See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/270427918

Lessons learnt from installation of field instrumentation

Article in ICE Proceedings Geotechnical Engineering · October 2014

DOI: 10.1680/geng.13.00054

CITATIONS READS

17 1,290

2 authors, including:

Michael Siu Wan

Geotechnical Consulting Group
12 PUBLICATIONS 234 CITATIONS

SEE PROFILE

All content following this page was uploaded by Michael Siu Wan on 01 May 2020.

The user has requested enhancement of the downloaded file.

Geotechnical Engineering Proceedings of the Institution of Civil Engineers

Lessons learnt from installation of field http://dx.doi.org/10.1680/geng.13.00054

instrumentation Paper 1300054
Wan and Standing Received 29/04/2013 Accepted 22/11/2013
Keywords: field testing & monitoring/site investigation/tunnels & tunnelling

ICE Publishing: All rights reserved

Lessons learnt from

installation of field
instrumentation
Michael S. P. Wan DIC, MSc, CEng, MICE Jamie R. Standing DIC, MSc, PhD, CEng, MICE
Geotechnical Engineer, Crossrail Ltd, London, UK; Postgraduate Senior Lecturer, Imperial College London, UK
Researcher, Imperial College London, UK

Imperial College in London, UK, as part of an Engineering and Physical Sciences Research Council (EPSRC) funded
research project and in collaboration with Crossrail urban railway project, is performing field monitoring research to
investigate how tunnelling affects existing tunnels. Comprehensive instrumentation was installed in Hyde Park and
bordering Bayswater Road, beneath which the new Crossrail tunnels were constructed in London Clay below the
existing London Underground Central Line tunnels. Surface and subsurface instruments were installed around the
Crossrail tunnel alignments to monitor the ground response to the tunnel construction. Monitoring systems of
sufficient resolution and accuracy were adopted to achieve high-quality data for assessing the tunnelling-induced
ground response and mechanisms of movement from earth-pressure-balance machine tunnelling. The installation of
surface and subsurface instrumentation took place in the summer of 2011. This paper describes and discusses the
installation of rod extensometers, in-place inclinometers and multi-level vibrating-wire borehole piezometers.
Selection of the appropriate cement–bentonite grout mixes for backfilling these borehole instruments is discussed,
as this is critical for representative measurements of ground response. Some practical challenges arising during the
installation process and how they were overcome are also described. Confidence in the instrument performance is
demonstrated using example monitoring results from the piezometer and extensometer installations.

1. Introduction intention is that lessons learned can be applied to future similar

Research is being undertaken by Imperial College London to instrumentation installation works.
investigate the effects of tunnelling on existing tunnels in London
Clay. The research opportunity arose with the Crossrail construc- 2. Borehole instruments and their
tion of 21 km of underground twin-bore railway tunnels through installation
central London where the new tunnels interface with existing The field research site is located near Victoria Gate, at the north-
networks of London Underground Limited (LUL) tunnels and west corner of Hyde Park, where the Crossrail west-bound and east-
other utility tunnels. The Crossrail western tunnel section in- bound tunnels pass beneath the LUL Central Line running tunnels
volves construction of twin tunnels using 7.2 m dia. earth- just east of Lancaster Gate. The actual crossing is directly beneath
pressure-balance tunnel-boring machines (TBMs) mostly in Lon- Bayswater Road. The location of the instrumentation site in plan
don Clay. The tunnel drives below the existing LUL Central Line and the Crossrail tunnel longitudinal section are shown in Figure 1.
tunnels near Hyde Park enabled a field investigation to be made
of the effect of new tunnel construction on both the ‘greenfield’ 2.1 Instrumentation plan
ground and the ground in the proximity of existing cast iron A plan of the instrumentation layout is shown in Figure 2. There
segmental lined tunnels. The research, involving instrumentation are in total 38 boreholes, each accommodating one or more
and monitoring fieldwork in Hyde Park and Bayswater Road, instruments. Extensometer boreholes and inclinometer boreholes
aims to provide a high-quality case study of tunnel construction are paired so that both vertical and horizontal ground displace-
in London Clay and the mutual interaction of both the new and ments can be determined at the same locations. Piezometers and
existing tunnels. spade cells have been installed around the Crossrail tunnels to
measure pore-water pressures and changes in earth pressures.
This paper describes the instrumentation scheme and the installa-
tion works. A particular focus is given to the selection of grout Most of these instruments are aligned in a main instrument array
mixes used for the installation of borehole instrumentation. transverse to the Crossrail tunnels. In the main array, the
Practical challenges faced during the installation works are instruments are positioned so that they cover, in the southwest part,
discussed. Example monitoring data from the instruments are the ‘greenfield’ ground response and, in the northeast part, the
presented to illustrate their proper functioning and accuracy. The ground response under the influence of the existing Central Line

1
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

Crossrail ea
nd st-bound
st-bou
n e ea
al Li
C entr
LUL nd
t- bou Crossrail wes
t-bound
e wes
tra l Lin
Cen Instrumentation site
LUL
N
rk
Hyde Pa

25 m 0 25 m 50 m

Instrumentation site To Bond Street Station

To Paddington Station

1803 mm ⫻ 920 mm sewer

1575 mm ⫻ 787 mm sewer 1530 mm ⫻ 736 mm sewer Middle level sewer 1676 mm int. dia. Cross passage 2 ⫹ sump

130

120

110
Reduced level: mATD

100

90

80 LUL Central Line

Crossrail east-bound
London Clay tunnels
alignment
70

? ? ? ?
60
? ~60 m below ground
Lambeth Group
50
1850 1900 1950 2000 2050 2100 2150 2200 2250 2300 2350 2400
Chainage: m east-bound

