A feasibility study of the measuring accuracy and capability of wireless sensor networks in tunnel monitoring

Abstract

Fire disasters and the deterioration of tunnel structures are major concerns for tunnel operation and maintenance. Traditional wired monitoring systems have many drawbacks in terms of installation time, overall cost, and flexibility in tunnel environments. In recent years, there has been growing interest in the use of wireless sensor networks (WSNs) for the monitoring of various structural monitoring applications. This paper evaluated the feasibility of applying a WSN in the monitoring of tunnels. The monitoring requirements of tunnels under explosion and combustion fire scenarios are analyzed using numerical simulation, and the maximum possible distance for temperature sensors is derived. The displacement monitoring of tunnels using an inclinometer is investigated. It is recommended that the inclinometer should be installed in the 1/4 span of the tunnel structure. The maximum wireless transmission distances in both outdoor and tunnel environments were examined. The influences of surface materials and sensor node locations on the data transmission distance in tunnel environments were also investigated. The experimental results show that the data loss in tunnel environments is approximately three times that in outdoor environments. Surface material has a considerable influence on the transmission distance of radio signals. The distance is 25 ˜ 28 m for a raw concrete surface, 20 m for a brick surface, and 36 m for a terrazzo surface. The transmission distances along the middle of quarter points are approximately 0.9D (D is the transmission distance in the center of the tunnel), and the relative error is less than±3%. The transmission distances at different locations along the bottom exhibit significant differences, decreasing from the middle to the corner point, with distances of approximately 0.8D at the quarter points and minimum distances of approximately 0.55D at the corner points.

Keywords

wireless sensor network (WSN) / tunnel / monitoring / feasibility study

Introduction

Tunnels have been used extensively in developing urban underground spaces for transportation and utility networks. Fire disasters and deterioration of tunnel structures are major concerns for tunnel operation and maintenance. A real-time monitoring system can be installed to ensure the safety of tunnel structures and operations. Conventionally, these systems have been wired systems, with the wiring constituting a considerable proportion of the installation cost. According to bridge engineering estimates, the installation of wires may cost as much as 25 percent of the overall budget and account for 75 percent of the installation time, on average [1,2]. For example, the cost of installing over 350 sensing channels on the Tsin Ma suspension bridge in Hong Kong is estimated to have exceeded $8 million [3]. As access time and engineering hours are severely limited in tunnel environments, the disadvantage of installing a wired monitoring system is particularly significant [4].

Like conventional systems, wireless sensor networks (WSNs) involve sensors and data loggers [5]. However, the wires and associated installation costs are replaced with radio connections, which could result in significant cost savings. In recent years, there has been growing interest in the use of WSNs for monitoring various structural monitoring applications [6-8]. Prototypes of WSNs have recently been tested in tunnels. For example, the use of frequency diversity (FD) at 2.4 GHz was investigated to improve link performance for WSN deployment in tunnels [9,10]. WSNs have been installed for structure performance monitoring in a London Underground tunnel of the Jubilee Line [11,12] and for disaster prevention monitoring in Snow Mountain Tunnel, Taiwan [13]. Miniature micro-electro-mechanical sensors (MEMSs) have also recently been introduced for the performance monitoring of underground structures [12]. Because the scale of WSN testing in tunnels is relatively small, field testing will continue to be an important validation environment for assessing the true merits and limitations of WSNs as the technology matures.

To evaluate the feasibility of applying WSNs in the monitoring of tunnels, the fire monitoring requirements in tunnels were first analyzed. Then, the displacement monitoring of tunnels using an inclinometer was investigated, and an indoor experiment was conducted to derive the accuracy of wireless inclinometer sensors. Finally, this paper examined the maximum wireless transmission distance in both outdoor and tunnel environments.

Wireless sensor network

A WSN is a self-organized wireless network composed of a large number of sensor nodes that interact with the physical world. Various low-power and cost-effective sensor platforms have been developed based upon recent advances in wireless communication and microsystem technologies.

The hardware used in this paper is a commercial off-the-shelf (COTS) WSN system manufactured by Crossbow Technology Inc. It consists of a USB base station board (MIB520), a processor and radio board (MPR2400CA), a sensor board (MDA300CA) with embedded temperature and humidity sensors, a tilt sensor (CXTLA02), and a Linear Variable Differential Transformer (LVDT) displacement sensor (DCTH4000) manufactured by RDP Electronics Ltd. The WSN hardware is shown in Fig. 1.

