You are on page 1of 9

1 INTRODUCTION

The movement of existing tunnels when new tunnels


are driven nearby is important to tunnel owners and
provides an important consideration for the design of
new tunnelling systems in urban areas. This paper is
based on current research on the subject undertaken
in the School of Engineering at the University of
Birmingham.

An established non-numerical method, capable of
predicting ground movement in three dimensions
caused around a new tunnel face, has been adapted
for undertaking predictions of movement of existing
tunnels.

A method of predicting contours of ground move-
ment around the face of a new tunnel is presented.
The contours assist in visualising the movement tak-
ing place in an existing tunnel in close proximity.
Estimation of the curves of both settlement and rota-
tion by the method are discussed and compared with
data from case histories. Marked asymmetry in
movement curves, associated with the skew, is
noted.

It is suggested that these methods may be used for
predictive purposes as an alternative to three dimen-
sional finite element analyses which, at their present
stage of development, appear to yield variability of




results as well as being relatively expensive and
time-consuming.
2 PREDICTION OF GROUND MOVEMENT
Equation 1, originally from Attewell et al . (1986),
later modified by New and Bowers (1994), may be
used for assessing the vertical movement at any
point in the ground relative to a new straight level
tunnel face.
1
1
]
1

,
_

,
_

,
_


i
X x
ERF
i
X x
ERF
i
y
EXP
i
V
S
s f s
2 2 2 2 2
2
2


(Equation 1)

z t
t
t
dt e z ERF
0
2 2
) (

(Equation 2)
where i is the standard deviation or trough width pa-
rameter in the longitudinal and transverse directions;
vertical displacement is S, at a point (x, y, z); X
s
and
X
f
are the starting and final locations of the tunnel
face; V
s
is the volume loss per unit length, equiva-
lent to the volume of the settlement trough, and
ERF(z) is the error function, where t is a dummy
Estimating the effects of new tunnelling on existing tunnels in soft
ground
M. L. Cooper
Civil Engineering, School of Engineering, University of Birmingham
D. N. Chapman
Civil Engineering, School of Engineering, University of Birmingham
C. D. F. Rogers
Civil Engineering, School of Engineering, University of Birmingham
ABSTRACT: Ground movements caused by tunnelling in urban areas in soft ground are an important consid-
eration to designers of new tunnels as they can affect subsurface structures, such as existing metro line tun-
nels, close to the new tunnelling works. There is often a need for relatively simple and quick methods to as-
sess the likely damage to existing structures caused by the new tunnelling works. The alternative is three-
dimensional finite element analyses, which are relatively expensive and time consuming.

This paper provides guidelines for designers on how to estimate likely movements in existing tunnels. In par-
ticular, these guidelines give information on how to estimate the settlement and rotation of the existing tun-
nel(s) using existing semi-empirical methods. These guidelines are based on extensive analysis of case his-
tory data where monitoring has been undertaken inside existing tunnels during new tunnelling work.
variable. To apply the equations, values of i and V
s

must be known or assumed.

Values of i and V
s
from previous tunnel projects can
be used. An alternative when tunnelling in clay for
estimation of i is to use the methods reported by
Mair et al . (1993).





Based on Figure from Attewell et al (1986)
Figure 1: Idealised contours of a normalised settlement trough




0 5 10 15 20 25 30
Centre line distance - m
-30
-25
-20
-15
-10
-5
0
D
e
p
t
h

-

m
4.1m outside diameter Piccadilly Line tunnels
9m span
Concourse
tunnel
Ground level

Figure 2: Transverse section with settlement contours from
Concourse at Central Terminal Area Station, Heathrow Ex-
press.

Figure 1 presented by Attewell et al (1986) is a plan
illustrating contours of settlement at the (horizontal)
ground surface in multiples of 0.1 of the maximum
settlement. It has been found that Figure 1 also ap-
plies to settlement contours on any horizontal plane
in the ground above the tunnel. Other formulae for
assessing the horizontal movement at any point in
the ground relative to the face of a new straight level
tunnel, and corresponding contour plans, are pre-
sented by Attewell et al (1986). However, these are
not discussed in the present paper.

