Key techniques for evaluation of safety monitoring sensors in water conservancy and hydropower engineering

Yan XIANG*, Lin WANG, Zhan-jun WANG, Hui YUAN, Yun-jie GUAN

1. Department of Dam Safety Management, Nanjing Hydraulic Research Institute, Nanjing 210029, P. R. China

2. Xiaolangdi Dam Project Construction and Management Bureau of Ministry of Water Resources, Zhengzhou 450000, P. R. China

3. College of Hydrology and Water Resources, Hohai University, Nanjing 210098, P. R. China

Key techniques for evaluation of safety monitoring sensors in water conservancy and hydropower engineering

Yan XIANG*1, Lin WANG2, Zhan-jun WANG1, Hui YUAN1, Yun-jie GUAN3

1. Department of Dam Safety Management, Nanjing Hydraulic Research Institute, Nanjing 210029, P. R. China

2. Xiaolangdi Dam Project Construction and Management Bureau of Ministry of Water Resources, Zhengzhou 450000, P. R. China

3. College of Hydrology and Water Resources, Hohai University, Nanjing 210098, P. R. China

For the evaluation of construction quality and the verification of the design of water conservancy and hydropower engineering projects, and especially for the control of dam safety operation behavior, safety monitoring sensors are employed in a majority of engineering projects. These sensors are used to monitor the project during the dam construction and operation periods, and play an important role in reservoir safety operation and producing benefits. With the changing of operating environments and run-time of projects, there are some factors affecting the operation and management of projects, such as a certain amount of damaged sensors and instability of the measured data. Therefore, it is urgent to evaluate existing safety monitoring sensors in water conservancy and hydropower engineering projects. However, there are neither standards nor evaluation guidelines at present. Based on engineering practice, this study examined some key techniques for the evaluation of safety monitoring sensors, including the evaluation process of the safety monitoring system, on-site detection methods of two typical pieces of equipment, the differential resistor sensor and vibrating wire sensor, the on-site detection methods of communication cable faults, and a validity test of the sensor measured data. These key techniques were applied in the Xiaolangdi Water Control Project and Xiaoxi Hydropower Project. The results show that the measured data of a majority of sensors are reliable and reasonable, and can reasonably reflect the structural change behavior in the project operating process, indicating that the availabilities of the safety monitoring sensors of the two projects are high.

water conservancy and hydropower engineering; safety monitoring sensor; Xiaolangdi Water Control Project; Xiaoxi Hydropower Project

1 Introduction

At present, China has built about 87 000 reservoirs, and some new water conservancy and hydropower engineering projects are constructed every year. These reservoirs play significant roles in flood control, irrigation, water supply, power generation, and ecological environment improvement. For the evaluation of construction quality and the verification of the design of projects, and especially for the control of dam safety operation behavior, it has been generallyadvised since the 1990s to arrange different amounts of dam safety monitoring sensors when designing new engineering and reinforcement engineering projects. These sensors are used to monitor the project during the construction and operation periods, and play an important role in reservoir safety operation and producing benefits (Wu 2003). With the changing of operating environments and run-time of projects, there are some factors affecting project operation and management, such as a certain amount of damaged sensors and instability of the measured data. In addition, whether the measured data of sensors meet the needs of engineering safety monitoring and whether they truly reflect the real operation behavior of dams directly impact the judgments and decisions of managers. Therefore, systematically evaluating safety monitoring sensors to fully control the monitoring system operation state is an essential element of dam safety management. Liu et al. (2007) took the appraisal of safety monitoring system of the Xiaolangdi Water Control Project on the Yellow River as an example, and discussed the system appraisal frame, the appraisal method, the appraisal software function, and the key technologies for software development. Frangopol et al. (2008) presented an approach for the efficient inclusion of monitoring data in the structural reliability assessment process. Curt and Talon (2011) assessed the quality of data used during dam reviews by using expert knowledge and the ELECTRE TRI method. At present there has been no systematic study on key techniques for evaluation of safety monitoring sensors in water conservancy and hydropower engineering projects. Based on engineering practice, this study systematically examined the evaluation process of a safety monitoring system, field detection methods of differential resistor sensors and vibrating wire sensors, field detection methods of communication cable faults, and a validity test of the monitoring data. Then, two case studies were carried out.

