Open Access
Add to BibSonomy
Abstract
A depth-discrete groundwater monitoring well is crucial to observing groundwater contamination and subsurface environments. To address this issue, we developed a multilevel monitoring system (MLMS). Because optical fiber sensors are small, have low voltage requirements, and have minimal signal loss over a long distance, we used fiber Bragg grating (FBG) technology to develop a MLMS to observe the depth-discrete aquifer status. The developed FBG sensors and MLMS were examined by a laboratory test and two field tests, respectively. The results show that the FBG piezometer and thermometer accuracies are 0.2% and 0.4% full-scale, respectively. The MLMS can be easily installed in a 2-inch well without a sealing process and can successfully measure the depth-discrete aquifer status at the selected fully-penetrated wells during the two injection events at the study site. The analysis of the collected data and their corresponding injection event reveals the possible structure of the subsurface hydraulic connections at the study sites. These results demonstrate that the FBG MLMS can be an alternative subsurface monitoring system, which has the advantage of a relatively low cost, good data collection efficiency, and environmental sustainability.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Understanding the spatiotemporal distributions of groundwater pressure and temperature is essential to evaluating groundwater resources, such as, among others, groundwater flow and shallow geothermal resources [1–4]. A conventional groundwater monitoring well measures aquifer status, including groundwater pressure, temperature, pH, and electrical conductivity) by opening a screen interval on the casing at a specific depth in a single borehole. The sensors (e.g., piezometer and thermometer) can measure the aquifer status at a depth of the well screen [5]. The measurements of aquifer status from these conventional wells represent an average value within the screen interval [6]. The method of constructing this type of monitoring well is well-developed and broadly used in subsurface environmental investigations [7], including groundwater resource monitoring, contaminated site remediation, seawater intrusion, geothermal energy exchange. [8,9]. For some complicated groundwater systems, developing a well monitoring system to observe the depth-discrete aquifer status and collect water samples is essential. Moreover, a depth-discrete monitoring system allows the monitoring of the fracture flow in bedrock [10] for single or multiple fracture systems at the depths of interest, including sampling, level measurements, and hydraulic tests. Accordingly, a depth-discrete monitoring system provides a more complete and detailed groundwater delineation than the traditional full-screened well system.
Besides providing a long-term groundwater monitoring capability, a depth-discrete monitoring system is more efficient and economical than a traditional well to collect detailed aquifer status (e.g., groundwater pressure, temperature, and concentrations) during the sequential hydraulic tests, known as hydraulic and tracer tomography [11,12]. These tomography tests usually require a group of wells to collect the pressure/concentration dataset during the sequential pumping or injection events (short-term monitoring). These collected datasets are then analyzed by geostatistical methods to characterize the heterogeneous hydraulic properties for the complicated sediment/fracture aquifer system. Because a depth-discrete monitoring well can measure the aquifer status at various depths during a test, the size of the dataset (e.g., pressure) collected from a depth-discrete monitoring well group is much greater than that collected from the traditional well group; therefore, a finer-resolution tomographic survey can be conducted. In other words, using depth-discrete monitoring well group for tomography tests leads to higher spatial resolution and more accurate three-dimensional hydraulic properties than using traditional wells.
There are two conventional approaches to monitoring the depth-discrete aquifer status [13,14]: (1) a monitoring well cluster, which consists of several traditional monitoring wells spaced alongside; the screens of these wells must be placed at the depth interval corresponding to their target aquifers; (2) a nested well, which assembles multiple well casings in a single borehole; the screen of each well casing is placed at its target aquifer. Also, the vertical hydraulic connection between any of the two well casings is interrupted by a grout seal. These two types of well groups can successfully observe the depth-discrete aquifer status of the multi-aquifer system. However, constructing a well cluster is costly and time-consuming, and developing a nested well poses the risk of failure in interrupting the vertical hydraulic connection as the number of monitoring zones increases. Accordingly, a more economical and efficient monitoring method, the multilevel monitoring system (MLMS), has been developed to satisfy the requirement of monitoring the multi-depth discrete aquifer status for complicated aquifer systems.
