foundations after simulation of underpassing by the second TBM (values in m)

123
3698 Geotech Geol Eng (2018) 36:3683–3705

FLAC3D 2.10 FLAC3D 2.10

Step 25045 Model Perspective Step 51906 Model Perspective
17:07:36 Sat Mar 08 2014 17:27:45 Sat Mar 08 2014

Center: Rotation: Center: Rotation:

X: 1.752e+001 X: 40.000 X: 1.963e+001 X: 40.000
Y: 4.003e+001 Y: 0.000 Y: 3.618e+001 Y: 0.000
Z: -1.371e+001 Z: 40.000 Z: -1.562e+001 Z: 40.000
Dist: 4.500e+002 Mag.: 3.91 Dist: 4.500e+002 Mag.: 3.91
Ang.: 22.500 Ang.: 22.500

Sketch Sketch
Magfac = 0.000e+000 Magfac = 0.000e+000
Null zones only Null zones only
Linestyle Linestyle
Principal Stresses Principal Stresses
Magfac = 0.000e+000 Magfac = 0.000e+000
Compression Compression
Linestyle Linestyle
Max Compression = -1.069e+005 Max Compression = -1.125e+005
Max Tension = 0.000e+000 Max Tension = 0.000e+000

Itasca Consulting Group, Inc. Itasca Consulting Group, Inc.

Minneapolis,MNUSA Minneapolis, MN USA

Fig. 19 Calculated principal stress tensors on gate’s foundation

5 Comparison of Design Predictions procedure, but also ensure that the modified stress
with Construction Data regime due to tunnel excavation would not affect
the foundation area of the Gate during stoppage,
The TBM drive of approximately 120 m from the e.g. in case of partial loss of the applied face
launching shaft to the completion of Gate under- pressure. This position was selected 10 m before
passing has been divided into 3 parts: the TBM crossing the projection of the Gate’s
foundation and the concept is presented in Fig. 20.
1. Initial Drive (0–60 m or first 50 rings). Along this
3. Gate Under-passing (94–120 m of rings 79–100).
drive, focus has been given on launching, check-
Slow, stable and continuous boring below the
ing and resolving main operation issues of the
Gate, with application of face pressure as per the
TBM, coordinating the construction, monitoring
tunnel design and the evaluation of the calibration
and design teams and first assessment of ground
drive and performance of grouting works with
response to tunnelling. Shield friction and grout-
high diligence and using experience from the
ing effectiveness, causing heave on the surface
calibration drive. Exact footprint of the Gate was
above the machine and increased settlement
above rings 87–93.
behind it respectively could be recognized during
this process.
Presentation of construction and monitoring data are
2. Calibration Drive (60–94 m or ring 51–78). This
focused on the comparison with the design consider-
drive was used to fine-tune TBM operation,
ations and results as regards face pressure and
resolve/limit shield friction, improve grouting
settlements of the ground and the Gate.
performance and check design assumptions and
revise them, if necessary. Evaluation of data from
5.1 Face Pressures
TBM operation and surface monitoring had to be
constant and analytical and therefore the machine
Figure 21 presents the distribution of applied face
had to slow down or stop frequently for this
pressures (recorded face pressures at middle sensors)
purpose. The end of this drive was the final
for the last part of the calibration drive and the under-
possible stoppage of the TBM for any repair and
passing of the Gate by TBM1 (Up Line) and TBM2
maintenance works, before re-launching for the
(Down Line), in comparison to the recommended face
under-passing of the Gate, along which the
pressure range as per the combined analytical-com-
machine had to proceed with slow and constant
putational method (design). The following are
speed. Therefore, selection of the end position for
observed:
this drive had to be decided carefully, so as to
maximize the length of the valuable calibration

123
Geotech Geol Eng (2018) 36:3683–3705 3699

Fig. 20 Concept for selection of safe distance before the Gate for TBM final stoppage before under-passing, for maintenance purposes

2.0 2.0
Pressure at Middle Sensor (bar)

1.8 1.8
1.6 1.6
1.4 1.4
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
TBM1 TBM2
0.2 0.2
(Up Line) (Down Line)
0.0 0.0
75 80 85 90 95 100 105 110 75 80 85 90 95 100 105 110
Ring No Ring No

Fig. 21 Variation of average face pressure of middle sensors per ring excavation versus the range recommended by the design for
TBM1 (Up Line—Left) and TBM2 (Down Line—Right)