The WSN assembled in this study uses a low-power, low-data-rate, short-range communication protocol known as ZigBee, which is based on the IEEE802.15.4 standard. The protocol allows for confirmation messages to be sent; a sensor will retransmit the data if they are not successfully communicated, ensuring a robust system. The WSN operates in the 2.4 GHz ISM band, which has the advantage that it does not require a license.

Study of the tunnel monitoring method

Displacement monitoring method using an inclinometer

The employment of traditional displacement monitoring methods, such as displacement gauges, in tunnels is prohibitive because the displacement gauge may interfere with tunnel operation. To circumvent this problem, displacement monitoring using an inclinometer is investigated. For a rectangular-shaped tunnel, as shown in Fig. 2(a), the relationship between the displacement and the deflection of the structure can be obtained using material mechanics:

(1)w_{\max }={\displaystyle \frac{{\displaystyle \frac{\mathrm{5}{l}^{\mathrm{4}}}{\mathrm{384}}}-{\displaystyle \frac{l^{\mathrm{4}}}{\mathrm{54}\sqrt{\mathrm{3}}}}}{{\displaystyle \frac{\mathrm{1}}{\mathrm{4}}}l{x}^{\mathrm{2}}-{\displaystyle \frac{\mathrm{1}}{\mathrm{6}}}{x}^{\mathrm{3}}+{\displaystyle \frac{\mathrm{1}}{\mathrm{12}}}{l}^{\mathrm{2}}x-{\displaystyle \frac{\mathrm{1}}{\mathrm{12}}}{l}^{\mathrm{3}}}}\theta ,

where x is the distance between the inclinometer and the abutment, l is the span of the structure, q is the inclination angle at location x, and w_max is the maximum displacement.

A loading experiment on a 2.2-m concrete beam is conducted to verify the displacement calculated from Eq. (1) and that obtained from structure responses, as shown in Fig. 3. The inclinometer is installed at the location that has a distance of 0.35, 0.4, 0.45, and 0.55 m to the center. The calculated and measured displacements are shown in Fig. 4.

It can be observed from Figs. 4(a)-(d) that the error between the calculated and measured displacement is minimized when the inclinometer is installed on the 1/4 span of the structure (0.55 m to the center). For this case, the relative error varies between 2.2 and 8.9%. Therefore, displacements of a tunnel can be deduced from the monitoring results of inclinometers. It is recommended that the inclinometer be installed in the 1/4 span of the tunnel structure, not in the center of the structure’s span. When x = 1/4l, Eq. (1) can be rewritten as:

(2)\mathrm{\Delta }{w}_{\max }=({\displaystyle \frac{\mathrm{5}}{\mathrm{19}}}-{\displaystyle \frac{\mathrm{64}}{\mathrm{171}\sqrt{\mathrm{3}}}})l\cdot \mathrm{\Delta }\theta =\mathrm{0.0471}l\cdot \mathrm{\Delta }\theta ,

Fire monitoring method based on numerical simulation

To obtain the minimum accuracy requirement for temperature sensors and guide the installation of temperature sensors in tunnels for fire disaster monitoring, a numerical model of a fire disaster in a tunnel is established in Smart-Fire software to simulate the temperature distribution. The tunnel is 7.5 m in width, 3.7 m in height and 100 m in length, and the fire is 30 m from the end of the tunnel. The simulated fire is 5 MW and is 4.5 m in length by 2.0 m in width and 1.5 m in height. The numerical model of the tunnel fire simulation is shown in Fig. 5.

The fire disaster is calculated in 1) explosion mode, where Q = 5 MW when t = 0, wind speed= 0 m/s; 2) explosion mode, where Q = 5 MW when t = 0, wind speed= 1.5 m/s; 3) combustion mode, where Q = 0.188t², wind speed= 0 m/s; and 4) combustion mode, where Q = 0.188t², wind speed= 1.5 m/s. The temperature change results are shown in Fig. 6.

The tunnel temperature changes faster in explosion mode than in combustion mode. In explosion mode, the temperature above the fire is approximately 510°C after 5 s, while the corresponding temperature is only 90°C after 60 s in combustion mode, as shown in Figs. 6 (a) and (c).

Wind speed in the tunnel has a major influence on the temperature distribution. The temperature curve is asymmetric when the wind speed is greater than zero, as shown in Fig. 6(b) and (d). In fire combustion mode, the temperature distribution range after 45 s is within 40 m with a wind speed of 0 m/s, while the distribution range is 70 m with a wind speed of 1.5 m/s. However, the highest temperature decreases from 55 to 10°C when the wind speed increases from 0 to 1.5 m/s. According to the numerical simulation results, the relationships between the accuracy of the sensor, the maximum distance of the sensors, and the alarm time are listed in Table 1.

