International Journal of Engineering Trends and Technology (IJETT) - Volume 67 Issue 4 - April 2019

Geotechnical Engineering Design of a Tunnel

Support System - A Case Study of Bheri
Babai Headrace Tunnel
Deepika Ghorasaini#1, Akal Bahadur Singh*2
#1
Student, M.Sc. in Geotechnical Engineering, Institute of Engineering, Tribhuwan University, Nepal
*2
Professor, Department of Civil Engineering, Institute of Engineering, Tribhuwan University, Nepal

Abstract
The analysis of rock mass strength, geological to divert 40 m3/s of water from Bheri River to the
problem related to tunnel construction and designing Babai River to irrigate 51,000 ha of agricultural land
appropriate support system for Bheri Babai Headrace of Bardia and Banke district of Nepal and also
Tunnel (BBDMP) is presented in the paper. Tunnel thereby generate 48 MW of electricity making it a
squeezing and rock burst is checked before design of multipurpose project. The tunnel alignment lies in
support system. Rock mass classification system (Q Surkhet and Bardia district of Nepal. The Tunnel is
and RMR) are used as empirical method to obtain divided into six sections.[1]
recommended support system. For analytical solution
Convergent confinement method is used to determine The overburden, Q and RMR value of the sections are
required support pressure to resist displacement. presented below:
Finite element modelling using Phase 2 version 7.013 TABLE 1:Input data for Bheri Babai headrace tunnel
is used to check the ground condition and stress Chainage Overburden RMR Q
concentration around the tunnel providing support (m)
systemfor resisting the deformation. The displacement 1+600 560 31 2.4
and radius of plastic zone is reduced after installation 2+800 618.3 41 2.8
of support shown by obtained results. The first two
5+300 408.5 56 4.5
method i.e. empirical and analytical provide first
6+200 346.6 29 1.5
estimate to design while numerical method can be
8+400 132.8 35 2.6
used to verify the performance of support system.
Hence design method providing maximum factor of 12+200 120 44 4.3
safety should is recommended.
Since the chainage 6+200m has least RMR and Q
Keywords - Tunnels, Case study, Support, Squeezing, value and Bheri Thrust also lies in this regionso the
Rock burst, Empirical Method, Analytical Method, design of this section is presented in this paper.
Numerical Modelling II. TUNNELLINGPROBLEMS
I. INTRODUCTION Among various hazards to be faced during
Tunnels are flexible structures during construction construction of tunnel squeezing and rock burst is
of any infrastructure. Tunnel prove to be the shortest among them. According to ISRM squeezing of rock is
alignment proving more efficient and effective the time dependent large deformation which occurs
medium of construction. Due to the short route and around a tunnel and other underground openings and
economic feasibility most of the hydropower tunnels is essentially associated with creep caused by
of Nepal choose tunnels for water conveyance. With exceeding shear strength (limiting shear stress).
ease comes challenges, on the other hand various Experience shows that deeper an opening is made in
hazards are associated with underground construction. rocks; more vulnerable it becomes to rock burst. The
Some of them are swelling, squeezing, rock burst, rock burst is defined as any sudden and violent
spalling etc which are major problems faced during expulsion of rock pieces from an apparently stable
underground tunnelling in weak rocks. Tunnelling in opening. [2]
weak rocks is an iterative process. The impending Criteria for squeezing is checked in the tunnel section
failure that may arise due to the construction of tunnel that is considered. Empirical and semi-empirical
is resisted by proper designing of tunnel support approaches were checked to determine probability of
system. squeezing. Singh (1992)determined squeezing
The aim of this paper is to analyse the geological phenomenon on the basis of Barton’s Q values of
condition and to obtain required support system. rock mass and the present overburden. [3] Goel
Bheri Babai Headrace Tunnel is the first tunnel to be (1994)approach expressed squeezing phenomenon on
built on Siwalik region of Nepal. The BBDMP aims the basis of rock mass number width of tunnel and
height of overburden.[4] Rock mass number is Q