• The applied face pressure values are generally 2.0

Pressure at Middle Sensor (bar)

within the recommended pressure range as per the 1.8

design. During TBM1 advance, face pressures 1.6

1.4
generally lie towards the lower bound of the design
1.2
range, whereas for TBM2, recorded face pressure
1.0
values lie towards the upper bound of the design
0.8
recommendations. Mean face pressure was main-
0.6
tained within the reasonably expected variations 0.4
around the design values, especially during under- 0.2 TBM1 - Average Pressure TBM2 - Average Pressure Range as per Design
passing of the Chandpole Gate, as it is more clearly 0.0
75 80 85 90 95 100 105 110
presented in Fig. 22. Ring No
• Required increase of face pressure to balance the
increased stresses below the Gate due to the weight Fig. 22 Variation of average face pressure per ring excavation
of the structure has been gradually applied in both versus the range recommended by the design for TBM1 (Up
Line) and TBM2 (Down Line)
drives.

123
3700 Geotech Geol Eng (2018) 36:3683–3705

• Recorded pressure on the left sensor of TBM1 was Figure 24 presents similar results (face pressure vs.
systematically higher than the corresponding stroke) for more TBM1 ring-long excavations below
pressure of the right sensor. Specifically, left the Chandpole Gate. Maintenance of pressure, prac-
sensor pressure readings generally lied on the tically constantly, at the design levels from the
average pressure of the range recommended by the commencement of ring excavation is obvious.
design, whereas the readings of the right sensor
lied somehow below the lower bound of the 5.2 Surface Settlements
suggested range.
A comprehensive online monitoring system was
Variation of face pressure during the last ring-long
installed during the Gate under-passing process.
excavation (ring 78) before temporary stoppage of
Dense surface settlement markers as well as move-
TBM1 at the end of the calibration drive is illustrated
ment markers installed on specific positions on the
in Fig. 23 (left). Frequent stoppage of the machine
Gate structure were continuously monitored and the
caused a loss of face pressure due to sitting of the muck
monitoring results were available and constantly
in the chamber during standstill and therefore pressure
evaluated by the key involved persons of the Con-
had to be increased during the 100–150 mm of stroke.
tractor, the Designer and the Client. Figure 25 illus-
Applied pressure was generally within or closely
trates the measured surface settlements above specific
around the suggested range as per the design. How-
ring positions below the Gate, for specific relative
ever, variation of pressure was relatively important
positions of the machine during TBM1 under-passing.
and this can be attributed to the frequent stops and the
Specifically, settlements recorded above rings when
absence of a small stretch of uninterrupted excavation,
each ring position was (a) 2 rings ahead of the
which is necessary to stabilize the face pressure to the
cutterhead, (b) at cutterhead position, (c) at last placed
desired level.
ring position and (d) 6 rings behind last placed ring,
Right graph of Fig. 23 illustrates the variation of
are presented versus the final surface ground surface
middle sensor pressures during the first ring-long
settlements along the Gate.
excavation of TBM1 below the Gate (ring 87) in
The following observations can be made on
comparison with the recommended range as per the
Fig. 25:
design. It is easily observed that due to the slow and
constant advance of the machine along the preceding • Settlement measured ahead of and above the
approximately 9 m, there is no considerable loss of TBM1 cutterhead is practically zero, as recorded
pressure at the beginning of ring excavation, whereas values fluctuate within the accuracy margins of the
pressures are much more uniform and generally within monitoring process.
or slightly below the recommended range as per the
design.

2.0 2.0
Pressure at Middle Sensor (bar)

1.8 1.8
1.6 1.6
1.4 1.4
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
Le Sensor
0.4 0.4
Right Sensor
0.2 Ring 78 0.2 Range as per Design Ring 87
0.0 0.0
0 250 500 750 1000 1250 0 250 500 750 1000 1250
Stroke (mm) Stroke (mm)

Fig. 23 Variation of face pressure of middle sensors along excavation of last ring of calibration drive versus the machine stroke

123
Geotech Geol Eng (2018) 36:3683–3705 3701

2.0 2.0
Pressure at Middle Sensor (bar)

1.8 1.8
1.6 1.6
1.4 1.4
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
Le Sensor
0.4 0.4
Right Sensor
0.2 Ring 90 0.2 Range as per Design Ring 91
0.0 0.0
0 250 500 750 1000 1250 0 250 500 750 1000 1250

2.0 2.0
Pressure at Middle Sensor (bar)

