Available online at www.sciencedirect.

com
Tunnelling and
Underground Space
Technology
incorporating Trenchless
Technology Research
Tunnelling and Underground Space Technology 23 (2008) 638–650
www.elsevier.com/locate/tust

Influences of the valley morphology and rock mass strength

on tunnel convergence: With a case study of Khimti 1
headrace tunnel in Nepal
Gyanendra L. Shrestha a,*, Einar Broch b
a
Department of Geology and Mineral Resources Engineering, NTNU, Trondheim, Norway
b
Department of Geology and Mineral Resources Engineering, Alfred Getz vei 2, NTNU, Trondheim 7491, Norway

Received 30 January 2007; received in revised form 27 November 2007; accepted 14 December 2007
Available online 4 March 2008

Abstract

Underground structures are constructed at the bottom of the valley sides for various purposes and for different reasons. Hydropower
projects and transport tunnels are some of the examples of such structures. In this paper, literatures on topographical effects on the in situ
stresses in valley and fjord sides are reviewed. An attempt is made to correlate stress anisotropy problems with the valley side topography
by using Phase2 numerical modelling. Based on an underground construction case study, fifteen in situ stress measurements and the
Phase2 analysis, stress induced problems have been found to be influenced by the valley morphology. This influence can be monitored
by the convergence measurement and by the stress measurement. In addition to the overburden height, the total valley height and the
slope need to be considered in the assessment of the stress induced problem. The second aspect dealt with is the influence of the rock
strength on the tunnel convergence. In the Khimti 1 headrace tunnel and 66 cases from 15 countries, it has been observed that the tunnel
convergence is larger in the weaker rocks than in the stronger rocks though they may have similar Q-values. Rock type such as gneiss or
phyllite (corresponding to the rock mass strength) is not considered in the Q-system but it has influence on the convergence that takes
place in underground works. Thus, it also needs to be considered in the assessment of potential convergence of an underground structure.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Squeezing; Topography effect; Rock stress; Rock mass strength; Convergence

1. Introduction
Many underground structures have been constructed at
the bottom of valley sides for various purposes and for dif-
ferent reasons. Hydropower projects and transport tunnels
are some of the examples of such structures. Hydroelectric
power is produced by utilising the potential head of water
flow. Maximum potential head is achieved by placing the
powerhouse at the bottom a valley. Thus many hydro-
power plants are located on the valley side. For economical
and environmental reasons, powerhouse and pressure tun-
*
Corresponding author. Present address: 18 Kilcrea Park, Magazine
Road, Cork, Ireland. Tel.: +353 21 4964133.
E-mail addresses: glsanu@yahoo.com (G.L. Shrestha), einar.broch@
geo.ntnu.no (E. Broch).
nel are also constructed underground. As they are less vis-
ible, underground structures also provide protection
against war hazards and sabotage. There are over 500
underground power stations in the world and about 200
of them are located in Norway. The first underground pow-
erhouse in Norway (Såheim II) was constructed in 1916.
However, almost all of the 200 underground power stations
were built between 1950 and 1990 (Broch, 2000).
Similarly, roads and railways at the bottom of steep val-
leys or along steep fjords, are placed underground to avoid
construction difficulties such as slope instability or prob-
lems during operation such as rock falls and snow ava-
lanches. Especially such tunnels are most efficient
solution to reach the villages located at the end of the fjord
with steep hills on the either sides. However, in the
early days of underground constructions in valley sides,

0886-7798/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tust.2007.12.006
 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650
attention was not paid to the stress induced problems.
After 1950 only, the underground work professionals
started recognising the stress induced problems caused by
the valley side topography.
In this paper literature on topographical effects on the
in situ stresses in valley and fjord sides are reviewed. An
attempt is also made to correlate stress anisotropy prob-
lems with the valley side topography by using Phase2
numerical modelling.
The influence of the rock mass strength on the tunnel
convergence is observed in the plotting of ‘Rock type ver-
sus overburden height’ based on the 66 converging tunnels
from the fifteen countries. In the headrace tunnel for the
Khimti 1 hydropower project in Nepal it was observed
(Table 4) that the convergence was significantly larger in
the weaker rocks (such as schist) than in the stronger rocks
(such as gneiss) though they had similar Q-values. Rock
type (corresponding to the rock mass strength) is not con-
sidered in the Q-system.
2. Influence of valley side topography on tunnel convergence
According to an old Norwegian rule of thumb, stress
induced problems may arise when the valley side inclina-
tion exceeds 25° and the depth of the valley is 500 m or
more as shown in Fig. 1. This rule of thumb is used in a
geological environment where Precambrian and Palaeozoic
H
h ley heights varying from 300 m to 700 m but maintaining a
>25o
Fig. 1. Effect of valley side on stress condition in tunnel.
Fig. 2. A typical Phase2 model for assessing valley side effect on the stress anisotropy.
639
rocks are dominating. Typical rocks are gneisses, granites,
diorites and micashists, and the rocks may in general be
classified as stiff and strong rocks.
In the case of a tunnel in a steep valley side the stresses
at the tunnel level might be much higher than induced by
the vertical overburden depth alone. For example in
Fig. 1 the overburden height above the tunnel is ‘h’, but
the stress level around the tunnel is much higher than that
caused by the vertical overburden ‘h’ alone. Stress evalua-
tion should therefore be associated with the height ‘H’.
Further to the rule of thumb, an attempt has been made
here by using Phase2 code for the design of underground
structures to be constructed in valley side. Similar
approach was used by Broch (1981) and presented as
‘Design chart for unlined pressure shafts based on a finite
element model”. It was based on the concept that nowhere
along an unlined pressure shaft or tunnel shall the internal
water pressure exceed the minor principal stress in the sur-
rounding rock mass.
In the present study, a number of idealised, but still typ-
ical valley sides have been analysed. One example of an
idealised model is shown in Fig. 2. In this case the inclina-
tion of the valley side is a = 35°, the bulk density of the
rock mass cr = 2.7 t/m3, Poisson’s ratio m = 0.2 and
the rhor/rvert ratio is 0.5. H is the depth of the valley, h is
the vertical overburden above the tunnel. d is the horizon-
tal distance of the location from the valley side (see Fig. 1).
In order to simplify the analysis, the rock mass has been
considered to be elastic, isotropic and homogeneous. The
details of the coordinates and geometrical input data used
in the Phase2 numerical analysis are given in the Appendix
A.
The Phase2 model has been used for five cases with val-
constant slope angle of 35°. In each case, stress anisotropy
values have been obtained for the different horizontal dis-
tance (d) and vertical overburden heights (h) keeping H
and a same.
640 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650