A MLMS is usually designed for accurate depth-discrete groundwater pressure measurement and water quality sampling. A MLMS well usually consists of several ports and packers used to install sensors and interrupt the vertical hydraulic connection between any of the two monitoring zones, respectively. Because a MLMS has less borehole drilling than a cluster/nested well, the cost, time, and carbon footprint can be much reduced for depth-discrete monitoring. Several commercial MLMS have been developed, such as the Waterloo system and the continuous multichannel tubing (CMT) system from Solinst Canada Ltd., the Westbay system from Westbay Instruments, and the FLUTe system from Flexible Liner Underground Technologies [13]. A CMT system consists of 7 tubes, which can be placed at the target depth, and sensors (maximum up to 7 sensors) installed to measure the groundwater pressure. Because the tube is narrow, only a small size sensor (diameter < = 1 cm) can be installed in the CMT. The vertical hydraulic connections in the CMT well are disrupted by the grout seal (tubes are fixed in the well). CMT needs a 10 cm well for installation, and the diameter itself is 4 cm. This system is easy to maintain because a damaged sensor can be replaced by a new one. Westbay system employs a packer system for depth-discrete measurements and can be installed in the 76 mm to 240 mm well. This system has been installed into a 131 m-deep well and includes 36 ports (monitoring zones) [15]. The pressure transducers can be installed in each port to monitor groundwater pressure. FLUTe has been installed in 150 m-deep wells, and the well diameter ranges from 96 to 145 mm. Each monitoring well includes 12–15 monitoring intervals (custom designed) [16]. Overall, most of these systems integrate several piezometers and operation components (e.g., valve, tube, pumper, and samplers) placed downhole to observe the aquifer status and sample groundwater at the depth of interest. These commercial MLMS also use transducers to convert analog signals into digital signals to maintain the measurement quality for deep monitoring wells.
Currently, optical fiber sensing technology is mature and has been used to develop a variety of environmental sensors [17–19]. Compared to the electronic sensors, optical fiber sensors have several attractive advantages, such as small size, lower weight, the possibility of spectral measurements, high acid and alkali resistance, a waterproof nature, immunity to electromagnetic interference, a low voltage requirement, no explosion risk, a serial multiplexing capability (i.e., several sensors share one fiber and use the same light source), and minimal signal loss over a long distance (no transducer requirement). Three types of optical fiber technologies have been used for environmental sensor development. The first type is the optical fiber sensors developed based on Fabry-Pérot Interferometry (FP) [20–22]. These studies employed FP sensors with CMT to conduct hydraulic tomography. They used CMT tubes to connect the depths of interest to the land surface; thus, the FP sensors can be placed near the water table. This system can measure small pressure variations and thus can characterize the hydraulic properties using a low pumping rate. Moreover, this system is easy to maintain (e.g., a damaged sensor is easily replaced with a new one), and the dynamic range requirement of the interrogator is relatively low. However, the primary drawback of the FP piezometer is that the measurement accuracy is sensitive to the temperature and stability of the light source. A most common method to maintain measurement accuracy is to conduct a field experiment in a short duration. Accordingly, a suitable temperature compensation method and a stable light source for FP sensors are necessary for a stable and long-term measurement [23].
The second optical fiber technique is fiber Bragg grating (FBG) technology. FBG sensors have been used in various fields, such as landslide monitoring [18], ground settlement monitoring [24], structural health monitoring [25,26], instant pollutant detection [27], and temperature and pressure monitoring [18,28]. Moreover, several studies have employed FBG piezometers to measure groundwater pressure under various environments, such as marine sediment [29], soil slope [18], alluvial fan area with subsidence [30], and in a large-scale dike [31]. For accurate field pressure measurement, the FBG pressure sensors are usually developed by packaging method (e.g. packaging a FBG in polymer [32]) or by elastic elements method (e.g. using elastic cylinders [33], cantilevers [34], or metal diaphragms [18]). Moreover, these FBG pressure sensors require temperature compensation to improve measurement accuracy because they are sensitive to both temperature and strain [35]. In this study, we employed metal diaphragms method [18] to developed our FBG pressure sensor, and the temperature compensation method is introduced in the section 2.3.