1.8 1.8
1.6 1.6
1.4 1.4
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 Ring 92 0.2 Ring 93
0.0 0.0
0 250 500 750 1000 1250 0 250 500 750 1000 1250
Stroke (mm) Stroke (mm)

Fig. 24 Variation of face pressure of middle sensors along excavation of below the Chandpole Gate versus the machine stroke for
TBM1

8
2 Rings ahead of Cuerhead
6 Above Cuerhead
Above Last Placed Ring
Measured Selements (mm)

4 6 Rings behind Last Placed

Final Surface Selement
2

-2

-4

-6

-8

-10
86 87 88 89 90 91 92 93 94 95 96
Ring No

Fig. 25 Variation of measured surface settlements below the settlement values for cutterhead at ring position 110 (approx.
Gate for various relative positions of the TBM1 versus the final 20 m, i.e. 3D ahead of the Gate)
surface settlement along the Gate (left) and snapshot of recorded

• Only a minor increase of settlements is observed only locally reach 7 mm, whereas practically
above the TBM1 shield up to the placement of the along the Gate are within a range of 4–6 mm.
last ring. • Properly performed grouting process is reflected
• Settlements are mainly recorded during the grout- on the effective control of settlements, which,
ing process, resulting to final settlement values that combined with some skin friction effects of the

123
3702 Geotech Geol Eng (2018) 36:3683–3705

shield, result to the recording of some heave in the grouting consumption was approximately 15% higher
area just after the Gate (where the weight of the than the theoretical volume. Most of the excess
structure is no longer applied). grouting volume can be attributed to some grouting
• Recorded settlement results are in remarkable intake from the granular silty sand formation. Espe-
agreement with the theoretically calculated surface cially along the tunnel portion exactly below the Gate,
settlements, as per the combined analytical-com- the grouting volume was almost equal to the theoret-
putational method, presented in the previous ical one.
section, i.e. 6 mm for the ground surface and Regarding the measured displacements of the Gate
5 mm for the Gate foundation. from the installed movement markers on both phases
of the monument, as it is evident from the results of the
Regarding TBM2, experience from TBM1 underpass- online monitoring system presented in Figs. 28 and
ing and the respective calibration drive, resulted to the 29, the maximum recorded displacements did not
underpassing of the Gate with only minor settlement, exceed 2 mm.
which did not exceed 2 mm above TBM2, as shown in
Fig. 26 (transversal section).
This result can be attributed to the maintenance of 6 Conclusions
face pressures at the upper limit of the design
recommended levels, as analytically shown in the Jaipur Metro Phase 1B, along its initial tunnel part,
previous section, the further improvement of shield under-passed the very important heritage masonry
skin friction effects and the optimization of the structure of the Chandpole Gate, part of the old city
grouting process, in terms of timing, grouting quan- pink walls. Two 6.6 m diameter tunnels were exca-
tities and grouting pressures. Figure 27 below illus- vated with EPB machines through low cohesion
trates the primary grouting consumption during TBM2 alluvium silty sand under the Gate, having a soil
drive, in comparison with the theoretical annulus cover of only 4.5 m from the Gate’s foundation.
volume between the excavation profile and the The vast importance of the structure to be under-
segmental lining extrados per ring. It can be observed passed, necessitated a strict, detailed and structured
that the actual grouting volume is in any case similar process from the investigation, through the design and
or slightly higher than the theoretical one, indicating a the construction process. Investigation of the soil
smooth grouting process with minor overexcavation conditions around the structure’s foundation, from
and effective filling of the annular gap. The average both the Designer and the Contractor, and subsequent

Fig. 26 Calculated, as per

the combined analytical-
computational method, vs
measured surface
settlements

123
Geotech Geol Eng (2018) 36:3683–3705 3703

Fig. 27 Comparison 5
between the primary
grouting consumption per
ring during TBM2 drive and
the theoretical grouting 4

Primary Groung Volume (cum)

volume (dashed line). Rings
corresponding to the drive
below the Chandpole Gate
3
(rings 87–94) are also shown