Figure 1. Location of instrumentation site in plan and Crossrail

tunnel longitudinal section

tunnels. Different responses would be expected for these two areas. schematically in Figures 3(a) and 3(b) respectively. Each cross in
There is another instrument array on the pavements of Bayswater Figure 3(a) represents an anchor at which the subsurface vertical
Road parallel to the LUL Central Line tunnels. These boreholes displacements are measured: these are largely distributed around
have been positioned as close as possible to the Central Line the Crossrail tunnels. Installing them around the existing Central
tunnel to observe how the presence of the existing tunnels affects Line tunnels was precluded by site constraints such as heavy
the ground response, compared with ‘greenfield’ conditions. traffic and Thames Water pipes on Bayswater Road. Nevertheless,
it was possible to install some borehole instruments under the
The positions of the installed rod extensometers and in-place pavement of Bayswater Road, where the Central Line tunnel will
inclinometers along the main array are presented in section influence the ground response. As shown in Figure 3(b), there is

2
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

A
el
Surface monitoring point ound tunn
al Line east-b
line Z LUL Centr
00

nel
und tun
21

al Line west-bo
LUL Centr

r Road
Bayswate
00
21

00
22
HP31A Crossrail east-bound tunnel
HP32
e Drive
Carriag
North

Surface monitoring
point line Y
Surface monitoring point
00

North Ride
22

Rod extensometer
In-place inclinometer
N
Crossrail west-bound tunnel
A

Vibrating-wire piezometer
Standpipe piezometer
Spade-shaped pressure cell Surface monitoring
25 m 0 25 m 50 m
point line X

Figure 2. Plan of instrumentation in Hyde Park and Bayswater

Road

a similar arrangement for the inclinometer boreholes in the main monitoring team gained from previous research projects (e.g.
array. The inclinometer and extensometer boreholes are paired Jubilee Line extension and Channel Tunnel rail link projects)
(about 2.5 m apart), to facilitate interpretation of the displace- played an important role in the selection process.
ment data: this is particularly useful when determining resultant
displacement vectors and ground strains. 2.2 Rod extensometers
2.2.1 Measurement principle and performance
In the main instrument array, piezometers and spade cells were Subsurface vertical displacements were measured using multi-
installed, as shown in Figure 3(c), around the Crossrail east- level rod extensometers. A maximum of eight anchors, positioned
bound tunnel to monitor changes in pore-water pressure and total at different depths, attached to stainless steel rods were installed
horizontal earth pressure as the TBMs approach and pass. There in a single borehole. The steel rods were adjusted so that the rod
are two multi-level vibrating-wire (VW) type piezometer bore- tips were just below the reference head at the ground surface
holes within Hyde Park, with six piezometer sensors at different (Figure 4). The depths of the rod tips below the reference head
depths in each. A conventional standpipe piezometer was also were measured using a dial gauge and the reduced level of the
installed just 2 m away from one VW piezometer borehole, at the reference head itself by precise levelling. By combining the dial
same distance from the Crossrail tunnel, to verify the long-term gauge and precise levelling measurements, the reduced level of
pore-water pressures recorded with the VW piezometers. each anchor at each survey time could be determined.

Instrument types were selected for the Hyde Park site considering Rod extensometers generally have a much higher accuracy than
the magnitude of changes predicted by empirical methods. magnetic extensometers. As the dial gauge to be used has an
Although other instruments and techniques may be available to accuracy of 0.05% full-scale range (100 mm), the accuracy is
perform the tasks, the experience of the Imperial College field 0.05 mm, compared with the 1 mm accuracy typically

3
 Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

Hyde Park Bayswater Road

HP29 HP30
HP16 HP17 HP18 HP19 HP20 HP21 HP22 HP23 HP24 HP25
0
Made ground
5 Terrace Gravel
London Clay Unit C1
10

15
Depth below ground: m

London Clay Unit B2

20

25

30

35
London Clay Unit B1
40 London Clay Unit A3ii
Crossrail Crossrail
London Clay Unit A3i
45 west-bound east-bound
tunnel tunnel Extensometer anchor
50
London Clay Unit A2

⫺60 ⫺55 ⫺50 ⫺45 ⫺40 ⫺35 ⫺30 ⫺25 ⫺20 ⫺15 ⫺10 ⫺5 0 5 10 15 20 25 30 35 40 45 50
Distance from Crossrail east-bound tunnel axis: m
(a)
Hyde Park Bayswater Road

HP3 HP5 HP7 HP14 HP15

HP1 HP2 HP4 HP6 HP8 HP9 HP10
0
Made ground
5 Terrace Gravel
London Clay Unit C1
10

15
London Clay
20 Unit B2
Depth below ground: m

25
Existing LUL
30 Central Line tunnels

35
London Clay Unit B1
40 London Clay Unit A3ii

45 Crossrail Crossrail
west-bound east-bound London Clay Unit A3i
tunnel tunnel
50
Uniaxial tilt sensor carriage. London Clay Unit A2
55 Axis perpendicular to Crossrail
east-bound tunnel
60
Biaxial tilt sensor carriage.
Axis perpendicular to Crossrail
65
east-bound/west-bound tunnel

⫺60 ⫺55 ⫺50 ⫺45 ⫺40 ⫺35 ⫺30 ⫺25 ⫺20 ⫺15 ⫺10 ⫺5 0 5 10 15 20 25 30 35 40 45 50
Distance from Crossrail east-bound tunnel axis: m
Figure 3. (Continued on next page) (b)