2 Evaluation process of monitoring system

For systematically evaluating the safety monitoring system of water conservancy and hydropower engineering projects, what should be done first is to collect research data on monitoring items and sensors (calibration, installation, and embedment), and to evaluate the rationality and integrity of existing safety monitoring items and the sensor layout based on engineering practice, and then to clarify whether the existing safety monitoring items meet the need of engineering safety monitoring. An automatic monitoring system has been established based on the automation technology. It includes an automatic data acquisition system and an automatic analysis and evaluation system. The data acquisition system is composed of a communication device, collection terminals, peripheral equipment, a data acquisition software, signal and control lines, communication and power lines, and other components. Therefore, the evaluation of the monitoring system also involves the layout and protection of communication lines, the layout of data acquisition devices, and the evaluation of communication modes and the network structure. In addition, validity check and rationality analyses of the monitoring data are especially important in evaluation of the monitoringsystem. A large amount of observational data was accumulated from the construction period to the running period of the project, and the reliability of measured results is critical for accurate evaluation of project operation behavior. We needs not only to evaluate the integrity, reliability, and rationality of the monitoring data, but also to analyze whether the monitoring data reflect the real operation situation of the project based on the design and construction data. The specific evaluation process of the safety monitoring system is shown in Fig. 1.

Fig. 1Evaluation process of safety monitoring system

3 Detection methods of typical monitoring sensors

Field detection of monitoring sensors is a basic task of evaluating the monitoring system. First, the operation situation of dam monitoring sensors must be understood based on the historical data and field observation, and then the data on the sensor layout, original parameters, observed data from sensors, and corresponding design drawings must be collected. In addition, reasons for damage of some sensors are textually researched. Effective secondary instruments certified by measurement (such as a frequency meter, digital reader, digital bridge, multimeter, and megger) are applied to comprehensive field detection of monitoring sensors. This study examined field detection methods of conventional differential resistor sensors and vibrating wire sensors in water conservancy and hydropower engineering projects in detail.

3.1 Differential resistor sensor

A digital bridge is used to measure the resistance and the resistance ratio of differential resistor sensors. Under normal circumstances, the resistance should be the sum of the measured resistance at 0℃ (reading in the card) and the resistance change caused by the temperature change (between 30 ? and 35 ? without the cable resistance), and the resistance ratio should be between 9 500 and 10 500. If the resistance is too high or infinite, it is considered an open circuit; if the resistance is too low or close to zero, it is considered a short circuit or geo-gas anomaly; and if the resistance is within the normal range but there is no reading, it is generally considered sensor failure. The field detection process of the differential resistor sensor is shown in Fig. 2.

Fig. 2Field detection process of differential resistance sensor

3.2 Vibrating wire sensor

As vibrating wire sensors are sealed and embedded in the building, their maintenance and troubleshooting are limited to periodic checks of the cable connections and cleaning up of the cable head (Liu et al. 2007; Zhang and Liu 2008). When detecting circuits with a multimeter (checking the coil resistance), under normal circumstances, the total resistance composed of the coil resistance and the cable resistance can be obtained by looking up the corresponding manufacturer information. Like the GK-4500 osmometer produced by Geokon, under normal circumstances, the coil resistance is 190 ? ± 5 ?, and the resistance of 100-m cable is about 8 ?. If the total resistance is too high or infinite, it is considered an open circuit; if the resistance is too low or close to zero, it is considered a short circuit or geo-gas anomaly; and if the resistance is within the normal range but there is no reading, it is generally considered a sensor failure, which can be solved by appropriate methods to excite the vibrating wire. The field detection process of the vibrating wire sensor is shown in Fig. 3.