ISSN: 2231 – 5381 http://www.ijettjournal.org Page 33

 International Journal of Engineering Trends and Technology (IJETT) - Volume 67 Issue 4 - April 2019

value with SRF 1. The summary for this method is Poor. The chart provided by Barton is used
shown below. todetermine the support required for the tunnel.
TABLE 2:Summary of squeezing criteria of Empirical IV. TUNNEL SUPPORT DESIGN USING
Approach ANALYTICAL APPROACH
Approach Squeezing Non-squeezing
condition condition The method is also known as Convergent
Singh’s H <350Q1/3 H >350Q1/3 Confinement Analysis which determines
approach displacement of the tunnel and gives the support
Goel’s H<(275N0.33) B0.1 H> (275N0.33) pressure that can control the displacement. Carranza-
approach B0.1 Torres and Fairhust said CCM has three components:
Longitudinal Displacement Profile (LDP), Ground
Among various semi-empirical method to determine reaction Curve and SCC.[8]
tunnel squeezing Jethwa.et.al (1984) approach is used
to check the tunnel squeezing. The uniaxial A. Longitudinal Displacement Profile
compressive strength and in-situ stress is considered LDP is the graphical representation of radial
during the analysis. The summary for this method is displacement that occurs along the axis of
shown is in table. unsupported cylindrical excavation i.e. for the
sections located ahead of and behind tunnel face.
TABLE 3:Summary of squeezing criteria according to
Jethwa et al. (1984) [5] B. Ground Reaction Curve
Nc Condition GRC is the relationship between decreasing
<0.4 Highly squeezing internal pressure pi and increasing radial displacement
0.4-0.8 Moderately Squeezing of tunnel wall ur. The Relationship depend upon
0.8-2.0 Mildly Squeezing mechanical property of rock mass and can be
>2.0 No Squeezing obtained from the elasto-plastic solution of rock
deformation around an excavation.
Criteria for rock burst was checked using empirical
approaches by Hoek and Brown and Grimstad and C. Support Characteristics Curve
Barton. Support characteristics curve is the plot between
increasing pressure Ps on the support and increasing
TABLE 4:Calculation for rock burst criteria
radial displacement ur of the support.
σ1= σh = 1.5+1.2 σv (MPa) Sengupta (1998) for
overburden less than V. CALCULATION OF AVAILABLE SUPPORT
σ3= σh = 1 + 0.5 σv (MPa) 400m [6]
A. Available support for shotcrete or concrete lining
σ1= σh =2.8+1.48 σv Hoek & Brown
The maximum support pressure developed by
(MPa) (1980) for
concrete or shotcrete lining can be calculated from the
σ1= σh= 2.2 +0.89σv overburden up to
following relationship which is based on the theory of
(MPa) 100m [7]
hollow cylinders.
σ (R−t )2
σv= γH pmax
s = cc 1 − c

Vertical Stress 2 R2
The stiffness constant Ks is as follows:
σϴmax = 3σ1 - σ3 Ks =
Ec R 2 −(R−t c )2
Where,
1− υ C R 1− υ C R 2 +(R−t c )2
σϴmax = 3σ3 – σ1 Kirsch Solution
Ec elastic modulus of concrete
υc is Poisson’sratio
σϴr = (Ak-1) σz Hoek & Brown R is external radius of tunnel (m)
(1980). A and B are tc is thickness of the concrete orshotcrete
σϴw = (B-k) σz excavation σcc is unconfined compressive strength of the
constants. shotcrete or concrete

III.TUNNEL SUPPORT DESIGN USING B. Available support for ungrouted bolts and cables
EMPIRICAL APPROACH The maximum pressure provided by the support
system, assuming that the bolts are equally space in
Q-system is very common for rock mass the circumferential direction, is given by:
classification for tunnelling. The other system used in pmax
T
= bf
s sc sl
the research paper is Rock Mass Rating (RMR)
system. According to Q system the rock mass And the stiffness is given by:
1 4l
classification is Poor rock for the section considered. = sc sl +𝑄 Where,
Ks πd b E s
According to RMR system the classification is also
db is the bolt or cable diameter (m)

ISSN: 2231 – 5381 http://www.ijettjournal.org Page 34

 International Journal of Engineering Trends and Technology (IJETT) - Volume 67 Issue 4 - April 2019

l is the free length of bolt or cable (m) The GSI Value is calculated using following relation:
Tbf is the ultimate load obtained from a pullout test GSI = RMR – 5
(MN)
Q is a deformation load constant for the anchor and TABLE 5:Input data for Phase2 for Numerical
head (m/MN) Modelling
Es is Young’s modulus of bolt or cable (MPa) Rock Mass Property Values
Sc is the circumferential bolt spacing (m) Rock Type Mudstone
Sl is the longitudinal bolt spacing (m) Poisson’s Ratio 0.15
σci (MPa) 11.89
C. Available Support for SteelSet mi 8
The maximum support pressure of the set is: GSI 24
As σy D 0.5
pmax
s = And the
sc R Ei (MPa) 4000
E
stiffness is: K = mb 0.2144
SR 2
Where, s 0.00003
σyis yield strength of steel (MPa) a 0.5334
Es is the Young’s modulus of the steel (MPa) Unit Weight (KN/m3) 23
As is the cross-sectional area of the section (m) Vertical Stress 7.971
Sc is the set spacing along the tunnel axis(m)
R is the radius of the tunnel (m) VII. RESULT AND DISCUSSION