1
Groung Volume Consumpon

Theorecal Groung Volume

0
70 75 80 85 90 95 100 105
Ring No

Fig. 28 Movement markers locations at the East phase of the Gate and measured displacements from the online monitoring system
after completion of TBM 1 drive

evaluation of the grouting potential below the Gate to analyses, through which the complete construction
enhance soil’s arching characteristics, resulted in the process was simulated to predict surface and Gate
decision to proceed without any soil treatment, as its settlements. Results of the numerical analyses, in
effectiveness was doubtful. terms of induced settlements and plastic zone around
Therefore, for the elaboration of an effective the tunnels, initially verified the internal support
design, a combined analytical-computational method- pressure input values from analytic solution, as surface
ology is used (proposed herein as a new method) for settlements and yield range ahead of and above the
assessing the tunnel internal support pressure and tunnel were minor. An important conclusion for the
prediction of the surface and Gate’s settlements. The particular case was that in order to retain surface
analytic solution of Massinas and Sakellariou (2009) settlements to minimum accepted values, the required
was used as a tool for calculating the required internal internal support pressure was close to the critical.
support pressure. The derived values for the support Thus, no plastic zone was allowed to be developed,
pressure then used as input for detailed 3-D numerical due to the sensitivity and importance of the structure.

123
3704 Geotech Geol Eng (2018) 36:3683–3705

Fig. 29 Movement markers locations at the West phase of the Gate and measured displacements from the online monitoring system
after completion of TBM 2 drive

As regards the construction process, the calibration based on the application of the closed form solution
drive of the TBM before under-passing the Gate was of Massinas and Sakellariou (2009) for deriving the
proved very valuable for the improvement/fine-tuning support pressure, that is used as input for calculating
of TBM operations and the verification of design the surface settlements through 3-D numerical
assumptions. Recorded heave above the TBM shield analysis.
was related to skin friction and was limited along the
calibration drive, whereas design face pressure rec- Acknowledgements The authors of the present study would
like to express their gratitude to Jaipur Metro Rail Corporation
ommended range was applied and found appropriate. (JMRC) and the general contractor Continental Engineering
Improvement of grouting operations was also possible Corporation (CEC) and especially the TBM team, for providing
and a valuable learning curve of the TBM team as well us with the essential information during elaboration of the
as the cooperation of the Contractor’s and Designer’s present study. The authors would like also to express their
special gratitude to Mr. Gabriele Tidone (First Project Director
staff was built up.
of CEC) for his valuable support and guidance during the
During under-passing of the Gate (a) the slow and elaboration of the design.
steady process of the machines, (b) the application of
the appropriate face pressure as per the design,
especially in TBM2, (c) the minimization of shield References
skin friction (especially in TBM2) and (d) and the
constant improvement and synchronization of grout- Anagnostou G, Kovári K (1994) The face stability of slurry-
ing operations, resulted in surface settlements and shield-driven tunnels. Tunn Undergr Space Technol
9:165–174
Gate deformation that was within the anticipated Anagnostou G, Kovári K (1996) Face stability conditions with
range, as per the design’s prediction derived form the earth-pressure-balanced shields. Tunn Undergr Space
combined analytical-computational methodology that Technol 11:165–174
is adopted. Especially during underpassing of TBM2, Atkinson JH, Potts DM (1977) Stability of a shallow circular
tunnel in cohesionless soil. Géotechnique 27:203–215
recorded deformation was minor. During both TBM Broms BB, Bennermark H (1967) Stability of clay at vertical
drives, face pressure was maintained at the levels openings. J Soil Mech Found Div 93(SM1):71–95
recommended by the design. Recorded surface settle- Burd HJ, Houlsby GT, Augarde CE, Liu G (2000) Modelling
ments ahead of and above the machine were within the tunnelling-induced settlement of masonry buildings. Proc
Inst Civil Eng Geotech Eng 143(1):17–29
accuracy limits of the monitoring procedure, thus Burghignoli A, Di Paola F, Jamiolkowski M, Simonacci G
verifying the results of the new proposed combined (2010) New rome metro line C: approach for safe-guarding
analytical-computational methodology; which is ancient monuments. In: Invited lecture, international

123
Geotech Geol Eng (2018) 36:3683–3705 3705

conference on geotechnical challenges in megacities, space obeying Mohr-Coulomb criterion. Géotechnique