4 Detection technology of communication cable

At present, communication cables are commonly used in safety monitoring instruments of hydraulic engineering projects. In general, the reasons for communication cable faults include the quality defects of the cable itself, construction impacts (grounding or disconnection due to cable core splices during construction), external forces or human impact (other projectconstruction, vehicles, and building deformation), and obstacles caused by natural disasters (Ma et al. 2007, 2011). On the basis of fault reasons of communication cables, the faults can be divided into four types according to the property of the line fault: insulation faults, break line faults, mixed-line faults, and grounding faults.

Fig. 3Field detection process of vibrating wire sensor

4.1 Detection steps and methods of cable line fault

The cable line fault detection includes fault property diagnosis, measurement of fault location, and determination of fault points (Dai and Wang 2007). The property and severity degree of the cable line fault are first determined using the megger, multimeter, and other secondary instruments. Then, the cable fault distance is measured using special testing instruments, and the minimum fault range is primarily identified. Finally, based on the measured results of instruments, the minimum fault range is determined, and then the fault point is precisely positioned. Common methods of cable troubleshooting include the bridge method, playback method, leak detection method, pulse reflection method, and comprehensive testing. According to field practice of the safety monitoring equipment of water conservancyand hydropower engineering projects, the pulse reflection method is widely applied in cable fault detection. It mainly consists of sending voltage pulses to the cable and determining the fault point according to the theory that the time difference between the sent impulse and the reflected impulse at the fault point is directly proportional to the distance from the sending pulse point to the fault point.

4.2 Reflected impulse waveform of typical faults

Seven typical faults that produce different reflected impulses with different waveforms are as follows:

(1) Core wire disconnection fault: it produces a positive reflected impulse and the amplitude of the reflected impulse is comparatively large, as shown in Fig. 4(a).

(2) Shielding layer disconnection fault: The reflected impulse at the breaking point of the shielding layer is positive, and its waveform is similar to that of the core wire disconnection fault, as shown in Fig. 4(b).

(3) Induction coil fault: The reflected impulse of the induction coil or poor contract is positive and the amplitude of the reflected impulse waveform is similar to that of the core wire disconnection fault, as shown in Fig. 4(c).

(4) Mixed-line (including mixed with other lines and self-mixing) fault: The negative reflected impulse at the fault point can be seen in the case of a mixed line fault, as shown in Fig. 4(d).

(5) Grounding fault: The negative reflected pulse at the fault point can be seen in the case of grounding fault and its waveform is similar to that of a mixed-line fault, as shown in Fig. 4(e).

(6) Soaking fault: It generally produces a gentle negative reflected impulse whose waveform is shown in Fig. 4(f).

(7) Mismatch fault: There is a positive reflected impulse at a misconnection point, and a negative reflected impulse appears at another misconnection point, as shown in Fig. 4(g).

Fig. 4Reflected pulse waveforms of typical faults

5 Validity check of sensor measured data

Due to issues related with observers, instruments, and a variety of external conditions, there are inevitably errors in the sensors’ original measured data. Therefore, error analysis and validity check of the original measured data should be carried out in order to judge their reliability. The validity check includes: (1) checking the operation method to examine whether the main measurement methods are qualified; (2) determining whether the performance of sensors is stable and normal; (3) checking whether the physical meanings of the measured data are reasonable; and (4) checking consistency, correlation, continuity, and symmetry: continuity implies that a variety of observed data should be continuous changing without jumping when the loading conditions and other external conditions remain constant, and consistency implies that the change trends of continuously accumulated data should be consistent. The measured value check is the most important among the four types of check mentioned above.

The observations under the same condition are called equal-precision observations. By calculating the difference between the observed value and the true value, namely the true error, the reliability of the observed values can be determined. For equal-precision observation sequences, the observation accuracy can be determined by the mean square error (standard deviation) of a complete series of observed values. The calculation formula of the mean square error is expressed as

where δiis the true error of the ith measured value, and n is the number of measured values. However, due to the fact that the true error of the measured values is generally unknown, it is usually replaced by the residual error of measured values. For a column of measured valuesthe residual error can be expressed as

where x is the the arithmetic mean value of the column of measured values. Thus, the mean square error represented by the residual error can be written as