In this case, the stiffness of the combined system is Assessment for Squeezing and Rock Burst
determined as the sum of the stiffness of the Condition was done for Chainage 6+200m.
individual components.
K = K 1 + K2 TABLE 6: Checking of squeezing criteria for chainage
Where, K1= stiffness of the first system and K2= 6+200m
stiffness of the second system. Goel Singh Jethwa et. al
Approach Approach approach
VI. TUNNEL SUPPORT DESIGN USING Squeeze Safe Mild squeeze
NUMERICAL MODELLING
TABLE 7: Checking for rock burst for chainage
Numerical methods available for problem solving 6+200m
in geotechnical engineering are Finite Element σc / σϴ For Roof For Wall
Method (FEM), Spectral Element Method (SEM), Hoek and Severe Heavy support
Finite Difference Method (FDM), Finite Volume Brown Spalling Required
Method (FVM), Discrete Element Method (FEM). Grimstad and High stress, Moderate
[10] Barton [9] Usually Slabbing After
Finite Element Method (FEM) is a technique which favourable to 1 hour
approximates the solution of governing differential stability
equations in the mathematical model by dividing the
domain into meshes or grids and applying simpler The support was designed considering the above
equations to individual elements or nodes in the mesh condition for squeezing and rock burst. The empirical,
to approximate the solution by minimizing the analytical and numerical modelling for the section
associated error function. [11] was performed.
Phase2is a two-dimensional elasto-plastic finite
element program. It is used for calculating stresses A. Empirical Method
and displacements around underground openings, and The support system as suggested by Q and RMR
can be used to solve a wide range of mining, System is given in table below.
geotechnical and civil engineering problems. The
detail assessment is using computer software is TABLE 8:Support system suggested by Q System
carried out for chainage 6+200m. The properties are Bolts Shotcrete Steel Sets
adopted as much as possible to real values. The blast 25 mm diameter 7.5 cm thick steel None
damage factor was introduced in Hoek and Brown 2.5 mm long fibre reinforced
failure criterion in 2002, the constant is determined as grouted rock shotcrete
follows.[12]
GSI −100
mb = mi exp TABLE 9: Support system suggested by RMR System
28−14D
GSI −100 Bolts Shotcrete Steel Sets
s = exp
9−3D
1 1 −GSI /15
a = + (e - e−20/3 )
2 6

ISSN: 2231 – 5381 http://www.ijettjournal.org Page 35

 International Journal of Engineering Trends and Technology (IJETT) - Volume 67 Issue 4 - April 2019

Systematic 100-150mm in Light to TABLE 11: Support system suggested by Phase2

Bolts 4-5m crown and 100 medium ribs Steel Ribs (I Beam M Shotcrete
long, spaced 1- mm in sides. spaced 1.5 Type)
1.5 m in crown m where Sectional 0.203m Thickness 300mm
and wall with required. depth
wire mesh. Area 0.0012 Poisson 0.25
m2 Ratio
B. Analytical Method Youngs 2 × 105 Compressive 25
Ground reaction curve and Support reaction curve Modulus MPa Strength MPa
was prepared for the section. Factor of Safetyof 2.09 Poisson 0.25 Tensile 3 MPa
was obtained after combined bolt, shotcrete and steel Ratio Strength
sets was provided. Compressive 250 MPa Rock-Bolts (Fully
Strength Bonded, 3m long)
TABLE 10: Total combined support parameters Tensile 400 MPa Diameter 25 mm
Maximum support pressure (Pmax) 2.676 Strength
Maximum Stiffness (MN/m) 17584.90 Weight 8.9 Bolt 2 × 105
Support installed 80mm Kg/m Modulus MPa

9 Shorcrete 90.00

Rock Bolts
8 Steel Sets
70.00
Combined
7 GRC
LDP
Internal Pressure, pi (MPa)