Contours showing the idealised calculated ground
settlement on a section perpendicular to a new tun-
nel, derived by using Equation 1, are shown in Fig-
ure 2. Details of the particular case history repre-
sented in this figure are given in Section 3.1.
Contours of ground settlement from borehole exten-
someter measurements on a section perpendicular to
a new tunnel have been presented in various publica-
tions including Cording & Hansmire (1976), Glos-
sop (1976) and Deane & Bassett (1995). Similar
contours derived from centrifuge tests were pre-
sented by Mair (1979). Figure 3 shows calculated
ground settlement contours on a section parallel to a
new tunnel axis, which is useful as an aid to visualis-
ing the settlement and rotation occurring in an exist-
ing tunnel when an new tunnel is driven beneath it.
The contours were derived using data from the Pic-
cadilly Line tunnel described in Section 3.1, and
therefore correspond to Figure 2.

The calculations assume that the existing tunnel lin-
ing provided no resistance to movement. The pattern
of rotation does not reflect the skew of the crossing.
Figure 3 shows that when the existing tunnel is 10 to
15m in front of the new tunnel face, an anticlock-
wise rotation is induced and that this rotation in-
creases to a maximum just behind the face and re-
duces to zero at about 30m behind the face.



-30 -25 -20 -15 -10 -5 0 5 10 15 20
Distance from tunnel face - m
-30
-25
-20
-15
-10
-5
0
D
i
s
t
a
n
c
e

b
e
l
o
w

G
r
o
u
n
d

-

m
Ground Level
Axis
New 9m outside diameter tunnel
Tunnel advance direction
Relative positions of 4.1m outside
diameter Piccadilly Line tunnel
New 9m span Concourse tunnel
Tunnel advance direction

Figure 3: Longitudinal section with settlement contours from
Concourse at CTA

The rate of change of rotation is greatest when the
contours are closest together, in front of the face,
Contours represent settlement in mm
Contours represent settlement in mm
Contours represent settlement
in multiples of 0.1 of the maximum
and reduces as the rotation becomes clockwise (i.e.
the anticlockwise rotation is progressively lost) be-
hind the face. Exercises to predict the idealised rota-
tion of an existing tunnel by this method could be
carried out for tunnels at any depth and with any
ground parameters to supplement the methods de-
scribed in Sections 4 and 5.
3 CASE HISTORIES
3.1 Heathrow Express, CTA
Figure 4 is a plan and Figure 5 is a section of the
crossing at Heathrow Central Terminal Area (CTA),
see Cooper and Chapman (1999). Three new 9m di-
ameter parallel tunnels were driven below the inner
and outer 3.81m diameter running tunnels of the
Piccadilly Line near the existing station. The verti-
cal clearance between the new and existing tunnels,
which were all in London Clay, was approximately
7m. The skew angle of the crossing was -21
o
.


Based on information from BAA
Figure 4: Site Plan at Heathrow CTA


Based on information from BAA
Figure 5: Cross section at Heathrow CTA

The existing Piccadilly Line tunnels were lined with
unbolted precast concrete expanded segments. The
new concourse tunnel was driven by the Sprayed
Concrete Lining method (SCL), or NATM, using the
"left hand drift and enlargement" sequence, the face
of the enlargement being some distance behind the
left-hand drift.

Only the effects of the first tunnel driven, the con-
course, are considered in making a comparison be-
tween predicted and measured values. Shortly after
completion, the well-publicised collapse occurred,
but this event had no effect on the monitoring of the
Piccadilly Line.
3.2 Old Street Reconstruction
Figure 6 is a plan and Figure 7 is typical section
through the tunnels at Old Street, described by Coo-
per et al . (2000). This project involved the replace-
ment and enlargement of London Underground Lim-
iteds (LUL) Northern Line twin tunnels near Old
Street Station.


Based on information from LUL
Figure 6: Site plan at Old Street

The tunnels were previously lined with 3.56m di-
ameter cast iron bolted segments, which had cor-
roded and cracked over time due to the presence of
acidic ground water.