The fire monitoring requirements can be derived from Table 1. For example, for an alarm time of less than or equal to 15 s in explosion mode, the maximum distance of the sensors is 50s in explosion mode, the maximum distance of the sensors is 50-60 m, given that the accuracy of the temperature sensor is 1°C. For an alarm time of<= 60 s in combustion mode, the maximum distance of the sensors is 2 , 3-5 , 10-20, and 20-40 m, given that the accuracy of the temperature sensor is 1, 0.5, 0.3, and 0.1°C, respectively.

To help people escape from a tunnel fire as quickly as possible, it is vital to deploy the wireless temperature sensors along the cross section of the tunnels. The numerical simulation results show that the temperature changes much faster at the tunnel crown than at the other points, as shown in Figs. 7(a) and (b). It is recommended that wireless temperature sensors be installed in the center of the tunnel structure.

To summarize, tunnel monitoring methods for deformation and fire based on WSNs are feasible. The wireless inclinometer and temperature sensors should be installed in the 1/4 and 1/2 span of the tunnel structure as often as possible.

Experimental studies

Data transmission distance for outdoor and tunnel conditions

To guide the installation of wireless sensors in tunnels, the maximum wireless transmission distance in both outdoor and tunnel environments is studied. The outdoor and tunnel environment is shown in Fig. 8. The tunnel is approximately 7.5 m in width by 3.7 m in height by 100 m in length.

The wireless data transmission loss, which is defined as data received divided by data transmitted, can be approximated with Eq. (3).

(3)L=A{(X/\mathrm{100}\lambda )}^{\mathrm{2}}+B,

where L is the percentage of data loss, X is the distance between data transmitter and receiver, l is the wavelength of the transmission radio, and A and B are coefficients to be determined.

The experimental results are shown in Fig. 9. The A and B coefficients in Eq. (3) for outdoor environments are 0.0365 and -0.87319, respectively. The A and B coefficients for tunnel environments are 0.104 and -0.29946, respectively. It can be observed from Fig. 9 that the data transmission distance is 60 and 20 m for outdoor and tunnel environments if there is no data loss. The distances are 65 and 25 m for outdoor and tunnel environments if the data loss is less than 10%. Therefore, the data loss in tunnel environments is approximately three times the data loss in outdoor environments. This means that the maximum distance between wireless sensors in tunnel environments is only 1/3 of the distance in outdoor environments.

The influence of surface materials on the data transmission distance in tunnels

It is common to decorate tunnel surfaces with different materials. The influences of the surface material on the data transmission distance of radio signals are investigated. The experiments were conducted in tunnels with five different surface materials – lime, terrazzo, asphalt, brick, and concrete, as shown in Fig. 10. Data transmission distances in the tunnels with the five surface materials are shown in Fig. 11.

It can be observed from Fig. 11 that the surface material has a considerable influence on the transmission distance of radio signals, with a minimum distance of 20 m in brick environments and a maximum distance of 36 m in terrazzo environments. The transmission distance in tunnels with a raw concrete surface is 25 - 28 m, which lies in the middle of the ranges for the five surface materials in this experiment. Using Eq. (3), the A and B coefficients are derived and shown in Table 3.

Influences of sensor location on the data transmission distance

To study the influences of sensor location on radio transmission distance along the tunnel, sensor nodes are deployed on nine locations (from location 0 to location 8), as shown in Fig. 12. Locations 1 to 8 are all on the tunnel walls. In each case, the antenna of the sensor node is perpendicular to the wall, as shown in Fig. 12. Location 0 is in the center of the tunnel, which is designed to provide a basis for comparison.

For a given data loss percentage value of 25%, the transmission distance is 28 m in location 0, denoted as D. The data transmission distances at locations 1, 2, 3 and 6 are 24.5, 26, 25, and 25 m, respectively. The data transmission distances at locations 4, 5, 7 and 8 are 15.5, 22.5, 22, and 16 m, respectively.

It can be concluded from the results that the variation in the transmission distance along the middle of each side is very small. The transmission distances are approximately 0.9D, and the relative error is less than±3%. Transmission distances at five locations along the bottom exhibit significant differences. The ratios of the five distances to D are shown in Fig. 13. It can be observed that the transmission distances decrease from the middle to the corner point, with distances of approximately 0.8D at the quarter points and minimum distances of approximately 0.55D at the corner points. Therefore, installation of sensor nodes at corner points should be avoided as much as possible. The installation of sensor nodes at quarter points is optimal in terms of inclinometer monitoring and the data transmission distance of sensor nodes.