In situ stress measurement data have been compiled
from the literature for one case in Norway, one case in
Austria and thirteen cases in Italy. These data are given
in Table 1.
A comparison has been made between the Phase2 model
output and the fifteen measured in situ stress values and the
result is given in Fig. 3. The figure has been plotted to show
the level of the stress anisotropy in relation to the product
of the H * tan a/h. In the Phase2 modelling results, H and a
are directly affecting and h is inversely affecting the stress
anisotropy values.
Fig. 3 shows that stress anisotropy given by the Phase2
modelling results are comparable to those measured at 15
project sites. However, all the measured in situ stress
anisotropy data are below four except for one project. In
fact, when stress anisotropy exceeds approximately four,
then the rock mass gets distressed and in situ stress anisot-
ropy comes below the theoretical (given by the modelling)
values. Phase2 modelling results gives the theoretical trend
of the relation between stress anisotropy and the valley
topography.
mass are considered. On the basis of the trend line, it can
be noticed that squeezing takes place at lower overburden
depth in weaker rocks and at higher overburden depth in
stronger rocks, indicating the influence of rock mass
strength on tunnel convergence.
The figure shows that squeezing is frequently encoun-
tered in mudstone. Shale, schist, phyllite and tuff are the
other rock types where squeezing is common. However,
squeezing is observed in diorite and quartzite also, but only
when those rock masses are altered and overburden is rel-
atively high. Some of the squeezing cases in gneiss in Fig. 4
are in altered condition. Metabasic rock types in the figure
are semi-metamorphosed basic (igneous) rocks with chlo-
rite and schist nature. Many of the squeezing cases took
place in or close to fault zones. Rock types mentioned in
the figure are the host rocks if there is a fault zone. How-
ever, there are some cases where squeezing took place even
if there was no fault zone present. In the figure only the ver-
tical overburden is considered. Topographical effects of
valley side are not included as such information has gener-
ally no been available.

3. Sixty six tunnel convergence cases from fifteen countries 4. The Khimti 1 headrace tunnel

Squeezing problems have been encountered in many
tunnels in different geological conditions around the world.
Information has been compiled from sixty six such squeez-
ing tunnelling cases from fifteen countries and is listed in
Appendix B. The information is also presented in Fig. 4.
The figure is organised with the average overburden height
in ascending order excluding the tunnels in Marl soil.
Squeezing takes place at different overburden depths for
different rock types.
A relation can be observed between the rock type and
respective overburden height. To demonstrate the trend
of the relation a line is roughly fitted (no correlation anal-
ysis has been carried out). The trend line does not consider
the tunnels in Marl soils; only squeezing tunnels in rock
Squeezing was one of the main problems encountered in
the headrace tunnel for the Khimti 1 hydropower project in
Nepal. Squeezing related data have been collected and
some of the available approaches have been used to analyse
the squeezing phenomena that occurred in this tunnel.
Validity of those approaches has been checked.
The Khimti 1 hydropower project is located in Janakpur
zone, central development region of Nepal. The project is
located approximately 100 km east of Kathmandu, see
Fig. 5. It is a ‘run of the river’ type of hydroelectric power
project with an installed generating capacity of 60 MW.
The power plant utilises a drop from 1272 m to 586 m
a.m.s.l. in the Khimti khola with the highest head in Nepal.
A concrete diversion weir diverts up to 10.75 m3/s of water