Moscow, pp 1–24 59(8):691–701
Chambon P, Corté JF (1994) Shallow tunnels in cohesionless Massinas S, Sakellariou MG (2010) Feasibility study for tunnel-
soil: stability of tunnel face. J Geotech Eng 120:1148–1165 building interaction by using the analytic solution for a
Chen RP, Tang LJ, Ling DS, Chen YM (2011) Face stability circular tunnel in an elastic-plastic half space. In: Petru-
analysis of shallow shield tunnels in dry sandy ground khin, Ulitsky, Kolybin, Lisyuk, Kholmyansky (eds) Pro-
using the discrete element method. Comput Geotech ceedings of the international geotechnical conference on
38(2):187–195 ‘‘Geotechnical Challenges in Megacities’’, Moscow, June
Davis EH, Gunn MJ, Mair RJ, Seneviratne HN (1980) The 2010, Russia, vol 3, GRF 190005, St. Petersburg,
stability of shallow tunnels and underground openings in pp 751–757
cohesive material. Géotechnique 30:397–416 Massinas S, Singh D, Saini MIS, Bhardwaj V, Aishwarya K
Detournay E, St. John CM (1988) Design charts for a deep (2015) Settlement analysis and monitoring instrumentation
circular tunnel under non-uniform loading. Rock Mech of Delhi Metro’s operational Line-2 tunnels during TBM,
Rock Eng 21(2):119–137 of new Line-8, underpass. In: ITA world tunnel congress,
Giardina G, DeJong MJ, Mair RJ (2015) Interaction between Dubrovnik, May 2015, Croatia
surface structures and tunnelling in sand: centrifuge and Migliazza M, Chiorboli M, Giani GP (2009) Comparison of
computational modelling. Tunn Undergr Space Technol analytical method, 3D finite element model with experi-
50:465–478 mental subsidence measurements resulting from the
Horn M (1961) Horizontaler Erddruck auf senkrechte extension of the Milan underground. J Comput Geotech
Abschlussflaechen von Tunneln. Landeskonferenz der 36(1–2):113–124
ungarischen Tiefbauindustrie (German translation by Mirhabibi A, Soroush A (2012) Effects of surface buildings on
STUVA, Düsseldorf) twin tunnelling-induced ground settlements. Tunn Undergr
Jeffery GB (1920) Plane stress and plane strain in bipolar Space Technol 29:40–51
coordinates. Trans R Soc Lond Engl Ser A Mollon G, Dias D, Soubra AH (2009) Probabilistic analysis and
221(1920):265–293 design of circular tunnels against face stability. Int J
Kim SH, Tonon F (2010) Face stability and required support Geomech 9(6):237–249
pressure for TBM driven tunnels with ideal face mem- Murayama S, Endo M, Hashiba T, Yamamoto K, Sasaki H
brane: drained case. Tunn Undergr Space Technol (1966) Geotechnical aspects for the excavating perfor-
25(5):526–542 mance of the shield mashines. In: The 21st annual lecture in
Kimura T, Mair RJ (1981) Centrifugal testing of model tunnels meeting of Japan Society of Civil Engineers
in soft soil. In: Publications Committee of X.ICSMEE Potts D, Addenbrooke TI (1997) A structure’s influence tun-
(eds) Proceedings of the 10th international conference on neling-induced ground movements. Geotech Eng
soil mechanics and foundation engineering, Stockholm, 125(2):109–125
Balkema, Rotterdam, vol 1, pp 319–322 Prountzopoulos G (2012) Investigation of the excavation face
Leca E, Dormieux L (1990) Upper and lower bound solutions stability in shallow tunnels. In: Doctoral thesis. National
for the face stability of shallow circular tunnels in frictional Technical University of Athens (NTUA), School of Civil
material. Géotechnique 40(4):581–606 Engineering (In Greek)
Liu G (1997) Numerical modelling of damage to masonry Rampello S, Callisto L, Viggianni G, Soccodato FM (2012)
buildings due to tunnelling. In: Ph.D. Thesis, Oxford Evaluating the effects of tunnelling on historical buildings:
University, Oxford the example of a new subway in Rome. Geomech Tunn
Massinas S, Sakellariou MG (2006) A parametric study of the 5(3):275–299
tunnelling-induced surface settlements under 3D condi- Savin GN (1961) Stress Concentration around Holes’’, Perga-
tions. In: 2nd international conference on from scientific mon Press
computing to computational engineering, 2nd IC-SCCE, Strack OE (2002) Analytic solutions of elastic tunneling prob-
Athens, 5–8 July 2006, Greece lems. In: PhD thesis, Delft University Press, Delft
Massinas S, Sakellariou MG (2009) Closed-form solution for Vermeer PA, Ruse N, Marcher T (2002) Tunnel heading sta-
plastic zone formation around a circular tunnel in half- bility in drained ground. Felsbau 20(6):8–18

123