Conclusions

This paper evaluated the feasibility of applying WSNs in the monitoring of tunnels. The monitoring requirements of tunnels under explosion and combustion scenarios are analyzed using numerical simulation, and the maximum distance of temperature sensors is derived. The displacement monitoring of tunnels using inclinometers is investigated. It is recommended that the inclinometers be installed in the 1/4 span of the tunnel structure, not in the center.

The maximum wireless transmission distances in both outdoor and tunnel environments were examined. The influences of the surface materials and the sensor node locations on the data transmission distance in tunnel environments were also investigated. The experimental results show that the data loss in tunnel environments is approximately three times that in outdoor environments. The surface material has a considerable influence on the transmission distance of radio signals. The distance is 25 - 28 m for raw concrete surfaces, 20 m for brick surfaces, and 36 m for terrazzo surfaces. The transmission distances along the middle of quarter points are approximately 0.9D, and the relative error is less than±3%. The transmission distances at different locations along the bottom exhibit significant differences, decreasing from the middle to the corner, with distances of approximately 0.8D at quarter points and minimum distances of approximately 0.55D at corner points. The installation of sensor nodes at corner points should be avoided as much as possible. The installation of sensor nodes at quarter points is optimal in terms of inclinometer monitoring and the data transmission distance of sensor nodes.

References

[1] Lynch J P, Loh K J. A Summary Review of Wireless Sensors and Sensor Networks for Structural Health Monitoring. Shock Vib, 2006, 38(2): 91-128

[2] Lynch J P. An overview of wireless structural health monitoring for civil structures. Mathematical and Physical Sciences, 2007, 365(1851): 345-372

[3] Farrar C R. Historical Overview of Structural Health Monitoring. Lecture Notes on Structural Health Monitoring Using Statistical Pattern Recognition, Los Alamos Dynamics, Los Alamos, NM, 2001

[4] Bennett P J, Kobayashi Y, Soga K, . Wireless sensor network for monitoring transport tunnels. Geotechnical Engineering, 2010, 163(3): 147-156

[5] Chintalapudi K, Fu T, Paek J, . Monitoring civil structures with a wireless sensor network. IEEE Internet Computing, 2006, 10(2): 26-34

[6] Cho S, Yun C B, Lynch J P, . Smart wireless sensor technology for structural health monitoring of civil structures. Steel Structures, 2008, 8: 267-275

[7] Feltrin G, Meyer J, Bischoff R. Wireless sensor networks for long term monitoring of civil structure. In: Proceedings of the second International Conference on Experimental Vibration Analysis for Civil Engineering Structures. Porto, Portugal, 2007, 95-111

[8] Hirai C, Soga K. An experimental model of relay deployment planning tool for a wireless sensor networks system to monitor civil engineering structure. In: Proceedings of the 9th IASTED International Conference on Parallel and Distributed Computing and Networks (PDCN 2010). Innsbruck, February, 2010

[9] Liu R, Wu Y, Wassell I J, . Frequency Diversity Measurements at 2.4GHZ for Wireless Sensor Networks Deployment in Tunnels. In: Proceedings of 20th IEEE International Symposium on Personal Indoor Mobile Radio Communications (PIMRC 09). Tokyo, Japan, September, 2009

[10] Wu Y, Wassell I. Investigation of close-to-wall wireless sensors deployment using 2D finite-difference time-domain modeling. In: Proceedings of the 2nd International Conference on Wireless Communications in Underground and Confined Areas (ICWCUCA). Val-d’Or, Quebec, Canada, August, 2008

[11] Stajano F, Hoult N, Wassell I, . Smart bridges, smart tunnels: Transforming wireless sensor networks from research prototypes into robust engineering infrastructure. Ad Hoc Networks, 2010, 8(8): 872-888

[12] Soga K, Chaiyasarn K, Viola F, . Innovation in Monitoring Technologies for Underground Structures. In: Proceedings of 1st International Conference on Information Technology in Geo-Engineering, Shanghai, China, 2010

[13] Chang D T T, Tsai Y S., Lee, H C, . Wireless sensor network(WSN) using in tunnel environmental monitoring for disaster prevention. In: International Conference on Advanced Engineering Materials and Technology. AEMT 2011, 291-294: 3401-3404

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg