Wireless Sensor Networks for

Civil Infrastructure Monitoring

ISBN 978-0-7277-6151-4

ICE Publishing: All rights reserved

http://dx.doi.org/10.1680/wsncim.61514.161

Chapter 22
Tottenham Court Road station (UK)

This case study presents a novel method to monitor the real performance of temporary timber
structures in underground construction using wireless sensor network (WSN) technology.
The information and advice provided herein have been distilled from Xu et al. (2014,
2015).

22.1. Description
22.1.1 Asset
The asset comprises underground temporary timber structures.

22.1.2 Background
Timber structures are commonly used for underground temporary works. These offer a suit-
able supporting system to retain the ground in place during tunnelling and excavation.
As the loadings on these structures have not yet been thoroughly understood, especially for
structures that use a mixture of steel and timber, their design is often conservative and over-
engineered, resulting in health and safety risks to workers. Determining the real performance
of these structures would therefore be beneficial not only to reduce these risks but also to
reduce construction cost and time.

This problem has been investigated at Tottenham Court Road (TCR) station. The existing
facilities are currently being upgraded in order to accommodate the expected rise in passen-
ger flow interchanging between London Underground services and Crossrail in 2018. The
construction of the section being investigated was started in 2013, and involves a new access
passage from stair 14 to the Central Line platforms. Reinforced concrete and steel frames
together with timber were employed to support the ground after excavation. Grouting was
carried out immediately after the temporary works were finished.

22.1.3 Monitoring objectives

The objective was to monitor temporary works at TCR station during the stair 14 excavation.

22.2. Solution
Figure 22.1(a) illustrates the WSN deployment at TCR stair 14 to monitor the temporary
construction works. Currently, the WSN is composed of four SmartPlanks (Xu et al., 2014)
(i.e. timber planks (cross-section 225 mm × 100 mm) that have been instrumented with sen-
sors, Figure 22.1(b)), seven relay nodes (Figure 22.1(c)) and a data sink (Figure 22.1(d)). The
data sink is a low-power computer connected to a mobile phone 3G modem, which allows
the monitoring data to be downloaded to an off-site database server in the University of

161

Downloaded by [ Monash University] on [26/04/19]. Copyright © ICE Publishing, all rights reserved.
 Wireless Sensor Networks for Civil Infrastructure Monitoring

Figure 22.1 Wireless sensor network deployment at Tottenham Court Road station stair 14:
(a) network layout; (b) instrumented plank (SmartPlank); (c) relay node; (d) data sink/gateway (Xu
et al., 2014)

Cambridge. The data sink can also act as data logger, storing the monitoring data locally
when there is poor or no 3G signal on site. Finally, monitoring results are made readily
accessible through a website in a variety of formats.

The SmartPlanks, numbered 51, 52, 53 and 54, were carefully designed and manufactured to
suit the actual site conditions. Each instrumented plank has a 5 mm wire lead strain gauge
(from RS Components) mounted onto the timber surface on the bottom and top sides to
measure both tension and compression. A temperature gauge (MCP9700A) was also attached
alongside each strain gauge for temperature compensation. To avoid the use of inflammable
epoxy adhesives on site, the sensor units were glued onto a mounting plate
(60 mm × 20 mm × 5 mm) of aluminium alloy 6028 specifically designed and manufactured
for the project. Figure 22.2(a) shows the instrumented planks. Proper calibration of the strain
gauge sensor was carried out in the laboratory before actual deployment at the actual site
(Figure 22.2(b)).

The strain gauge units and temperature gauges were interfaced with a wireless sensor node
attached at an available space on the underside of the plank. Each node was powered by
either two AA-sized batteries mounted directly on the printed circuit board (PCB) of the
node or by an external pack of two D-sized lithium-thionyl chloride batteries for an extended
operating lifetime. Nodes also implement radio duty cycling in software in order to prolong
further the battery lifetime.

Wireless sensor nodes installed in the planks were configured to take and transmit strain and
temperature measurements every 15 min for the first week, and then hourly measurements
afterwards.

162

Downloaded by [ Monash University] on [26/04/19]. Copyright © ICE Publishing, all rights reserved.
 Tottenham Court Road station (UK)

Figure 22.2 (a) Instrumentation of SmartPlanks and (b) experimental set-up for calibration of
strain gauges

Monitoring activities were planned to last for 1.5 years until the Central Line possession in
2015. The following issues were encountered:

n Only construction workers were generally allowed to place the planks at the respective
loading position. Unfortunately, the SmartPlank 51 was somehow damaged, and
because it was located in an accessible location, it was unrecoverable and therefore
useless in terms of results. Briefing the people on site regarding how to handle the
SmartPlank would have prevented this problem.
n The planks were wrapped with Duck Ultimate Cloth Tape for protection against any
physical damage. Nevertheless, the PCB in SmartPlank 54 was severely affected by
corrosion caused by unexpectedly high humidity during grouting. Owing to the
construction works, it took a few weeks to replace the damaged component. A
waterproof package was then used for increased protection.
n External antennas with a gain of 5 dBi were used for a better radio communication
range. However, these were often damaged during the excavation works, and protection
was therefore required.
n The 3G signal on site was poor and unstable. The data sink was then used as the data
logger and data were retrieved manually.
n The data delivery rate suffered severe fluctuations, with a maximum value of 80.2%
and a minimum value down to zero. The reasons could be the radio shield within the
steel beams, the number and position of relay nodes, the antenna length, the positioning
and direction, and on-going construction work.

22.3. Monitoring results

A detailed description of the results for the performance of the planks is provided by Xu et
al. (2014). In summary, the temperature on the top surface of the planks rose by up to 33°C
around 10 h after grouting, which is consistent with the classical hydration process of the

163

Downloaded by [ Monash University] on [26/04/19]. Copyright © ICE Publishing, all rights reserved.
 Wireless Sensor Networks for Civil Infrastructure Monitoring

cement type used (Portland cement). The strain measurements showed that the planks experi-
enced a very large compression during grouting, but much less strain on the bottom.

REFERENCES
Xu X, Soga K, Nawaz S et al. (2014) Wireless sensor monitoring of timber structures in under-
ground constructions. In Information Technology in Geo-Engineering, Vol. 3: Advances in Soil
Mechanics and Geotechnical Engineering (Toll DG, Zhu H, Osman A et al. (eds)). IOS Press,
Amsterdam, The Netherlands, pp. 146–153.
Xu X, Soga K, Nawaz S and Gajia M. (2015) Performance monitoring of timber structures in
underground construction using wireless SmartPlank. Smart Structures and Systems. 15(3):
769–785.

164

Downloaded by [ Monash University] on [26/04/19]. Copyright © ICE Publishing, all rights reserved.