The affected linings were progressively removed
and replaced in each tunnel by a 4.68m diameter,
stainless steel segmental lining over a length of ap-
proximately 90m. The northbound enlargement is re-
ferred to as the NBLUL, which was constructed
first, and the southbound is referred to as the
SBLUL.


Based on Bracegirdle et al. (1996)
Figure 7: Typical section, Old Street

The Northern Line tunnels are sited beneath and ap-
proximately parallel to twin main-line 4.88m diame-
ter cast iron lined rail tunnels, owned at the time by
British Rail, having a clearance of about 5m. The
equivalent "skew" is therefore approximately 90
o
. In
this paper, the northbound tunnel is referred to as the
NBBR and the southbound is referred to as the
SBBR. The ground consists of London Clay at the
level of the main-line tunnels and variable water-
bearing sands and clays in the Lambeth Beds forma-
tion at the level of the Northern Line tunnels.
4 PREDICTION OF SETTLEMENT IN AN
EXISTING TUNNEL
Figure 8 shows the layout and sign convention as-
sumed in computing the settlement along an existing
tunnel. The skew angle was measured clockwise
from the line perpendicular to the tunnel axis at the
face. Distances along the transverse settlement
troughs (equivalent to y in Equation 1) were taken
perpendicular to the new tunnel axis and are normal-
ised by i (y/i in the figures). As noted on Figure 8,
the tunnel starting location (X
s
) has been assumed as
-12i, i.e. effectively at infinity. When computing
tunnel movements X
f,
the tunnel final location, is the
position stated on the curve (e.g. on Figure 10, when
assessing the curve with the face at -2i, X
f
was taken
as -2i).


Figure 8: Plan of orientation of new and existing tunnels

To illustrate the process used, the method of generat-
ing the line -2i on Figure 10 is described. Using the
spreadsheet Excel and a skew angle, the x and y co-
ordinates of points along the tunnel axis at longitu-
dinal intervals of 4m were tabulated. Using values
for the concourse tunnel at Heathrow CTA, i =
10.4m, X
f
= -2i, X
s
= 10i, Vs = 0.82m
2
/m, the set-
tlement was worked out for each point using Equa-
tion 1. Values of settlement were then normalised.

Figures 9 to 12 show the predicted vertical move-
ment along an existing tunnel caused by a new tun-
nel crossing beneath at varying skew angles and
with the face at varying distances from the crossing.

-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
-4 -3 -2 -1 0 1 2 3 4
ysq/i
S
/
S
m
a
x
Face at -2i
Face at -i
Face at 0
Face at +i
Face at +2i

Figure 9: Predicted normalised settlement in existing tunnel,
skew angle 0
o

When the skew is 90
o
, as occurs in the case history
at Old Street, the predicted settlement curves take
the form of the longitudinal settlement trough. This
is illustrated in Figure 12, which shows the predicted
settlement along a tunnel when a new
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
-4 -3 -2 -1 0 1 2 3 4
ysq/i
S
/
S
m
a
x
Face at -2i
Face at -i
Face at 0
Face at +2i
Face at +i
Face at +9i

Figure 10: Predicted normalised settlement in existing tunnel,
skew angle 20
o

-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
-4 -3 -2 -1 0 1 2 3 4
ysq/i
S
/
S
m
a
x
Face at -3i
Face at -2i
Face at -i
Face at 0
Face at +i
Face at +2i
Face at +3i

Figure 11: Predicted normalised settlement in existing tunnel,
skew angle 60
o

-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
-2 0 2 4 6 8 10 12 14 16
x/i
S
/
S
m
a
x
xf = 2i
xf = 4i
xf = 61
xf = 8i
xf = 10i
xf = 12i
xf = 13.8i

Figure 12: Predicted normalised settlement in existing tunnel,
skew angle 90
o

tunnel is driven beneath it, parallel and offset from it
by 0.56m. This corresponds to the settlement in the
NBBR tunnel at Old Street caused by the NBLUL
enlargement shown in Figure 7. Curves are shown
when the new face is at 2i, 4i, 6i, 8i, 10i, 12i and
13.8i from the beginning.

5 PREDICTION OF ROTATION IN AN
EXISTING TUNNEL
To predict rotation along the existing tunnel, the ver-
tical movement at pairs of points each side of the
tunnel on the horizontal axis has been calculated us-
ing Equation 1. The difference between movements
at these points, divided by their distance apart and
converted to degrees, gave the rotation. The sign
convention adopted was anticlockwise positive look-
ing along the existing tunnel from the right.

Figures 13 to 16 show rotation along the existing
tunnel caused by a new tunnel at skew angles of 0
o
,
20
o
, 60
o
and 90
o
, with the new tunnel face at pos i-
tions varying from -3i to +3i from the crossing.

0.00
0.02
0.04
0.06
0.08
0.10
0.12
-4 -3 -2 -1 0 1 2 3 4
ysq/i
R
o
t
a
t
i
o
n
,

d
e
g
r
e
e
s
Face at "0"
Face at +i
and -i
Face at +2i
and -2i
Face at +3i
and -3i

Figure 13: Predicted rotation in existing tunnel, skew angle 0
o

-0.06
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08
0.10
0.12
-4 -3 -2 -1 0 1 2 3 4
ysq/i
R
o
t
a
t
i
o
n
,

d
e
g
r
e
e
s
Face at -3i
Face at -2i
Face at -i
Face at 0
Face at +i
Face at +2i
Face at +3i

Figure 14: Predicted rotation in existing tunnel, skew angle 20
o

-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
-4 -3 -2 -1 0 1 2 3 4
ysq/i
R
o
t
a
t
i
o
n

-

d
e
g
r
e
e
s
Face at -3i
Face at -2i
Face at -i
Face at 0
Face at +i
Face at +2i
Face at +3i

Figure 15: Predicted rotation in existing tunnel, skew angle 60
o

Figure 16 shows the rotation along an existing tun-
nel caused by a new tunnel driven offset from it by
6.56m and parallel beneath it, as occurs approxi-
mately in the NBBR tunnel caused by the SBLUL
tunnel at Old Street. The sign convention for rota-
tion is anticlockwise positive looking in the direction
of driving.

-0.04
-0.03
-0.02
-0.01
0.00
-2 0 2 4 6 8 10 12 14 16
x/i
R
o
t
a
t
i
o
n
,

d
e
g
r
e
e
s
xf = 2i
xf = 4i
xf = 6i
xf = 8i
xf = 10i
xf = 12i
xf = 13.8i

Figure 16: Predicted rotation in existing tunnel, skew angle 90
o

Similar patterns of rotation occur irrespective of the
offset of the new tunnel, although the magnitudes
differ. The rotation patterns are of similar form to
the settlement curves on Figure 12.

6 VALIDATION BY MONITORING RESULTS
FROM CASE HISTORIES
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
-4 -3 -2 -1 0 1 2 3 4
ysq/i
S
/
S
m
a
x
Face at -2i
3-Sep-94, D = -
0.9i; E = -2.2i
Face at 0
22-Sep-94, D =
1.5i; E = -0.3i
Face at +i
30-Sep-94, D =
3.2i; E = 1.4i
Face at +3i
5-Oct-94, D = -
3.7i; E = 1.4i
Skew = -21
o
i = 10.3m
Smax = 29mm
S
t
a
t
i
o
n

E
n
d
D = drift face from crossing
E = enlargement face from crossing

Figure 17: Predicted and measured settlement caused by Con-
course in CTA Inner tunnel

Figure 17 shows the predicted and measured settle-
ment along the Outer Piccadilly Line tunnel at
Heathrow CTA caused by construction of the con-
course tunnel on four dates as the tunnel approached
and passed beneath it. The face distances from the
crossing are marked in the legend, "D" referring to
the drift face and "E" referring to the enlargement
face. The exact equivalent face distance from the
crossing cannot be found since the drift and en-
largement both contribute towards the settlement by
unknown amounts. However, taking the mean of the
drift and enlargement face positions, as the equiva-
lent, the curves representing predicted and measured
values correspond reasonably well.

Figure 18 shows the predicted and measured settle-
ments along the SBBR caused by the NBLUL
enlargement at Old Street. Since the lengths of the
enlarged sections were only approximately 90m, the
"S waves" at the beginning and ends of the troughs
form a significant proportion of the settlement
curves.