Distance to the face, X (m)

6 50.00

5
30.00
4

3 10.00

2
-10.00
1
FIGURE 2: Total displacement 75mmand Radius of
plastic zone 32.6m before support installation
0 -30.00
0 0.1 0.2 0.3 0.4 0.5 0.6
Total Displacement, ur (m)
FIGURE 1:GRC, LDP and SCC for chainage 6+200m

C. Numerical Modelling (PHASE2)

The displacement after preparing the model
was seen 75 mm. The radius of plastic zone was 32m.
So,in order to reduce the plastic zone and
displacement support was installed. Care is taken the
tunnel closure is not more than 4% of the tunnel span.
Support capacity diagram is generated for
determining the factor of safety of shotcrete and steel
ribs. For a given factor of safety, capacity envelopes
are plotted in axial force versus moment space and
axial force versus shear force space. Values of axial
force, moment and shear force for the liners are then FIGURE 3:Total displacement 18.51mm and Radius of
compared to the capacity envelopes. The computed plastic zone 9.171m after support installation
liners values must fall inside an envelope so that they
have a factor of safety greater than envelope values. VIII. CONCLUSION
Factor of Safety greater than 2 is accepted. Also,
Empirical method gives very low value of support
there should be no yielding of bolts and liners.
system. Analytical method provides quite fair result
but cannot meet the permissible requirement.
Integrating empirical, analytical and numerical

ISSN: 2231 – 5381 http://www.ijettjournal.org Page 36

 International Journal of Engineering Trends and Technology (IJETT) - Volume 67 Issue 4 - April 2019

modelling, a satisfactory support can be achieved.

The result form Phase modelling show displacement
and radius of plastic zone reduces significantly after
installation of support. The analytical GRC and SCC
helps to determine the appropriate time to install the
support. The empirical and analytical method lead to
determine first estimate of ground behaviour while
numerical modelling can be used to verify the
performance of the excavated ground. So,
recommending the support as suggested in Table 11.

ACKNOWLEDGMENT

The authorsgive sincere thanks to Institute of

Engineering, Tribhuwan University for the
encouragement and MSc. Geotechnical Engineering
coordinator Dr.Indra Prasad Acharya for his constant
help during publication of this paper. The authorsare
obliged to Department of Irrigation for providingwith
data required for completing the research.

REFERENCES
[1] www.bbdmp.gov.np
[2] Singh, B., & Goel R. (2006). Tunnelling in weak rocks.
Elsevier.
[3] Singh, B., Jethwa J., & Dube A. (1992). Correlation between
observed support pressure and rock mass quality. Tunnelling
and Underground Space Technology.
[4] Goel, R., Jethwa, J., & Paithanakar, A. (n.d), Indian
experience with Q and RMR systems, Tunnelling and
Underground Space Technology.
[5] Jethwa J., &Singh B. (1984). Estimation of ultimate rock
pressure for tunnel linings under squeezing rock conditions- a
new approach. Design and Performance of Underground
Excavations (pp. 231-8), Cambridge: ISRM.
[6] Sengupta, S. (1998). Influence of geological structures on In
Situ Stresses. Uttarakhand, India: Department of Civil
Engineering, IIT.
[7] Hoek, E.,& Brown, T. (1980). Underground Excavations in
Rock. Institution of Mining and Metallurgy, (p.527). London.
[8] Carranza-Torres C., & Fairhust, C. (2000). Application of
Convergence-Confinement method of tunnel design to rock
masses that satisfy Hoek-Brown Failure Criterion. Tunnelling
and Underground Space Technology.
[9] Grimstad, E., Barton, N. (1980). Updating the Q system for
NMT. International Symposium on Sprayed Concrete. Oslo
Norway: Norwegian Concret e Association Atkinson, K.E.
(2007). Numerical Analysis. Scholarpedia
[10] Zadeh, A., & Tackley, P. (2011). Computational methods for
geodynamics. Geophysical Journal International.
[11] Hoek, E., Carranza-Torres, C., & Corkum, B. (2002). Hoek-
Brown Failure criterion – 2002 Edition, NARMS-TAC
Conference, Toronto.
[12] M.S Rahul (2014). Case study on design and construction of
tunnel. International Journal of Engineering Trends
andTechnology, 373-380.13(8),

ISSN: 2231 – 5381 http://www.ijettjournal.org Page 37