Table 1
In situ stress measurements in Norway, Austria and Italy
S. No Projects H (m) h (m) Slope (a) Tan a H * Tana/h d (m) r1 r3 r1/r3 Reference
1 Heggura Road tunnel, 1290 670 53 1.33 2.56 505 24.8 6.6 3.76 Broch and Sorheim (1984)
station 6020, Norway
2 Felbertal, Austria 1250 70 40 0.84 14.98 83 25 5 5.00 Kohlbeck et al. (1980)
(Schist, Tungsten mine)
3 Entracque a, Italy 1500 380 38 0.78 3.08 486 15.1 5.5 2.75 Martinetti and Ribacchi (1980)
4 Entracque c, Italy 1500 160 38 0.78 7.32 205 6.1 2.3 2.65 Martinetti and Ribacchi (1980)
5 Entracque d, Italy 1500 480 38 0.78 2.44 614 10.1 2.9 3.48 Martinetti and Ribacchi (1980)
6 Piedilago, Italy 1600 390 40 0.84 3.44 465 8.3 2.5 3.32 Martinetti and Ribacchi (1980)
7 Roncovalgrande, Italy 1000 210 37 0.75 3.59 279 24.2 6.2 3.90 Martinetti and Ribacchi (1980)
8 Fiorano, Italy 1800 240 25 0.47 3.50 515 14.7 3.7 3.97 Martinetti and Ribacchi (1980)
9 Edolo, Italy 1700 250 25 0.47 3.17 536 35.6 14 2.54 Martinetti and Ribacchi (1980)
10 Pelos, Italy 350 120 15 0.27 0.78 448 8.5 4.8 1.77 Martinetti and Ribacchi (1980)
11 Salafossa, Italy 1200 280 27 0.51 2.18 550 8.3 4.2 1.98 Martinetti and Ribacchi (1980)
12 Raibl, Italy 590 400 47 1.07 1.58 373 56 17.7 3.16 Martinetti and Ribacchi (1980)
13 Piani di ruschio, Italy 750 100 29 0.55 4.16 180 5.6 3.3 1.70 Martinetti and Ribacchi (1980)
14 Timpagrande, Italy 300 120 32 0.62 1.56 192 3.4 1.1 3.09 Martinetti and Ribacchi (1980)
15 Taloro, Italy 550 240 25 0.47 1.07 515 10.6 7.8 1.36 Martinetti and Ribacchi (1980)
 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650 641

Phase2 data Measured stresses

10

6
1/ 3

0
0 2 4 6 8 10 12 14 16
Htan /h

Fig. 3. Influences of the valley height, slope and vertical overburden height, on the stress anisotropy.

1200

1000

800
Overburden (m)

600

400

200

0
Mud Sand Crushe Andes SerpentTuff Shale Phyllite Fracture Silt MetabaSchist Gneiss Diorite
Marl Flysch
stone stone/ d schist ite inite d stone sics (altered)(altered)
(soil) argilite
Shale quartzite
Rock types

Fig. 4. Relations between the various rock types and the overburden depths in tunnel squeezing cases. The bold straight line is the boundary line above
which stress problem is expected according to Norwegian rule of thumb.

from the river into a 7.9 km long headrace tunnel, and then
through a 913 m long, steel lined penstock shaft inclined at
45° to an underground powerhouse cavern (70 m long,
11 m wide and 10 m high). Finally the water discharges
to Tamakoshi River though a 1.4 km long tailrace tunnel.
The powerhouse cavern is 420 m under the ground surface
and 893 m inside from the entrance portal, see Fig. 6.
The rocks in the area are represented mainly by grey,
coarse-to very coarse-grained, porphyroblastic augen
gneiss, occasionally banded gneiss and granitic gneiss with
bands of very weak, green chlorite and bright grey talcose
schist parallel to the foliation at intervals of 5–15 m. The
area is influenced by several minor thrust faults character-
ised by very weak sheared schist with clay gouge (Fig. 6)
running parallel to the foliation plane. The foliation near
Adit 4 (the saddle of Pipal Danda) has steep dips (45°–
60°), whereas it is gently dipping (15°–35°) in the area
between Pipal Danda and the intake.
642 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650
The Q-method was used to classify the rock mass and to
estimate the tunnel support requirement. In the Khimti
project, the rock masses along the tunnel were divided into
five classes based on the Q-value assigned to it. On the
basis of the Q-value, tunnel support was provided using
the support chart given by Grimstad and Barton (1993).
Five rock classes and required tunnel support for each class
are given in Table 2. Distribution of the rock mass classes
for the total tunnel length of the project is given in Fig. 7.
The figure shows that the weak rock (i.e. Q < 1) content is
about 72%.
During the tunnelling, some stability problems were
encountered with the initially adopted support system. So
the initial support system was supplemented by adopting
adjustments based on the site-specific experiences. Those
adopted adjustments are summarised in Table 3. It was
found to be practical and effective.
Tunnel support estimate and support review were done
in three steps: (1) a tunnel log was prepared and Q-value
was estimated after each round of blast, (2) support
amount was recommended according to the established
Fig. 5. Location and layout map of Khimti 1 Hydropower project (BPC, 1993).
system and (3) observation, monitoring in the problematic
sections and support audit were carried out. Additional
supports were provided where needed as per the observa-
tion and monitoring.
Squeezing was one of the main problems encountered in
the tunnels. Tunnel convergence was measured by using
tape extensometer at more than 50 tunnel sections covering
the entire tunnels. However, other necessary data such as
overburden depth, support type, rock condition and tunnel
geometry are available only for 26 tunnel sections (Statk-
raft Eng. and BPC, 2001). These 26 sections also include
sections with significant squeezing phenomena. The con-
vergence measurement records include date of record, con-
vergence measurement up to tenth of millimetre and
graphical presentation of the records. Other tunnel conver-
gence related data have been extracted from the tunnel logs
and those are given in Table 4.
Influences of the valley side topography and the rock
mass strength, both have been observed in the Khimti 1
tunnel case. The tunnel sections at Adit 1 d/s 500 m,
515 m, 580 m, and; at Adit 4 u/s 974 m, 1013 m and
 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650 643

Fig. 6. Geological cross section showing faults along the tunnel alignment (Sunuwar et al., 1999).