4
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

Hyde Park Bayswater Road

HP31A
HP32
HP35

HP33
HP39
HP36

HP37
0
Made ground
5 Terrace Gravel
London Clay Unit C1
10

15
Depth below ground: m

20 London Clay Unit B2

25
Existing LUL
30 Central Line tunnels

35
London Clay Unit B1
40 London Clay Unit A3ii
Crossrail Crossrail Spade-shaped pressure cell
London Clay
45 west-bound east-bound Unit A3i
tunnel tunnel Vibrating-wire piezometer
50
London Clay
Standpipe piezometer filter
Unit A2
⫺45 ⫺40 ⫺35 ⫺30 ⫺25 ⫺20 ⫺15 ⫺10 ⫺5 0 5 10 15 20 25 30 35 40 45 50 55 60 65
Distance from Crossrail east-bound tunnel axis: m
(c)

Figure 3. Cross-section A–A: (a) cross-section of rod extensometer boreholes in the main instrument array; (b) cross-section of
proposed in-place inclinometer boreholes in the main instrument array; (c) cross-section of piezometer and spade cell boreholes in
the main instrument array

achieved with a conventional magnetic extensometer. The overall borehole through the ground above the London Clay (in particular
measurement accuracy is governed by that of the precise levelling the Terrace Gravels) was removed, care being taken to maintain
which is about 0.3 mm at best. Sub-millimetre accuracy was the grout level during this operation. Once the backfill grout had
essential at the Hyde Park research site as the maximum hardened, the reference head was installed and the headworks
anticipated settlement was only about 10 mm. constructed. Finally, the upper sections of the stainless steel rods
were replaced with adjustable rod heads which were set to about
2.2.2 Installation 50 mm below the reference head. The dial gauge has a range of
The rod extensometers were installed in boreholes formed by 50 mm, allowing measurements of maximum 50 mm settlement
cable percussion drilling. During final stages of borehole drilling, or heave of the rods and hence anchors relative to the reference
the stainless steel rods, poly vinyl chloride (PVC) sleeves and head.
hydraulic tubing for the anchors at different depths were laid out
and prepared on the ground surface. On completing the borehole 2.3 In-place inclinometers
drilling, a tremie grout pipe was first put down the borehole. The 2.3.1 Measurement principle and performance
assembly of the bottom-most anchor was then lowered down the Subsurface horizontal ground displacements were determined
borehole and, once at the required depth, the prongs of the anchor using in-place inclinometers. Each inclinometer borehole installa-
jacked out hydraulically by way of the tubing to secure the tion consisted of a series of micro-electro-mechanical system
anchor into the wall of the borehole. This procedure was repeated (MEMS) accelerometer sensors mounted on short carriages
for each of the remaining anchors. After these anchors were positioned at selected intervals down a conventional inclinometer
installed, the borehole was ready for backfilling with a cement– casing. Each accelerometer measured the capacitance change in
bentonite grout. Grout mix trials were conducted at the contrac- response to tilt and transmitted the voltage change to a data-
tor’s workshop before commencement of the site work to estimate logger. The MEMS tilt sensors were left in place during and after
the grout properties and pumpability of different design mixes. the tunnel construction so that automated data-logging was
The grout mixes used for various types of instruments are possible. A maximum of 16 sensors were installed in any one
discussed in Section 3. After grouting, casing used to support the single inclinometer casing. The horizontal displacement profile

5
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

Top view

Reference
holes with
machined guides
Dial gauge
Reference head
Levelling plug
Ground level

Cast-iron
lockable cover

Lean mix Rod tip

concrete

Stainless
steel rod

Guide
Rod bracket
sleeving

Edge of
borehole

Anchor head
(approximately
0·35 m long)

Multiple anchors in
a single borehole

Grout

Extendable
prongs

Figure 4. Details of multi-level rod extensometer (Nyren, 1998)

p
along the inclinometer casing could be deduced by integrating equate to individual repeatability 3 16, which is 0.16 mm over
the rotations over assumed appropriate gauge lengths between the 32 m (2 m spacing), or 0.32 mm over 64 m (4 m spacing).
sensors (see Figure 5).
2.3.2 Installation
The MEMS tilt sensors have a measurement resolution of 2 arc Conventional inclinometer casings were installed in the boreholes
seconds. Baseline monitoring after installation showed that formed by cable percussion drilling. Aligning the casing keyways
repeated readings on a number of tilt sensors over a few days fell to the desired orientation (perpendicular/parallel to the Crossrail
within a range of about 4 arc seconds, equivalent to about tunnel alignments) was achieved using a string stretching across
0.08 mm for a sensor spacing of 4 m (i.e. 0.02 mm/m). Since this two pins on the ground, set out by surveying. The boreholes were
potential error occurs independently for each individual sensor, backfilled with a cement–bentonite grout (discussed in Section 3)
the overall repeatability for a maximum of 16 sensors would and the in-place tilt sensors were installed in the inclinometer

6
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

Inclinometer casings at a later stage. Each MEMS tilt sensor is fixed within a
casing
protective aluminium tube, with O-ring seals at its ends, which is
mounted on a short carriage with wheels on one side and a leaf-
spring on the other. Each carriage was pushed along the keyways
MEMS tilt
sensor
of the inclinometer casing to the required depth using PVC rods
Cumulative
carriages horizontal of different lengths and held in position by the leaf-springs.
displacement Consecutive tilt sensors within a single borehole were connected
in series so that all the sensors were read using just one signal
cable coming out from the inclinometer casing. This was
connected to a data-logger placed within the headworks. The
frequency of the data logging was set and the data periodically
Horizontal displacement downloaded from the data-logger to a laptop.
⫽ L tan α

2.4 Vibrating-wire piezometers in fully grouted

Measured MEMS boreholes
Gauge
length, L rotation, α 2.4.1 Measurement principle and performance
The technology of VW type instruments has advanced in recent
years, resulting in better longevity and durability than before.
Historically, conventional Casagrande type standpipe piezometers,
which require large groundwater inflow to or outflow from the
standpipe, were installed within a borehole, surrounded by a sand
filter and sealed with a bentonite plug, the remainder of the
borehole being grouted (Figure 6(a)). When first developed and
Hole base
assumed stable
introduced, single VW piezometers were installed in the same
way (Figure 6(b)). Recently the technique of installing them
using the ‘fully grouted’ method, in which the VW piezometer
Figure 5. Schematic representation of in-place inclinometer units are embedded in a suitable cement–bentonite grout, has
become popular (Dunnicliff, 2008).