-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
-2 0 2 4 6 8 10 12 14 16
x/i
S
/
S
m
a
x
Face at 2.2i,
15-Jan-96
Face at 2i
Face at 6.4i,
27-Feb-96
Face at 6i
Face at 9.5i,
8-Apr-96
Face at 9i
Face at 11i,
21-Apr-96
Face at 11i
i = 7.8m
Smax = 5mm

Figure 18: Predicted and measured settlement caused by
NBLUL in Old Street SBBR tunnel

The SBBR tunnel is offset from the NBLUL by ap-
proximately 6m and the maximum settlement in the
SBBR tunnel is 5mm. The correlation between pre-
dicted and measured values is relatively poor and
this was attributed to the stiffness of the existing
tunnel causing a reduction in the settlement.

-0.10
-0.08
-0.06
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08
0.10
-4 -3 -2 -1 0 1 2 3 4
ysq/i
R
o
t
a
t
i
o
n

-

d
e
g
r
e
e
s
Face at -i
3-Sep-94, D = -
0.9i; E = -2.2i
Face at 0
22-Sep-94, D =
1.5i; E = -0.3i
Face at +2i
30-Sep-94, D =
3.2i; E = 1.4i
Face at +3i
14-Oct-94, D =
5i; E =3i
Skew = -21
o
i = 10.3m
vL = 1.3%
S
t
a
t
i
o
n

E
n
d
D = drift face from crossing
E = enlargement face from crossing

Figure 19: Predicted and measured rotation caused by Con-
course in CTA Inner tunnel

-0.09
-0.08
-0.07
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
0.00
0.01
0.02
-2 0 2 4 6 8 10 12 14
x/i
R
o
t
a
t
i
o
n

-

d
e
g
r
e
e
s
Face at 2.2i,
15-Jan-96
Face at 2i
Face at 6.4i,
27-Feb-96
Face at 6i
Face at 9.5i, 8-
Apr-96
Face at 9i
Face at 11i,
21-Apr-96
Face at 11i
i = 8.1m

Figure 20: Predicted and measured rotation caused by NBLUL
in Old Street SBBR tunnel

Figure 19 shows measured and predicted rotation
along the Outer Piccadilly Line tunnel at Heathrow
CTA during driving of the concourse tunnel. Curves
are presented at four different dates when the new
tunnel was passing beneath the existing. The face
distances at each date are shown in the legend, the
drift face being denoted as "D" and the enlargement
face as "E". As with the curves of vertical move-
ment, it is difficult to identify the equivalent face
position since both the drift and enlargement tunnels
contribute to the rotation by an unknown amount.
Anomalies in the comparison of measured and pre-
dicted rotations are considered in the later discus-
sion.

Figure 20 shows measured and predicted rotation
along the SBBL tunnel at Old Street (where the an-
gle of skew is approximately 90
o
, i.e. the tunnels are
parallel) caused by the NBLUL enlargement. Rota-
tion curves along the tunnel are plotted at four dif-
ferent dates and the enlargement face distance from
the start of the drive at each of the dates is given in
the legend. Whilst the general patterns of rotation
are credible, the measured values are generally
somewhat greater than predicted values. This could
be accounted for by the values of i and V
s
adopted in
the predictions.
7 DISCUSSION
7.1 Formulae
Equation 1 provides a convenient method of assess-
ing ground movement caused by a new tunnel, by
using a spreadsheet, without employing the more
time- and resource-consuming numerical methods.
The equation does not take into account the detailed
ground properties, but relies on the simple but well-
known parameters, face loss and trough width.
There is however evidence that, in many instances,
the results are as reliable as or more reliable than
those from numerical methods. In using this method
for prediction of movement of existing tunnels, the
assumption is implicit that the stiffness of the exist-
ing tunnel linings has negligible effect on the results.