Table 2
Recommended tunnel support for Khimti 1 hydropower project (for tunnel width 4 m considering ESR = 1.6)
Class Q and RMR value Support Description
I Fair and good >4 Spot bolting or unsupported Mainly competent and stable rock
>56
II Poor 1–4 Bolts in pattern 1.5 * 1.5 m Jointed and fractured strong rock with limited
clay and water
44–56 In fractured rock area, and mostly at crown 5 cm fibre
shotcrete shall be applied
III Very poor 0.1–1 Bolts in pattern 1.0 * 1.5 m Heavily jointed/fractured medium strong to weak
rock. Support to be applied not more than 1 pull
behind face. Reduced pull normally not needed
23–44 Fibre shotcrete: crown 10 cm and wall 5 cm
Spiling c/c 0.5 m in crown for Q < 0.4. Spiling length
4 m, or pull +1.5 m at 10°–15° and end fixed
IV Extremely poor 0.01–0.1 Bolts in pattern 1.0 * 1.2 m Weathered/weak rock, may be peeled with pocket
knife. Support to be applied immediately.
Reduced pull be utilised. Max pull length 2.5 m if
water in-leakage
Fibre shotcrete: crown 15 cm and wall 10 cm
3–23 Spiling c/c 0.4 m in crown and 1 m in walls. Spiling
length 4 m, or pull +1.5 m at 10°–15° and end fixed
with bolts or straps
V Exceptionally poor <0.01 Bolts in pattern 1.0 * 1.0 m Very weak rock normally containing >60% clay,
easily separated by fingers. Schist with water.
Support to be applied immediately. Reduced pull
be utilised. Max pull length 2.0 m if water in-
leakage
<3 Fibre shotcrete: 20 cm thick
Spiling c/c 0.3 m in crown and 0.7 m in walls. Spiling
length 3 m, or pull +1.5 m. If needed two layers one at
10°–15° and second at 30°–45° and end fixed with
bolts or straps

1045 m have similar Q-values and situated on the valley
side of topography under the similar vertical overburdens.
However, these two groups have different maximum topo-
graphical overburden heights. The former group has
1100 m and the latter has about 270 m. The convergence
range for the sections with the higher overburden is 32–
160 mm and that for the sections with the lower overbur-
den is in 4–47 mm (Table 5). This difference in the tunnel
convergence might have been caused by the difference in
the maximum topographical heights.
644 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650

Very poor Table 5

44% The valley side effect of topography on the Khimti 1 headrace tunnel
convergence at six tunnel sections
Location Q-value Vertical Total Tunnel
Extremely poor chainage overburden topographical convergence
21% (m) (m) overburden (m) (mm)
Exceptionally Fair to poor
poor 28% Adit 1 d/s 500 0.01 100 1100 160.2
7% Adit 1 d/s 515 0.0045 100 1100 110.0
Adit 1 d/s 580 0.008 111 1100 32.3
Fig. 7. Distribution of the rock mass classes based on the Q-value for the Adit 4 u/s 974 0.008 112 270 7.9
total tunnel length. Adit 4 u/s 1013 0.006 112 270 47.7
Adit 4 u/s 1045 0.008 112 270 4.0

Table 3
Supplementary design principles for squeezing ground
1. Concrete invert for the rock class V
2. Concrete invert for the rock class IV also if the deformation
measurement do not clearly indicate stable condition
3. If the shotcrete in the squeezing areas is highly cracked and
deformed, scaling and replacing with the steel fibre reinforced
shotcrete. If just minor cracks exist, add a layer (30 mm) of shotcrete
In Table 5, the convergences at the sections Adit 1 d/s
580 m and Adit 4 u/s 1013 m are not compatible in the
respective groups. Relatively low tunnel convergence
at the section Adit 1 d/s 580 m might have been caused
by the absence of the shear plane at that section whereas
all the other five sections are crossed by the shear planes.
The section at Adit 4 u/s 1013 m is crossed by the multiple
shear planes with clay filling up to 25 cm thick. It might
have caused the relatively higher tunnel convergence at
Adit 4 u/s 1013 m. Fig. 8 shows that the higher the total
overburden (1100 m) the larger the tunnel convergence
(160 mm) is and the lower the total overburden (270 m)
the smaller the tunnel convergence (7.9 mm) is.
It has been observed that squeezing is larger in the
weaker rock type such as schist than that in gneiss though
they have similar Q-values. For examples, tunnel sections
at Adit 4 u/s 974 m, 1013 m and 1045 m have Q values
0.008, 0.006 and 0.008 respectively. The section at Adit 4