Grout Grout

Grout

Bentonite seal Bentonite seal

Sand filler Sand filler

(a) (b) (c)

Figure 6. Schematic representation of: (a) traditional standpipe

piezometer; (b) diaphragm type piezometer with traditional
installation; (c) diaphragm type piezometer in fully grouted
borehole (after Contreras et al., 2008)

7
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

Vaughan (1969) suggests that for diaphragm type piezometers, ground. Mikkelsen (2002) argues that for extensometers, the
which require only very small amounts of water flow to grout should have stiffness equal to or smaller than the surround-
operate, granular filters can be omitted and the piezometer ing ground, while for inclinometers, a stiffer grout is desirable.
sensors can be fully backfilled by low-permeability grout. For piezometers fully grouted in a borehole, Vaughan (1969)
Mikkelsen (2002) discusses the practical side of the fully proves analytically that the error associated with pore-water
grouted method and recommends good practice and appropri- pressure measurement is insignificant if the grout permeability is
ately proportioned cement–bentonite grout mixes. Dunnicliff not larger than 20 times the ground permeability. More recently,
(2008) and Contreras et al. (2008) report successful installation Contreras et al. (2008) demonstrated using finite-element analy-
and monitoring of piezometers installed by the fully grouted ses that the grout permeability could be 1000 times higher than
method at various sites around the world. A schematic diagram that of the ground before significant errors in pore-water pressure
of a borehole piezometer installed using the fully grouted measurement occur. Therefore, three types of grout mixes were
method is shown in Figure 6(c). chosen for backfilling the extensometer, inclinometer and piezo-
meter boreholes. In addition to the stiffness and permeability
The fully grouted method has a major advantage in that the requirements, the grouts should also be readily mixable and liquid
straightforward backfilling procedure allows multi-level piezo- enough to be pumped through the hose and tremie pipe to the
meter sensors to be installed in one single borehole (very difficult borehole.
to achieve reliably with sand filters). This makes it possible to
obtain a pore-water pressure profile with depth at a single In order to select appropriate grouts and investigate their proper-
borehole location, significantly reducing the number of boreholes ties, a number of trial mixes were prepared in the contractor’s
required. workshop. Grout properties such as stiffness and permeability are
affected by many factors, such as the material composition (e.g.
2.4.2 Installation cement type, fineness of bentonite powder, water acidity), mixing
First, the porous filters of the piezometers were de-aired by equipment, mixing sequence and even mixing temperature. On
boiling in water. Prior to installation, pre-installation ‘zero’ site, despite these preliminary trials and investigations, the initial
readings were taken at ground level with the VW piezometer trial grout mixes were found to be too thick to be pumped
sensors in de-aired water. This allows future readings from these through the tremie pipe. Checks were made to ensure that the
sensors to be referenced to atmospheric pressure. On site, six materials were the same for the trial mixes at the workshop and
sensors were installed in a single borehole. A PVC grouting pipe the actual mixes on site; the same sequence of the mixing was
was used for backfilling each borehole and also for holding the followed; the same model of mixer was used too. A thicker grout
sensors at different depths. As the grouting pipe lengths, with consistency could result from small temperature differences
sensors attached, were lowered down into the borehole, their (winter to summer), but perhaps more significantly the trial grout
joints were connected and sealed using duct tape. Low-permeabil- mixes in the workshop were not pumped through the same length
ity cement–bentonite grout (see Section 3) was then pumped by of grout pipe as on site to check the pumpability. This latter
way of the grouting pipe to the base of the borehole until the effect was further compounded on site as the grout pump had
piezometer assembly was fully grouted. During backfilling, the additionally to overcome the hydraulic head of the wet grout
sensors were monitored by a hand-held VW readout unit to check within the borehole.
their responses to the wet grout. Finally, the cables were
connected to a data-logger housed within the headworks. Further grout mix trials were performed on site, using the same
materials and mixer but under the site environment and using a
3. Selection of grout mixes for backfilling stronger piston pump. Successful revised grout mixes were
boreholes determined, but this took time and led to delays in the overall
A key requirement for successful monitoring of subsurface programme of work.
ground response is the appropriate selection of grout for back-
filling the boreholes used for installing instrumentation. Ideally 3.2 Properties of trial grout mixes
the grout provides a substitute material for the soil that has been Ordinary Portland cement (OPC), sodium-bentonite powder and
removed from the borehole, such that the ground response is as pulverised fly ash (PFA) were used for the grout. For both the
close as possible to that of the undisturbed ground, with the initial workshop trial and the final site trial, a mixing sequence as
instrument installed integrally within it. In practice it is not recommended by Mikkelsen (2002) was followed: mixing bento-
usually possible to match realistically every aspect of the ground nite at the end after cement and fly ash. The mixing procedure
and so attention is focused on providing grout properties most started by adding a measured quantity of clean water into a
applicable for the quantity being measured. mixing pan. The proportioned cement was then gradually added
and mixed thoroughly, followed by adding and mixing the fly ash.
3.1 Background considerations Finally, the bentonite powder was slowly added, stirring the
Depending on the instrument types, the backfilling grouts should mixture constantly, to break down any lumps of bentonite that
have different mechanical properties relative to the surrounding formed.