The third technique is the optical fiber-based technology [36], such as distributed temperature sensing (DTS) and distributed acoustic sensing (DAS). DTS has been used for temperature measurement based on the Raman scattering effect [37]. The temperature data from DTS are calculated by the ratio of Stokes scattering (i.e., insensitive to the temperature) and anti-Stokes scattering, which is exponentially correlated to the temperature [38]. The temperature readings are spatial averages from discrete sections in the fiber [38,39]. On the other hand, DAS was developed for Rayleigh scattering and provides distributed strain sensing. Currently, DAS is used for real-time monitoring of pipelines, tunnels, wells, roads, railway tracks, and bridges, or in fire and security systems [40].
To properly develop the MLMS without electronic sensors, the FBG sensor is more appropriate than the optical fiber-based technology and FP technology, as a variety of FBG sensors can share the same fiber to measure different physical or chemical quantities, although DTS/DAS can measure only temperature or strain. More importantly, the DTS interrogator is at least four times more expensive than that of the FBG unit [41], while the DAS interrogator is even more expensive than the DTS unit. Accordingly, this study employed FBG sensors to develop the MLMS.
The FBG sensors are used to develop the piezometer and thermometer, as temperature measurements are becoming increasingly important in groundwater-related studies [42–46]. For example, many studies have used temperature as a natural tracer (i.e., because of minor environmental damage) to (1) investigate the interaction between surface water and groundwater [46], (2) estimate groundwater velocity [41], and (3) be an alternative index to investigate the transport pathway of a remediation agent or plume in the shallow aquifer of a contaminated site once their temperatures are significantly different from the groundwater temperature (e.g., in this study). Temperature monitoring is also essential in bioremediation as higher temperatures tend to enhance the bioremediation efficiency [47]. Also, understanding the temperature profile and groundwater head variation of a local aquifer is essential for ground-coupled heat-pump studies because temperature is a crucial state variable that affects the efficiency of geothermal energy exchange [48–50]. Accordingly, developing an FBG multi-level groundwater monitoring system that can measure depth-discrete groundwater pressure and temperature simultaneously is necessary for efficient subsurface environment investigation.
To efficiently and economically investigate the depth-discrete aquifer status (i.e., groundwater pressure and temperature), FBG technology is employed to develop a relatively economical multi-level sensing system with a simple installation process (no seal requirement) in this study. The proposed system here focuses on the simultaneous measurement of depth-discrete groundwater temperature and pressure, and the developed MLMS system is examined by a laboratory and two field tests. The capability and accuracy of the FBG MLMS for aquifer status investigation are demonstrated in this study.
2. Methods and materials
2.1 Theory of fiber Bragg grating
An FBG is made by a periodic variation of the fiber core refractive index. This periodic variation is formed by exposing a 1- to 20-mm-long segment of single-mode optical fiber to a spatial pattern of ultraviolet light [51,52]. When a wideband light source illuminates the FBG, a fraction of the light is reflected back upon interference by the FBG (Fig. 1). The wavelength of the reflected light, or the Bragg wavelength, $\Delta {\lambda _\textrm{B}}$ is related to the period of the index modulation, $\Lambda $, and the effective fiber core index of refractive, n, as expressed by [53]:
Longitudinal strains within the Bragg grating, ${\varepsilon _B}$, induced by variations in temperature or stress, can cause a change in $\mathrm{\Lambda }$ (Fig. 1) and, thus, a shifting of ${\lambda _B}$, with the following approximate relationships [53]:
where $\Delta T$ is the change of temperature in degree Celsius. ${\textrm{K}_s}$ and ${\textrm{K}_T}$ are strain and temperature sensitivities of FBG sensors, respectively and correlated to FBG’s central wavelength and environmental temperature. For the central wavelengths range from 0.83-1.55 $\mu m$ in room temperature, their corresponding ${\textrm{K}_s}$ and ${\textrm{K}_T}$ range from 0.64 to1.2 $pm/\mu m$ and from 6.8 to 13 $pm/^\circ C$, respectively [35]. In general, ${\textrm{K}_s}$ and ${\textrm{K}_T}$ in Eqs. (2) and (3) can be obtained by a calibration process.The relationship between strain $\varepsilon $ and its corresponding shift of can be defined as
The ${\lambda _B}$ of the FBG sensors developed herein ranged from 1520 to 1570 nm (10−9 m). The returned signal from every FBG sensor carries a unique wavelength ${\lambda _B} + \Delta {\lambda _B}$, making it possible to connect the multiple FBG elements on the same fiber. Most of the silica optical fiber breaks at a strain of 1%, which corresponds to a $\Delta {\lambda _\textrm{B}}$ of approximately 10 nm. Thus, a separation of ${\lambda _B}$ by 8 to 10 nm between two FBG sensors would be sufficient in most cases. Notice that the FBG is partially distributive because only the parts of the optical fiber with FBG are employed as strain sensors, which share the same optical fiber transmission line.