Table 4
Convergence related data for the Khimti 1 headrace tunnel
Location chainage Over burden Tunnel width Rock type Q rci Support provided Measured
(m) (m) (m) value (MPa) convergence
(mm)
Adit 1 d/s 475 98 4.0 AG/STS 0.08 <25 IV 30.9
Adit 1 d/s 500 100 4.2 Sheared SCS 0.01 <25 IV with concwall 160.2
Adit 1 d/s 515 100 4.2 Sheared SCS 0.0045 <5 V with concwall 110.0
Adit 1 d/s 580 111 4.3 Sheared TCS 0.008 <5 V 32.3
Adit 1 d/s 665 112 4.0 AG/S 0.06 25–50 III 11.8
Adit 2 d/s 441 126 4.0 AG 0.3 <100 III 1.3
Adit 2 d/s 601 138 4.0 STS 0.013 <5 V 7.5
Adit 2 d/s 895 198 4.0 G and CS 0.14 25–50 IV 11.5
Adit 2 u/s 1283 212 4.4 AG/SCS 0.04 <25 IV 1.0
Adit 2 u/s 1357 261 4.0 BG/CS 0.095 <50 IV 6.1
Adit 2 u/s 1730 95 4.0 AG/AG with clay gauge 0.065 <25 IV but not 11.6
applied
Adit 3 u/s 15 130 5.0 2 m shear band of AG and S 0.2 25–50 III 17.0
Adit 3 u/s 59 158 4.1 AG/S 0.23 25–50 III 13.2
Adit 3 u/s 200 276 5.0 AG 0.25 <50 III 38.7
Adit 3 u/s 210 276 5.0 AG 0.28 25–50 III 18.2
Adit 3 d/s 220 140 4.0 S 0.009 <5 V 32.0
Adit 3 u/s 235 284 5.0 G 0.09 50 IV 62.0
Adit 3 u/s 340 300 5.0 AG and SS 0.09 25–50 V 14.0
Adit 3 u/s 345 300 5.0 AG and SS 0.05 <25 V 9.0
Adit 4 u/s 503 225 4.0 GG and STS 0.14 5–25 III 9.7
Adit 4 u/s 550 218 4.0 CSS 0.07 25 IV 5.5
Adit 4 u/s 852 114 4.0 BG 0.47 50 III 1.0
Adit 4 u/s 876 114 4.0 BG with shear planes 0.6 50 III 9.7
Adit 4 u/s 974 112 4.0 Sheared AG 0.008 <1 V with mesh 7.9
Adit 4 u/s 1013 112 4.0 Sheared CTS with multiple 0.006 <1 IV mess@crown 47.7
SP
Adit 4 u/s 1045 112 4.0 Sheared AG with clay fill 0.008 <5 V 4.0
AG = Augen Gneiss, BG = Banded Gneiss, GG = Granitic Gneiss, G = Gneiss, STS = Sericite Talcose Schist, TCS = Talcose Chlorite Schist,
CS = Chlorite Schist, SS = Sericite Schist, CSS = Chlorite Sericite Schist, S = Schist, SP = Shear Plane.
 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650 645

2500

112 m
270 m

2000
Tunnel with convergence
7.9 mm

1100 m
1500
Elevation (m)

100 m

Tunnel with convergence

160 mm
1000

500

0
500 1000 1500 2000 2500 3000 3500 4000 4500
Topographical section of tunnel at Adit 4 u/s 974 m & Adit1 d/s 500 m

Fig. 8. Influence of the valley side effect of topography on the convergence at Adit 4 u/s 974 m and Adit 1 d/s 500 m, respectively (shown in the same
figure).

u/s 1013 m consists of sheared schist whereas other two
sections consist of sheared gneiss. Convergences measured
in those tunnel sections are 7.9, 47.7 and 4.0 mm, respec-
tively. Rock type difference is not considered in the Q-
system.
5. Conclusions
Underground structures are constructed at the bottom
of the valley side especially for hydropower and trans-
port tunnels. Based on the underground constructions
and the Phase2 analysis, stress induced problem have
been found to be influenced by the valley side topogra-
phy. This influence can be monitored by the convergence
measurement and by the stress measurement. In addition
to the overburden height, the total valley height and the
slope need to be considered in the assessment of the
stress induced problem.
The second aspect dealt in this paper is the influence of
the rock strength on the tunnel convergence. In the Khimti
1 headrace tunnel and 66 cases from 15 countries, it has
been observed that the tunnel convergence is larger in the
weaker rocks than in the stronger rocks though they may
have similar Q-values. Rock type is not considered in the
Q-system but it has influence on the convergence that takes
place in underground works. Thus, it also needs to be con-
sidered in the assessment of potential convergence of an
underground structure.