Figure 1 shows the contours on any horizontal plane
above the new tunnel expressed as multiples of 0.1
of the maximum settlement, with longitudinal and
lateral distances normalised by the trough width pa-
rameter i. Figure 1 could therefore be used as an al-
ternative to Equation 1 to produce the contours in
Figures 2 and 3, settlement curves in Figures 5, 6, 7
and 8 and rotation curves in Figures 9, 10, 11 and 12
graphically, by adjusting the scales, drawing the
tunnel configuration across the contours and reading
off the contour intercepts. It would only be neces-
sary to calculate the maximum settlement for each
condition by using the well-known Equation 3 (for
example in Mair et al. 1993).

,
_


2
2
2 2 2 i
y
EXP
V
S
s

(Equation 3)
The symbols are defined earlier. However, the de-
gree of precision attainable using this graphical
method would be less than that from using the for-
mula.

Contours of settlement in the field around the new
tunnel in Figures 2 and 3 are based on the geometry
of the tunnels at CTA, using measured parameters.
Whereas published records of contours on a trans-
verse section based on measurements are available,
no previous illustrations of contours on a longitudi-
nal section have been found. For design purposes,
Equation 1 could be used to produce ground settle-
ment contours for any configuration of tunnels.
7.2 Settlement
Settlement and rotation predictions derived by the
method herein do not take into account the stiffness
of the existing tunnel.

From Figures 9 to 12, it is apparent that when the
skew is zero, the settlement troughs are symmetrical
about the centre-line. However, when the skew an-
gle increases s ignificantly, the settlement troughs are
asymmetrical when the face is approaching and
passing beneath the crossing. The trough tends to
become symmetrical when the face is well past the
crossing and the asymmetry is thus transient. It is,
perhaps, surprising to find from Figure 11 that when
the skew is 60
0
and the face is at -3i, the settlement
is quite pronounced. This is in contrast to the situa-
tion when the skew is zero and the face is at -3i,
when the settlement is negligible.
7.3 Rotation
Figure 13 shows that when the skew is zero, rotation
is symmetrical. It increases to a maximum when the
new tunnel face is at the crossing and decreases to
zero when the new face is at infinity. Figures 14 and
15 show that the rotation curves become markedly
asymmetrical when the crossing is at a significant
skew angle. When the skew is at 60
o
(Figure 15),
the rotation pattern starts to take effect when the face
is at -3i. Figures 14 and 15 show that the rotation
curves approach the form of a sine wave when the
face is well past the crossing (approaching infinity).

Comparison with longitudinal section contours on
Figure 3 shows that when the existing tunnel is 10 to
15m in front of the new tunnel face, an anticlock-
wise rotation is induced. The anticlockwise rotation
increases to a maximum just behind the face and re-
duces to zero at about 30m behind the face.
7.4 Validation by measured results
Figure 17 suggests that the stiffness of the existing
tunnel lining at CTA is negligible. The existing tun-
nel lining consists of 3.8m diameter unbolted ex-
panded precast concrete segments. There is no grout
outside the lining and there is no mechanical linkage
between adjoining segments. The lining stiffness
would therefore be expected to be small. The
asymmetry effects caused by the 21
o
skew angle of
the crossing appear to be reflected in the measured
curves.

The linings of the BR tunnels at Old Street were
4.88m diameter heavy cast iron bolted linings with
an outside grouted annulus whose stiffness would be
expected to be relatively significant in comparison
with, for example, those at CTA. Figure 18 shows
that the correlation between the measured and pre-
dicted settlement is relatively poor at the end of the
drive and this is consistent with it being reduced by
the stiffness of the existing tunnel.

In Figure 19 the maximum measured rotation when
the face is at 0 exceeds the predicted value by ap-
proximately 30%. The minimum value when the
face is at approximately +3i on 14 October 1994 ex-
ceeds the predicted value by approximately 100%.
Thus, whilst the general shapes of the predicted
curves agree with the measured shapes, their magni-
tudes differ considerably. It is suggested that the
lack of conformity may have been due to local level-
ling errors since the computation of measured rota-
tion depends on small differences in level of the or-
der only 2 to 3mm.

In Figure 20 the measured rotations show consider-
able scatter. This may be accounted for by variation
in the resistance to rotation along the tunnel pro-
vided by the bolted cast iron tunnel lining.