The corresponding error analysis algorithms are programmed to analyze the errors of measured values and evaluate the accuracy and reliability of measured values. According to field detection, the historical process of the measured values and calculated results of mean square errors, combined with the accuracy of observation instruments, instrument measurement range, and corresponding monitoring technical specifications, the corresponding reliability evaluation criteria can be developed. In addition, the reliability analysis of differential resistor sensors include not only the analysis of the mean square error of observations mentionedabove, but also a comprehensive analysis of the errors of the forwardly and reversely measured resistance ratios of the instrument. According to the Technical Specification for Concrete Dam Safety Monitoring (DL/T 5178-2003) (SETCPRC 2003), the forwardly measured resistance ratio Z and reversely measured resistance ratio Z′ of the sensor can be obtained using a bridge, and then, the reliability of measured values can be evaluated by comparing Z + Z′ with 20 000 + A2± 2, where A= (1 0 000 ? Z)100.

6 Case studies

6.1 Xiaolangdi Water Control Project

The Xiaolangdi Water Control Project is a large multi-objective type-I project with the main objectives of flood control (including ice prevention) and sedimentation reduction, and other functions of water supply, irrigation, power generation, storing clean water, and discharging muddy water (Tang et al. 2005). In view of the characteristics of large-scale projects, a large quantity of chambers, and complex geological conditions, many items are monitored in each structure, which form a relatively comprehensive safety monitoring system. Vibrating wire sensors include osmometers, soil displacement meters, earth pressure meters, interface joint meters, strain gauges, joint meters, anchor dynamometers, plate strain meters, measuring weirs, zero stress-strain meters, reinforcement meters, inclinometers, and multi-point extensometers. The field detection methods of potentiometer-type embankment strain meters are similar to those of the vibrating wire sensors. Differential resistance sensors include strain gauges, zero stress-strain meters, osmometers, and anchor dynamometers. Tests were conducted with the detection methods of two types of typical sensors. The results show that, among the 2858 internal observation sensors installed in the Xiaolangdi dam, the mountain on the left bank of the dam, the Xigou dam, the intake tower, the orifice tunnel, the free-flow tunnel, the desilting tunnel, the inlet and outlet slopes, the plunge pool, the underground powerhouse, the spiral case, and the cubital tunnel of the Xiaolangdi Water Control Project, the measured values of 1865 sensors (accounting for 65.26%) is reliable and reasonable, reasonably reflecting the structural change behavior in the project operating process; the measured values of 255 sensors (accounting for 8.92%) are applicable after they are revised; the measured values of 257 sensors (accounting for 8.99%) can provide a reference for analysis; and the remaining 481 sensors (accounting for 16.83%) have been damaged and cannot be used for measurement. This indicates that the availability of the sensors in safety monitoring systems of the Xiaolangdi Water Control Project is high.

6.2 Xiaoxi Hydropower Project

The Xiaoxi Hydropower Project is located in Xinshao County, Shaoyang City, Hunan Province. It is a large multi-objective type-Ⅱ project with the main objective of power generation. The Xiaoxi dam is a concrete gravity dam with a maximum height of 44.5 m. Theproject layout from right to left is: a right bank, an indoor switching station of 110 kV, a run-of-the-river hydropower house, a spillway dam with eight holes, a vertical ship-lift, and a left bank dam. Monitoring items in the project include the external deformation, the foundation settlement, the uplift pressure, the joint change, the steel stress, the stress of anchor bar piles, the compressive stress of concrete, the bypass seepage, seepage of the dam foundation, the bedrock deformation, concrete temperature, and environment factors. According to the detection methods of differential resistance sensors, the safety monitoring system of the Xiaoxi Hydropower Project was evaluated in terms of the dielectric strength, resistance ratio, and resistance. The results show that among the 104 differential resistance sensors of the Xiaoxi Hydropower Project, 88 sensors work in normal conditions, accounting for 84.62%; 12 sensors are qualified, accounting for 11.54%; and four sensors are damaged, accounting for 3.85%. By using the validity check methods of the measured values from sensors, checks were carried out on displacement of the dam body, displacement of the dam foundation, joint change, uplift pressure, water levels upstream and downstream, and concrete temperature. The results show that the monitoring system of the Xiaoxi Hydropower Project is reliable: of the 289 monitoring sensors, the measured values of 232 sensors (accounting for 80.27%) are reliable and reasonable, which reasonably reflects the structural change behavior in the project operating process; 37 sensors (accounting for 12.80%) can provide a reference for analysis; and the remaining 20 sensors (accounting for 6.92%) have been damaged and cannot be used for measurement.