8
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

The initial workshop grout trial involved 12 mixes, of which six simplicity because of the varying nature of soil stiffness with
were not pumpable using a progression cavity pump that the strain). Similarly, grouts with a lower water–cement ratio and
contractor proposed to use on site. Samples of 100 mm dia. were hence lower void ratio generally have a lower permeability,
prepared for the other six pumpable grouts and tested in the although the bentonite content also influences permeability, as
laboratory for unconsolidated, undrained shear strength and per- can be seen from Figure 8. For instance, mix 6, having a lower
meability using triaxial apparatus. The details of the grout mixes void ratio but smaller bentonite content than mix 9, has a higher
and test results are summarised in Table 1 and Figures 7 and 8. permeability, suggesting that a reduction in the bentonite content
would result in an increase in the permeability.
As indicated in Table 1 and Figure 7, generally the shear
strengths of the grouts are insensitive to the confining cell The undrained shear strengths (measured from unconsolidated,
pressures. In the cases of mixes 1, 4 and 6 there was one sample undrained triaxial tests) of soil samples taken from six nearby
in each set of tests that had a significantly lower strength than the boreholes in Hyde Park are plotted against the sample depth
others. As there is no trend in the lower strength with cell below ground level in Figure 9. The representative undrained
pressure and as only a single sample was prepared for each cell shear strengths of grout mix 2 and mix 4 are also shown. It can
pressure, it is assumed that these lower strengths are not be seen that the strength of mix 2 is at the weaker end of the soil
representative. In general, the grout shear strength decreases with shear strength values, whereas over almost the full depth to 50 m,
increasing water–cement ratio (note that mixes with higher corresponding to the base of the deepest borehole at the site, mix
water–cement ratios (and PFA) produce grouts with higher void 4 has a larger strength than the soil. In terms of stiffness, the
ratios). The average secant moduli for all the mixes determined undrained secant moduli of the London Clay samples collected
from the same triaxial tests at 0.25% strain, as presented in Table from the Crossrail ground investigation site near Paddington were
1, range from 18 MPa to 120 MPa and follow in general the same measured from undrained triaxial tests to be between 20 MPa and
trend as the shear strength (note that, in practice, although 40 MPa at large strains (. 0.25%). Again, mix 2 has an
compatibility of stiffness of the grout and the ground is sought, undrained stiffness on the softer side of the soil stiffness range at
often comparisons are made in terms of undrained strength for large strain, whereas mix 4 is considerably stiffer than the soil. In

Mix no. 1 2 4 6 9 12

Mix proportion by 4.0: 1.0: 0.0: 4.0: 1.0: 0.0: 2.5: 1.0: 0.0: 2.0: 1.0: 0.0: 4.0: 1.0: 0.5: 2.5: 0.8: 0.9:
mass (water: 0.72 1.0 0.74 0.5 1.0 0.55
cement: fly ash:
bentonite)
UU triaxial Curing time: d 59 58 38 38 32 32
compression Average sample 5.14 4.93 3.50 3.03 4.16 2.76
testa void ratio
(Confining (100) 91 (100) 152 (200) 117 (200) 702 (100) 145 (200)398
pressure), UU (200) 56 (200) 156 (400) 421 (400) 691 (200) 146 (400) 410
triaxial shear (300) 114 (300) 166 (600) 420 (600) 489 (300) 166 (600) 375
strength: kPa
Average Esec at 18 22 110 80 40 120
0.25% strain: MPa
Permeability Sample void ratio N/A N/A 3.57 2.84 3.72 2.34
test in triaxial (Consolidation N/A N/A (200) (200) (100) (200)
apparatusb pressure: kPa), 6.3 3 109 1.5 3 109 9.4 3 1010 1.3 3 1010
coefficient of N/A N/A (400) (400) (200) (400)
permeability: m/s 4.4 3 109 1.2 3 109 6.0 3 1010 7.0 3 1011
N/A N/A (600) (600) (300) (600)
3.7 3 109 8.9 3 1010 4.2 3 1010 6.0 3 1011
a
UU: Unconsolidated undrained triaxial compression test on 38 mm specimens, in accordance with BS1377, Part 7: 1990: Test 8 (BSI, 1990b).
b
100 mm dia. specimens were consolidated under different cell pressures before permeability was tested under a constant pressure gradient. The
tests were performed in accordance with BS1377, Part 6: 1990, Clauses 5.3, 5.4, 5.5.2.2 to 5.5.2.7 and 6 (BSI, 1990a).

Table 1. Details of trial grout mixes and properties

9
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

800
W: Water
C: Cement
700 PFA: Fly ash
B: Bentonite

600
Mix 6 (2·0W:1·0C:0·5B)
Undrained strength: kPa

500
Mix 4 (2·5W:1·0C:0·74B)
400
Mix 12
300 (2·5W:0·8C:0·9PFA:0·55B)

Mix 2 (4·0W:1·0C:1·0B)
200
Mix 9
(4·0W:1·0C:0·5PFA:1·0B)
100

Mix 1 (4·0W:1·0C:0·72B)
0
0 100 200 300 400 500 600 700
Confining stress: kPa

Figure 7. Unconsolidated undrained triaxial compression test on

trial grouts: undrained shear strengths of 38 mm specimens of
trial grout mixes

1·0 ⫻ 10⫺8
W: Water
C: Cement
PFA: Fly ash
Coefficient of permeability: m/s

B: Bentonite Mix 4 (2·5W:1·0C:0·74B)

(void ratio ⫽ 3·57)
1·0 ⫻ 10⫺9
Mix 6 (2·0W:1·0C:0·5B)
Mix 9 (void ratio ⫽ 2·84)
(4·0W:1·0C:0·5PFA:1·0B)
(void ratio ⫽ 3·72)
1·0 ⫻ 10⫺10
Mix 12
(2·5W:0·8C:0·9PFA:0·55B)
(void ratio ⫽ 2·34)