Appendix A

Data for Phase2 (version: 6.00) modelling for constant slope 35° and H varying

S. No Coordinates Total height Angle

A B C D E F H a
(x/y) (x/y) (x/y) (x/y) (x/y) (x/y) (m) (°)
1 1428 714 714 0 428 1428 300 35
500 500 500 0 300 300
2 1571 714 714 0 571 1571 400 35
500 500 500 0 400 400
(continued on next page)
646 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650

Appendix A (continued)
S. No Coordinates Total height Angle
A B C D E F H a
(x/y) (x/y) (x/y) (x/y) (x/y) (x/y) (m) (°)
3 1714 714 714 0 714 1714 500 35
500 500 500 0 500 500
4 1857 714 714 0 857 1857 600 35
500 500 500 0 600 600
5 2000 714 714 0 1000 2000 700 35
500 500 500 0 700 700

Appendix B

Sixty six tunnel convergence cases from 15 countries

S. No Country Tunnel Minimum Maximum Rock type Length Type Sources

name overburden overburden (m)
(m) (m)
1 Japan Enrei 110 Mudstone 5994 Rail Aydan et al.
way (1996)
2 Japan Shirasaka-1 150 Mudstone 1296 Rail Aydan et al.
way (1996)
3 Japan Shirasaka-2 150 Mudstone 1765 Rail Aydan et al.
way (1996)
4 Japan Akakura 380 Siltstone, Mudstone 4220 Rail Aydan et al.
way (1996)
5 Japan Mikai 60 Mudstone 1000 Rail Aydan et al.
way (1996)
6 Japan Ibikijo 180 Mudstone 11350 Rail Aydan et al.
way (1996)
7 Japan Siekan 270 Andesite, Tuff, 25000 Rail Aydan et al.
Mudstone way (1996)
8 Japan Nabetachiy 280 Mudstone 9117 Rail Aydan et al.
ama way (1996)
9 Japan Nakayama 400 Tuff 14830 Rail Aydan et al.
way (1996)
10 Japan Orizume 200 Tuff and mud stone 2300 Road Aydan et al.
way (1996)
11 Japan Fujishiro 260 Mudstone 1823 Road Aydan et al.
way (1996)
12 Japan Inari 140 Serpentinite, Shale 1441 Road Aydan et al.
way (1996)
13 Japan Komadome 300 Tuff 2000 Road Aydan et al.
way (1996)
14 Japan Shinfuku 200 Mudstone, Tuff 2400 Road Aydan et al.
way (1996)
15 Japan Myojin 250 Tuff, Shale 3700 Road Aydan et al.
way (1996)
16 Japan Eno 170 Andesite 955 Road Aydan et al.
way (1996)
 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650 647

Appendix B (continued)
S. No Country Tunnel Minimum Maximum Rock type Length Type Sources
name overburden overburden (m)
(m) (m)
17 Japan Kofuchi 280 Tuff, Serpentinite, Slate, 4555 Road Aydan
Granite way et al. (1996)
18 Japan Enasan 1000 Diorite 8489 Road Aydan
way et al. (1996)
19 Japan Nousie 79 Tuff, Mudstone 2992 Road Aydan
way et al. (1996)
20 Japan Mineoka 90 Shale, Tuff 735 Road Aydan
way et al. (1996)
21 Japan Nigamine II 400 Micaschist 3831 Road Aydan
way et al. (1996)
22 Japan Tagami tunnel: 350 Squeezing in 800 m section 6988 Rail Yamazaki
10.5 m dia in sandstone with way et al. (2003)
sandstone blocks
23 Japan Fujikawa:1 90 m2 250 300 Squeezing in 100 m fault 4500 Rail Kobayashi
section in argillaceous way et al. (2003)
(clayey) crushed zone.
Host rocks are sandstone
and volcanic tuff breccia
24 Japan Jiyoshi 250 Squeezing in foliated and 2990 Road Yukio et al.
clayey Surpentinite rock way (2004)
with a fault. It had 90%
clay minerals including
50% surpentinite
25 Austria Arlberg 150 700 Gneiss (altered), Schist 14,000 Road Kovari and
way Staus
(1996);
Steiner
(1996)
26 Austria Karawanke 500 800 Schist Kovari and
Staus
(1996)
27 Austria Galgenberg 100 250 Phyllite Kovari and
Staus
(1996)
28 Austria Inntal 300 Phyllite Kovari and
Staus
(1996)
29 Austria Taurn 200 1100 Phyllite 6000 Road Kovari and
way Staus
(1996);
Steiner
(1996)
30 Austria Strenger tunnel 800 Squeezing in 1250 m 5800 Road John et al.
(twin). Total section, caused sometimes way (2005)
squeezing Length by few meter thick fault
is 2 * 1250 m and sometimes by
schistocity. Host rock is
phyllitic mica schist
31 Italy Fleres 550 Gneiss (altered) Kovari and
Staus
(1996)
(continued on next page)
648 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650