Consideration has been made elsewhere by the au-
thors of methods of predicting the asymmetry of a
settlement trough caused by the second of closely
spaced twin tunnels (Cooper et al . 2002) caused by
interaction between settlement fields of two tunnels.
8 CONCLUSIONS
In order to predict ground movement around tunnels
the general equations may be used to produce con-
tours of movement. Figures 2 and 3 show the gen-
eral form of the contours but the geometry and pa-
rameters (measured or assumed) at an individual site
must be entered to estimate the magnitudes.

To predict settlement along an existing tunnel, the
general equations may be used and the effects of a
skew crossing angle may be taken into account. Al-
ternatively a graphical method is described, although
this would be less precise. For a preliminary as-
sessment, since the results are normalised, Figures 9
to 12 may be used if the skew angle at the crossing
corresponds to one of those illustrated (i.e. 0
o
, 20
o
,
60
o
or 90
o
).

To predict rotation along an existing tunnel, the gen-
eral equations may be used and the effects of a skew
crossing angle may be taken into account. Figures
13 to 16 give a general idea of the anticipated shapes
of the rotation curves. However, they are of limited
use in assessing the magnitude of the rotation gener-
ally since the rotations are not normalised. It is rec-
ommended that rotations be assessed by using Equa-
tion 1.

The formulae from New and Bowers (1994) are pro-
posed for predictions as an alternative to numerical
methods, particularly at the conceptual stages of the
tunnelling process when the design is evolving.
9. ACKNOWLEDGEMENTS
The authors would like to thank the UK Engineering
and Physical Sciences Research Council for their
support in the research project, and London Under-
ground Ltd., Charles Haswell and Partners Ltd. and
BAA plc for their support and permission to use the
data.
10. REFERENCES
- Attewell, P.B., Yeates, J. and Selby, A.R.
1986, Soil movements induced by tunnelling
and their effects on pipelines and structures,
Blackie, Glasgow.
- Bracegirdle, A., Jefferis, S.A., Tedd, P.,
Crammond, N.J., Chudleigh, I. and Burgess
N. 1996, The investigation of acid generation
within the Woolwich and Reading Beds at
Old Street and its effect on tunnel linings,
Geotechnical aspects of underground con-
struction in soft ground, p. 653-658, City
University, London.
- Cooper, M.L. and Chapman, D.N. 1999. Set-
tlement, rotation and distortion of Piccadilly
Line tunnels at Heathrow, Proc. Int, Symp. on
Geotechnical Aspects of Underground Con-
struction in Soft Ground, p. 213-218, Tokyo.

- Cooper, M.L., Chapman, D.N., Rogers,
C.D.F. and Chan, A. H. C. 2000. Movements
of existing tunnels due to the enlargement of
Northern Line tunnels near Old Street, Proc.
World Tunnelling Congress 2000, Durban, p.
101-108, pub. by SAIMM.
- Cooper, M.L., Chapman, D.N. and Rogers,
C.D.F. 2002, Prediction of settlement in an
existing tunnel caused by the second of twin
tunnels, Proc. 81st TRB annual meeting,
January 2002, Washington DC, USA.
- Cording, E.J. and Hansmire, W.H. 1975,
Displacement around soft ground tunnels,
Proc. 5th Pan Am Conf Soil Mech. and
Found Eng. Buenos Aires, Vol. 4, p. 571-
633.
- Dean, A.P. and Bassett, R.H. 1995, The
Heathrow Express Trial Tunnel, Proc Inst.
Civil Engineers, Vol. 113, July, p. 144-156.
- Glossop, N.H. 1978, Soil deformations
caused by soft ground tunnelling, Ph D The-
sis, University Durham, UK.
- Mair, R.J. 1979, Centrifugal modelling of
tunnel construction in soft clay, Ph D Thesis,
University of Cambridge, UK.
- Mair, R.J., Taylor, R.N. and Bracegirdle, A.
1993. Subsurface settlement profiles above
tunnels in clays, Geotechnique, Vol 43, No
2, p 315-320.
- New, B.M. and Bowers, K.H. 1994, Ground
movement model validation at the Heathrow
Express Trial Tunnel, Tunnelling 94; p. 301-
329, pub. by IMM, London, UK.

You might also like