1·0 ⫻ 10⫺11
0 100 200 300 400 500 600 700
Confining stress: kPa

Figure 8. Permeability test on trial grouts: coefficient of

permeability of trial grout mixes

view of these observations, mixes 2 and 4 were adopted as bentonite grout samples in different projects, from both the
suitable design mixes for backfilling the extensometer and literature and their own measurements. Figure 10 is modified
inclinometer boreholes respectively. after Contreras et al. (2008) and shows the relationship between
the unconfined compressive strength and the water–cement ratio
The influence of cement and bentonite content on grout strength by weight, with the Hyde Park data also plotted (note that the
has been investigated by Contreras et al. (2008). They summarise Hyde Park data are derived from the measured, unconsolidated,
the unconfined compressive strength measured from cement– undrained shear strength, Su which theoretically equals half of the

10
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

Undrained shear strength, Su: kPa

0 100 200 300 400 500 600 700 800 900
0
Borehole 1
Borehole 2
10
Borehole 3
Borehole 4
Depth of soil sample below ground: m

Borehole 5
20
Borehole 6

30
Crossrail tunnel axis level

40

Deepest inclinometer borehole ⫽ 50 m

50

60
Mix 2: Mix 4:
Design grout mix for Design grout mix for inclinometer
extensometer Average Su ⫽ 420 kPa
70 Average Su ⫽ 160 kPa

Figure 9. Undrained shear strength of London Clay borehole

samples taken from Hyde Park

2000
Mix 6, having a coefficient of permeability of 1 3 109 m/s, was
Unconfined compressive strength: kPa

1800 Contreras et al., 2008

Hyde Park* considered to be less than 100 times more permeable than the
1600 surrounding London Clay (for which the lowest permeability was
1400 taken to be about 1 3 1011 m/s) and therefore adopted as
1200 suitable for backfilling VW piezometer boreholes.
1000
800 3.3 Grout mixes used on site
The lack of pumpability of the initial trial mixes on site using the
600 Trend line after Contreras et al. (2008) original progression cavity pump led to modifications being made
400
to the grout mixes for extensometer and inclinometer boreholes,
200 primarily involving reducing the bentonite quantity and adding
0 PFA. The pumpability of the revised mixes was tested by
0 2 4 6 8 10 12 14
Water–cement ratio by weight pumping the grouts into an abandoned 30 m borehole using a
stronger piston pump. The final grout mix proportions used are
Figure 10. Measured unconfined compressive strength of shown in Table 2. Owing to the tight installation programme, no
cement–bentonite grout samples from different project sites further strength tests on the modified mixes were performed. It
(after Contreras et al., 2008) (*Hyde Park data are derived from was considered that the small additional amounts of cementitious
measured unconsolidated undrained shear strength) materials used would only result in marginally increased values
of strength and stiffness of the modified grout mixes which
should not significantly affect instrument response and perform-
unconfined compressive strength). All the data points lie reason- ance.
ably well in one single trend line, although different types and
contents of bentonite were used in these projects, suggesting that Ideally grout mix trials should be performed using the same
it is the water–cement ratio that is the governing parameter for materials and equipment on site prior to the grout backfilling.
the grout strength. This would involve careful planning of site activities as it

11
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

Borehole backfill Water Cement Fly ash Bentonite

Extensometera 4.0 (4.0) 1.0 (1.0) 0.5 (0.0) 0.9 (1.0)
Inclinometera 2.5 (2.5) 0.8 (1.0) 0.9 (0.0) 0.5 (0.74)
VW piezometer 2.0 1.0 0.0 0.5
a
Bracketed values represent the initial design mix values.

Table 2. Final grout mix proportions by mass

typically takes weeks to obtain representative laboratory testing than those reported in the literature for the assessment of
results (all grouts should be tested after a minimum 28-day period seismically induced damage to buried pipelines; for example, the
to allow full cement strength to develop). Allowance in grout minimum value reported by O’Rourke and Ayala (1993) to have
pumpability should also be made to take account of the need to caused damage was about 100 mm/s. Nonetheless as a precau-
overcome the wet grout hydraulic head. tionary measure, the reduced strike rate was adopted with the
borehole no closer than 3 m from the pipe. No damage to the
4. Technical challenges water main was reported during the work. Clearly, when utilities
Achieving research-quality instrumentation installations at times pipes are expected, early engagement and close liaison with
leads to additional technical and sometimes contractual chal- utilities companies are vitally important.
lenges compared with those for more standard ground investiga-
tion and instrumentation projects. In this section, three practical 4.2 Time of completion of instrument installation
challenges encountered during the works (in addition to the issues Ideally, drilling, installation of instruments and backfilling of the
relating to grout pumpability) are discussed along with the borehole should be completed in as short a time as possible,
solutions adopted. preferably within one day, to minimise disturbance to the ground
around the borehole due to stress relief and desiccation. This
4.1 Encountering obstructions: Thames Water pipe helps ensure a high instrument conformance.
Prior to drilling on Bayswater Road, several inspection trenches
were excavated to expose an existing water pipe, as required by Finishing a 30 m or deeper borehole with instrument installation
Thames Water, to determine its exact size and location. It was and backfilling within a day was not usually practically feasible,
found to be a cast iron pipe of 600 mm diameter, with the depth as there was a strict control on the working hours (8 a.m. to 6
of the pipe crown ranging from 0.2 m to 0.6 m below ground p.m.) within Hyde Park. In such cases the installation was
level in the vicinity of the boreholes. The pipe is a major water completed over a two-day period. A borehole would be started
main and so Thames Water required the boreholes to be drilled in and left open overnight down to 30 m below ground, with the
a controlled manner outside a set distance from the pipe. borehole being cased over the depth of Terrace Gravels, taking
into account that the London Clay sub-unit B2 to this depth is
A useful measure of potential vibration-induced damage can be relatively more stable than the underlying A3ii London Clay sub-
obtained from peak ground/particle velocities (Bommer and unit (containing silt and sand partings) (Standing and Burland,
Alarcón, 2006). In order to assess potential effects on the water 2006). The borehole drilling was completed the next morning and
main of borehole drilling, a trial was performed within Hyde installation of the instruments and grout backfilling of the
Park, measuring ground vibrations within two 0.5 m deep pits borehole was finished by the end of the afternoon.
excavated at 1.5 m and 3.0 m from the borehole. Measurements
of peak particle velocity (ppv) were made using geophones at the Early discussions with the contractor about specific project
base of the pits during various steps of borehole drilling. Potential requirements and site restraints help to achieve workable solu-
mitigation measures to reduce vibration were investigated, includ- tions.
ing digging a deeper inspection pit before starting cable percus-
sion drilling, advancing the casing by surging instead of 4.3 Borehole diameters
hammering, reducing the rate of hammering and using more Smaller diameter boreholes usually are more advantageous than
water to facilitate penetration of the casing close to the pipe. It larger ones, because faster, more effective drilling can be
was found that the maximum vibration occurred when the casing achieved with less disturbance to the ground. On site, 200 mm
was hammered into the top of London Clay at about 5 m depth. boreholes were drilled initially, although the instruments were
A ppv of 4.4 mm/s was measured at 1.5 m from the borehole able to fit into 150 mm boreholes. The main reason was to cover
when the casing was rapidly struck. However, if a 2–3 s delay the risk of abandoning the borehole in case of borehole wall
was allowed between each casing strike, reduced ppv values of collapse: with a 200 mm cased borehole it is possible to continue
3.9 mm/s and 3.5 mm/s were measured at 1.5 m and 3.0 m drilling with 150 mm casing. If the borehole were to be started
respectively from the borehole. All of these values are far lower with 150 mm casing, there would be a risk of abandoning the