Appendix B (continued)
S. No Country Tunnel Minimum Maximum Rock type Length Type Sources
name overburden overburden (m)
(m) (m)
32 Italy San Vitale 50 150 Mudstone Kovari and
Staus (1996)
33 Switzerland Furka, 1100 Gneiss (altered) 15000 Rail Kovari and
section way Staus (1996);
36 m2 Steiner (1996)
34 Switzerland Gotthard 1050 1150 Gneiss (altered) Kovari and
Staus (1996)
35 Switzerland Gotthard 300 Schist Kovari and
Staus (1996)
36 Switzerland Vereina 700 Mudstone 20000 Rail Kovari and
way Staus (1996);
Steiner (1996)
37 Switzerland Vereina 350 450 Serpentinite 20000 Rail Kovari and
way Staus (1996);
Steiner (1996)
38 France/ Frejus width 1000 1800 Micaceous schist 12570 Road Panet (1996)
Italy 11 m way
39 USA Stillwater 600 Shale 12900 Aquedut Steiner (1996)
(Utah)
40 USA Moffat 300 350 Schist (sheared) 9830 Road Steiner (1996)
(Colorado) way
41 Algeria Sidi 65 Flyschs argilites 990 Rail Panet (1996)
Mezghiche, way
x-section
46 m2.
42 Nepal Khimti 100 130 Schist (sheared) 8000 Water Sunuwar et al.
way (1999).
43 Nepal Modi 70 90 Phyllitic schist, 2000 Water Author
Quarzite (sheared) way
44 Nepal Kali 620 Phyllite (sheared) 6000 Water Author
Gandaki way
45 India Chibro- 280 Crushed red shale Hoek (2001)
Khodri Dia with r
3m cm = 0.7 MPa.
Strain 2.8%
46 India Giri-Bata 380 Slates with r Hoek (2001)
tunnel. Dia cm = 0.8 MPa.
4.2 m Strain 7.6%
47 India Giri-Bata 240 Phyllites with r Hoek (2001)
tunnel. Dia cm = 0.7 MPa.
4.2 m Strain 9%
48 India Loktak 300 Shale with r Hoek (2001)
tunnel. Dia cm = 0.7 MPa.
4.8 m Strain 7%
49 India Maneri 410 Sheared metabasics Hoek (2001)
Bhali stage with r cm = 3 MPa.
II. Dia 7 m Strain 3%
50 India Maneri 480 Metabasics with r Hoek (2001)
Bhali stage cm = 3 MPa. Strain
II. Dia 2.5 m 2.5%
 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650 649

Appendix B (continued)
S. No Country Tunnel Minimum Maximum Rock type Length Type Sources
name overburden overburden (m)
(m) (m)
51 India Uri tailrace 400 500 Crushed graphitic 2000 Water Brantmark
Schist way (1998)
52 India Uri tailrace 400 450 Shale (eocene) 2000 Water Brantmark
way (1998)
53 India Maneri- 700 900 Metabasic (chlorite- 8560 Water Goel et al.
Uttarkashi schist). Strain 9% way (1995)
54 India Maneri Bhali 350 Fractured Quartzite. Hoek
stage 1 Strain 7.9% (2001)
55 Taiwan Pinglin (35– 150 200 Argillite 40236 Rail way
40 MPa)
56 Taiwan Maan 200 Sandstone/shale Hoek
headrace (2001)
57 Taiwan Maan Adit 200 Sandstone/Shale Hoek
A (2001)
58 Taiwan Mucha 110 Sandstone/shale Hoek
tunnel (2001)
59 Taiwan Pengshan 140 Sandstone/shale Hoek
tunnel (2001)
60 Venezuela Yacambu- 600 Graphitic Phyllite Hoek
Quibor (2001)
61 Norway Lrdal 1000 In 500 m thick fault 24500 Road Grimstad
tunnel zone with partly fresh (2000)
and partly weathered
Gneiss blocks and clay
with 65%–80% smectite
giving swelling pressure
up to 0.6 MPa. Host
rock is Precambrian
Gneiss
62 Turkey Bolu tunnel 60 70 Squeezing in clayey Approx Transit Dalgic
(twin): fault gorges. Host rock 3250 European (2002)
Elmalik is Flysoid series. Fault Motorway
Thrust is 50 m thick
63 Turkey Bolu tunnel 200 Squeezing in clayey Approx Transit Dalgic
(twin): fault gorges and 3250 European (2002)
Asarsuyu breccia. Host rock is Motorway
Thrust combination of
Metasedimentary
(Shale and phyllite) and
Metacrystalline
(Amphibolite and
Gneiss). Fault is 150 m
thick
64 Iran Taloun 300 Squeezing in contact 5000 Free Yassaghi
Service zones of Andesite- Motorway and Salari-
tunnel basalt and Tuff rocks, Rad
mainly in altered and (2005)
foliated tuff. No Fault.
Contact zone is 10 m
thick but squeezing
section is 100 m long
(continued on next page)
650 G.L. Shrestha, E. Broch / Tunnelling and Underground Space Technology 23 (2008) 638–650

Appendix B (continued)
S. No Country Tunnel Minimum Maximum Rock type Length Type Sources
name overburden overburden (m)
(m) (m)
65 Greece Patras 20 45 In Soil. Poor quality 270 By-pass Kontogianni
by-pass argillaceous marls with road, and Stiros
tunnel 1. sandy intercalation and completed (2002)
Dia 10 m ground water. GSI 20 to 1999
20
66 Greece Patras 30 In Soil. Poor quality marls 150 By-pass Kontogianni
by-pass with sand and silt. GSI < 20 road, and Stiros
tunnel 2. completed (2002)
Dia 11 m 1998
67 Canada Myra 1020 Squeezing occurred up to 60 m Transport Li and
Falls 60 cm in altered Rhyolite. from drift Marklund
Mine The section was 60 m from stope tunnel (2004)
the nearest mining stope