7 Conclusions

In view of the fact that there is no standard for the evaluation of safety monitoring systems, this study examined some key techniques for evaluation of safety monitoring sensors based on engineering practice. The evaluation process of the safety monitoring system was presented. The field detection methods of differential resistance sensors and vibrating wire sensors were developed. The main reasons for communication cable line faults were analyzed, and the test steps and methods of cable line faults were presented, which play an important role in ensuring the reliability of the monitoring system. There are inevitable errors in original observation data. The validity check methods of sensor measured data were put forward in order to evaluate the reliability of the original measured data. These techniques for evaluation of safety monitoring sensors were applied in the Xiaolangdi Water Control Project and the Xiaoxi Hydropower Project. The results show that the measured values of a majority of sensors are reliable and reasonable, and can reasonably reflect the structural change behavior in the project operating process, indicating that the availabilities of the safety monitoring sensors of the two projects are high.

Curt, C., and Talon, A. 2011. Assessment and control of the quality of data used during dam reviews by usingexpert knowledge and the ELECTRE TRI method. Journal of Computing in Civil Engineering, 1-25. [doi:10.1061/(ASCE)CP.1943-5487.0000187]

Dai, Q., and Wang, L. 2007. Research on the cable detecting technology of the evaluation of the monitoring instruments. Journal of Hydraulic Engineering, 38(s1), S671-S674. (in Chinese)

Frangopol, D. M., Strauss, A., and Kim, S. 2008. Bridge reliability assessment based on monitoring. Journal of Bridge Engineering, 13(3), 258-270. [doi:10.1061/(ASCE)1084-0702(2008)13:3 (258)]

Liu, C. D., Ma, F. H., Xiang, Y., and Dai, Q. 2007. Appraisal of safety monitoring system of dam and development of the software. Journal of Hydraulic Engineering, 38(s1), S214-S219. (in Chinese)

Ma, F. H., Xiang, Y., and Liu, C. D. 2007. Appraisal of Safety Monitoring System for Xiaolangdi Water Control Project. Nanjing: Nanjing Hydraulic Research Institute. (in Chinese)

Ma, F. H., Xiang, Y., and Liu, C. D. 2011. Appraisal of Safety Monitoring System for Xiaoxi Hydropower Project. Nanjing: Nanjing Hydraulic Research Institute. (in Chinese)

State Economic and Trade Commission of the People’s Republic of China (SETCPRC). 2003. Technical Specification for Concrete Dam Safety Monitoring (DL/T 5178-2003). Beijing: China Power Press. (in Chinese)

Tang, H. H., Li, M. A., and Qi, Z. F. 2005. Practice of initial operation and management of Xiaolangdi Multipurpose Project. Water Power, 31(12), 83-85. (in Chinese)

Wu, Z. R. 2003. Safety Monitoring Theory of Hydraulic Structures and its Application. Beijing: Higher Education Press. (in Chinese)

Zhang, H., and Liu, C. D. 2008. Study on appraisal technique of safety monitoring system for Xiaolangdi Project on the Yellow River. Hydro-Science and Engineering, 30(3), 75-80. (in Chinese)

(Edited by Yan LEI)

This work was supported by the National Natural Science Foundation of China (Grants No. 51179108 and 50909066) and the Key Research Foundation of Nanjing Hydraulic Research Institute (Grant No. Y711007).

*Corresponding author (e-mail: yxiang@nhri.cn)

Received Sep. 9, 2011; accepted Jan. 20, 2012