12
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

whole borehole if a borehole collapse occurred beyond the Figure 12 shows how the negative excess pressures dissipated
contractor’s control. with time at different depths for HP32, resulting eventually in the
measured pore-water pressures developing a near-hydrostatic
Drilling rates are much slower for 200 mm compared with profile. Also plotted in Figure 12 are the post-equilibration pore-
150 mm boreholes, especially when stiffer strata or claystones water pressures at the two response zones of a nearby standpipe
are encountered. In order to minimise the ground disturbance piezometer HP31A (6 m apart), which shows that both the VW
and improve the construction progress, it was agreed that an piezometer borehole and standpipe piezometer borehole are
engineer’s instruction should be given to the contractor to drill measuring the same pore-water pressure profile. The different
boreholes of 150 mm dia. Any abandoned hole and redrilling magnitudes of negative excess pressure measured at different
due to borehole collapse induced by factors uncontrollable by depths of the boreholes indicate that the grout backfill formed a
the contractor would be payable works. This was a re- good seal between the piezometer units: otherwise similar
allocation of the risk from the contractor to the client. Based magnitudes of negative excess pressure would have been ex-
on records and experience of drilling in the Hyde Park area, pected. Subsequent measurements of pore-water pressure change
the chance of encountering water strike was deemed to be during the Crossrail TBM passage showed that their magnitudes
low. No collapse within any of boreholes occurred at the site decreased with the distance of the piezometer units from the
during the works. TBM, also indicating a good functioning of the grout seal
between the piezometer units within the boreholes.
5. Preliminary monitoring results
In this section selected data from measurements made using the 5.2 Immediate surface settlement at extensometer
VW piezometers and extensometers are presented to provide boreholes
basic examples of the monitoring data. In the case of an extensometer borehole, the potential adverse
effect of a grout backfill stiffer than the surrounding soil would
5.1 Initial VW piezometer measurements be that the grout column deforms less, making the settlement
Results from measurements of pore-water pressures made during measured at the extensometer reference head appear smaller than
the early weeks after VW piezometer installation in borehole that of the ground. At the Hyde Park site, as shown in Figure 2,
HP32 (see Figures 2 and 3(c) for location) in Hyde Park are an array of extensometer boreholes runs parallel to a line of
shown in Figure 11. Negative excess pore-water pressures around shallow surface monitoring points (line Y) with a separation of
the boreholes were generated from stress relief and also shearing 5 m. Each surface monitoring point comprises a 1.5 m deep,
of the over-consolidated London Clay during the borehole 100 mm dia. concrete column with a monitoring point installed in
drilling. The equilibration of the excess pressure was complete the top. Thus a direct comparison can be made between the
about 1 month after instrument installation. surface settlements measured at the extensometer reference heads
and at the surface monitoring points, when the two TBMs passed
beneath them. For each tunnel construction, the TBM control was
essentially the same over the 5 m distance between the two lines
Piezometer of instruments and clearly ground conditions are the same. The
tip depth:
immediate incremental settlement troughs induced by construc-
350 42·0 mbgl
tion of the first and second TBMs measured by precise levelling
37·2 mbgl along these two lines of instruments are shown in Figures 13(a)
300
and 13(b) respectively. The measured magnitudes of surface
Pore-water pressure: kPa

250 settlement at every offset position of the instruments lying

31·0 mbgl transverse to the TBM axis are essentially the same, suggesting
200 that the grouted extensometer boreholes conformed well with the
24·5 mbgl
surrounding ground.
150
18·0 mbgl
6. Summary and conclusion
100 12·0 mbgl As part of a collaboration field monitoring research project
involving Imperial College and Crossrail, high-quality borehole
50 instrumentation was successfully installed in Hyde Park and
Bayswater Road. Reasons for instrument selection are given and
0 discussed in the paper. An important factor for achieving good
0 10 20 30 40 50 60 70 80 90 100
Time after installation completion 7 June 2011: d conformance of the instrumentation to the surrounding ground
conditions is the choice of grout for backfilling the boreholes.
Figure 11. Initial monitoring results of VW type piezometers from Ideally for extensometers and inclinometers the grout should
borehole HP32 (mbgl: metres below ground level) have a compatible stiffness to the ground. Equally grout and
ground permeabilities should be compatible for piezometer

13
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

5 Terrace Gravel
London Clay Unit C1

10

Hydrostatic line No. of days since

Water table at 4·5 mbgl installation completion:
2 6 10
15

23 37 54
Depth below ground: m

20
Measurement at response zone
(24·5 m below ground) of
standpipe piezometer HP31A
after equilibration
25

London Clay Unit B2

30
Measurement at response zone
(37·0 m below ground) of
standpipe piezometer HP31A
after equilibration
35
London Clay Unit B1

40 London Clay Unit A3ii

London Clay Unit A3i

45
0 100 200 300 400 500
Pore-water pressure: kPa

Figure 12. Measured pore-water pressure profile along the depth

of borehole HP32

installations. A number of grout mixes were trialled and the Acknowledgements

final choices were made according to strength and permeability The authors wish to acknowledge Crossrail and the Engineering
criteria, as well as practical mixing and pumping considerations. and Physical Sciences Research Council (EPSRC) who are major
Other issues that had to be taken into account are also described sponsors of the research project at Imperial College. Many thanks
(sensitivity of a cast iron water main, time constraints and are owed to the Imperial College research team, in particular
optimum borehole diameter). Example data are presented to Professors John Burland and David Potts, for their contributions
confirm good operation of instruments and selected grout for in discussing the choice and layout of the instrumentation. A
backfilling them. thoughtful review of the instrumentation plan from Mr John

14
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

⫺1

⫺2
Vertical displacement: mm

⫺3

⫺4

⫺5

⫺6
Extensometer reference head

⫺7 Surface monitoring point

⫺8
⫺45 ⫺40 ⫺35 ⫺30 ⫺25 ⫺20 ⫺15 ⫺10 ⫺5 0 5 10 15 20 25 30 35
Distance from TBM1 (WB) axis: m
(a)
0
⫺1
⫺2
⫺3
⫺4
Vertical displacement: mm

⫺5
⫺6
⫺7
⫺8
⫺9
⫺10
⫺11
Extensometer reference head
⫺12
Surface monitoring point
⫺13
⫺14
⫺60 –55 –50 –45 –40 –35 –30 –25 –20 –15 –10 –5 0 5 10 15 20
Distance from TBM2 (EB) axis: m
(b)

Figure 13. Measured incremental surface ground settlement at

extensometer reference heads and surface monitoring points for:
(a) first TBM (west-bound); (b) second TBM (east-bound) (note:
negative values of vertical displacement indicate settlement)

Dunnicliff is also greatly appreciated. The support provided by work. Special thanks are reserved for the experienced crew of
the Royal Parks and Westminster Council during the installation drillers, especially Mr Ian Shortland, whose vast experience of
work is gratefully acknowledged. It would not be possible to drilling in the London Basin proved to be essential in achieving
achieve the successful Hyde Park instrumentation without the high-quality installations.
diligent work by the contractor, Fugro Engineering Services, and
specialist sub-contractors Construction Monitoring Control Sys- REFERENCES
tems (CMCS) Limited and itmsoil. The engineer, Arup Atkins JV, Bommer JJ and Alarcón JE (2006) The prediction and use of peak
in particular Dr Vicky Hope and Mr Paul Braddish, made ground velocity. Journal of Earthquake Engineering 10(1):
enormous efforts in both the procurement and supervision of the 1–31.

15
Geotechnical Engineering Lessons learnt from installation of field
instrumentation
Wan and Standing

BSI (1990a) BS1377: British standard methods of test for soils for Geotechnical News 26(2): 38–40.
civil engineeing purposes, Part 6: Consolidation and Mikkelsen PE (2002) Cement-bentonite grout backfill for borehole
permeability tests in hydraulic cells and with pore pressure instruments. Geotechnical News 20(4): 38–42.
measurements. BSI, London, UK. Nyren R (1998) Field Measurements Above Twin Tunnels in
BSI (1990b) BS1377: British standard methods of test for soils for London Clay. PhD thesis, Imperial College, University of
civil engineeing purposes, Part 7: Shear strength tests (total London, London, UK.
stress). BSI, London, UK. O’Rourke MJ and Ayala G (1993) Pipeline damage due to wave
Contreras IA, Grosser AT and ver Strate RH (2008) Geotechnical propagation. ASCE Journal of Geotechnical Engineering
Instrumentation News – The use of the fully-grouted method 119(9): 1490–1498.
for piezometer installation, Part 1 and Part 2. Geotechnical Standing JR and Burland JB (2006) Unexpected tunnelling
News 26(2): 30–37. volume losses in the Westminster area, London.
Dunnicliff J (2008) Geotechnical Instrumentation News – Geotechnique 56(1): 11–26.
Discussion of ‘The use of the fully-grouted method for Vaughan PR (1969) A note on sealing piezometers in boreholes.
piezometer installation’ by Contreras et al. (2008). Géotechnique 19(3): 405–413.

WHAT DO YOU THINK?

To discuss this paper, please email up to 500 words to the
editor at [email protected]. Your contribution will be
forwarded to the author(s) for a reply and, if considered
appropriate by the editorial panel, will be published as a
discussion in a future issue of the journal.
Proceedings journals rely entirely on contributions sent in
by civil engineering professionals, academics and students.
Papers should be 2000–5000 words long (briefing papers
should be 1000–2000 words long), with adequate illustra-
tions and references. You can submit your paper online via
www.icevirtuallibrary.com/content/journals, where you
will also find detailed author guidelines.

16
View publication stats