References ground: a case history of the Strenger tunnel, Austria. Proceedings

Aydan, O., Akagi, T., Kawamoto, T., 1996. The squeezing potential of
rock around tunnels: theory and prediction with examples taken from
Japan. Rock Mech. Rock Eng. 29 (3), 125–143.
Butwal Power Company Pvt. Ltd. (BPC). 1993. Khimti Khola Hydro-
electric Project, Nepal, Feasibility Study, Final Report, vol. 1.
Brantmark, J., 1998. Rock support in weak rock – a study based on
the URI project. Ph.D. thesis, Royal Institute of Technology,
Sweden.
Broch, E., 1981. Design of underground openings for hydroelectric power
plants. South African Tunnel. 4 (1), 14–26.
Broch, E., 2000. Unlined high pressure tunnels and air cushions surge
chambers. In: Stacey, T.R. (Ed.), Tunnels under Pressure AITES-ITA
2000 World Tunnel Congress. South African Institute of Mining and
Metallurgy, Johannesburg, pp. 63–72.
Broch, E., Sorheim, S., 1984. Experiences from the planning, construction
and supporting of a road tunnel subjected to heavy rockbursting. Rock
Mech. Rock Eng. 17, 15–35.
Dalgic, S., 2002. Tunnelling in squeezing rock, the Bolu tunnel, Anatolian
motorway, Turkey. Eng. Geol. (67), 73–96.
Goel, R.K., Jethwa, U.L., Paithakan, A.G., 1995. Tunnelling through
young Himalayas – a case history of the Maneri–Uttarkashi power
tunnel. Eng. Geol. 39, 31–44.
Grimstad, E., 2000. Squeezing in Lrdal tunnel surprised the tunnel
experts (in Norwegian). In: Proceedings of the FJELLSPRENG-
NINGSKONFERANSEN/BERGMEKANIKKDAGEN/GEOTEK-
NIKKDAGEN, Oslo, Norway. pp. 36.1–36.11.
Grimstad, E., Barton, N., 1993. Updating the Q-system for NMT. In:
Proceedings of the International Symposium On Sprayed Concrete,
Fagernes, Norway.
Hoek, E., 2001. Big tunnels in bad rock. ASCE Journal of Geotechnical
and Geoenvironmental Engineering, 2000 Terzaghi Lecture, Seattle,
pp. 726–740.
John, M., Spondlin, D., Ayaydin, N., Huber, G., Westermayr, H., Mattle,
B., 2005. Means and methods for tunnelling through squeezing
of the RETC – 2005, USA.
Kobayashi, T., Sato, J., Kusumoto, F., 2003. Design and excavation of
large-scale tunnels in squeezing rock. In: Saveur (Ed.), Proceedings of
the (Re)Claiming the underground space.
Kohlbeck, F., Roch, K., Scheidegger, A., 1980. In situ stress measure-
ments in Austria. Rock Mech. 9, 21–29.
Kontogianni, V., Stiros, S., 2002. Prediction and observations of
convergence in shallow tunnels: case histories in Greece. Eng. Geol.
63, 333–345.
Kovari, K., Staus, J., 1996. Basic considerations on tunnelling in
squeezing ground. Rock Mech. Rock Eng. 29 (4), 203–210.
Li, C.C., Marklund, P.I., 2004. Field tests of the cone bolt in mines
of Boliden. In: Proceedings of the FJELLSPRENGNINGSKONF-
ERANSEN/BERGMEKANIKKDAGEN/GEOTEKNIKKDAGEN,
Oslo, Norway, pp. 35.1–35.12.
Martinetti, S., Ribacchi, R., 1980. In situ stress measurements in Italy.
Rock Mech. 9, 31–47.
Panet, M., 1996. Two case histories of tunnels through squeezing rocks.
Rock Mech. Rock Eng. 29 (3), 155–164.
Statkraft Engineering AS and Butwal Power Company Ltd. (BPC), 2001.
Khimti 1 Hydropower Project, Project Completion Report.
Steiner, W., 1996. Tunnelling in squeezing rocks: case histories. Rock
Mech. Rock Eng. 29 (4), 211–246.
Sunuwar, S., Shrestha, G., O’Neill, B., 1999. In: Proceedings of the
International Symposium on Engineering Geology, Hydrogeology, and
Natural Disasters with Emphasis on Asia organised by NGS and IAEG.
Journal of Nepal Geological Society, 2000, vol. 22, pp. 227–236.
Yamazaki, M., Hironaka, T., Ito, K., Kato, T., 2003. Tunnelling and
countermeasure design in unexpected squeezing rock zone. In: Saveur
(Ed.), Proceedings of the (Re)Claiming the underground space.
Yassaghi, A., Salari-Rad, H., 2005. Squeezing rock conditions at an
igneous contact zone in the Taloun tunnels, Tehran-Shomal freeway,
Iran: a case study. Int. J. Rock Mech. Min. Sci. 42, 95–108.
Yukio, K., Kazuyoshi, Y., Motohiko, T., 2004. Supports for tunnelling in
squeezing ground. WTC 2004. Tunnel. Underground Space Technol., 19.