2.2 Theory of FBG piezometer and thermometer
A schematic view of an FBG piezometer is shown in Fig. 2. We employed the FBG to detect the deflection of a metallic diaphragm inside the piezometer caused by changes in pressure against the atmosphere, and the diaphragm's stiffness manipulates the range of the piezometer. Because temperature variation causes a change in FBG wavelength, measurement error from temperature variation occurred in the pressure measurements. This leads to the requirement to develop the relationship between temperature variation and wavelength change to adjust the measured wavelength of the piezometer (see next section). A typical interrogation system can detect a 1 pm (10−12 m) shift of the Bragg wavelengths. An FBG breaks when stretched by a strain equivalent to approximately 8000–10,000 pm in wavelength variation. Depending on the required safety margin, the maximum allowable pressure was designed to correspond to 1000–6000 pm of the FBG wavelength variation. Figure 2 also shows that a steel case covers the piezometer to fix the steel diaphragm and protect the FBG sensor from the additional stretch caused by environmental noise (e.g., lateral earth pressure) or improper operation. The thermometer (Fig. 3) is developed based on Eq. (3) and the constant in this equation can be derived through the calibration process introduced in the next section
2.3 Calibration of FBG sensors
Developing the relationship between the Bragg wavelength and pressure/temperature is essential before an FBG sensor is used to measure temperature/pressure. To develop the relationship for each FBG thermometer, a constant temperature and humidity chamber (MHK-120AK, Taichy Technology Ltd.) is used to calibrate this relationship under a periodic temperature with constant humidity condition. According to Eq. (2), this relationship is linear [36], and the constant coefficient in Eq. (2) can be obtained through linear regression.
To develop the relationship for each FBG piezometer, a sealed chamber that was controlled pneumatically and monitored with a highly accurate pressure gauge was used to calibrate the relationship between the positive pressure and Bragg wavelength. Because this relationship is linear [35], linear regression was employed to develop it. The primary function of the FBG piezometer is to measure positive pressure ${P_P}$; however, temperature variation causes the deviation of the reflected wavelength of the FBG in the piezometer, resulting in the measurement error in ${P_P}$. The total change of the Bragg wavelength from the pressure and temperature change can be expressed as:
where $\Delta {\lambda _{PT}}$ represents the total change of the Bragg wavelength; and $\Delta {\lambda _P}$ and $\Delta {\lambda _T}$ are the change of the Bragg wavelength caused by pressure and temperature, respectively.To address the measurement error, the FBG piezometer is placed in the MHK-120AK to develop the relationship between temperature variation and wavelength change. Thus, $\Delta {\lambda _T}$ can be measured. Accordingly, the true $\Delta {\lambda _P}$ can be obtained by
Note that every FBG piezometer must adjust the measurement error based on the temperature variation (this operation is called temperature compensation), and the derived $\Delta {\lambda _P}$ is used to calibrate the piezometer under the given pressure from the pressure gauge.
2.4 Multi-level sensing system
The developed FBG MLMS is used to measure the depth-discrete groundwater temperature and pressure inside 2-inch full-screened wells. Each component of the MLMS consists of an FBG piezometer and an FBG thermometer (Fig. 4) developed based on the metal diaphragms method [18] (see section 2.2). These FBG sensors are installed inside a stainless-steel tube, in which two holes are created to connect with the groundwater flow. To measure the groundwater pressure and temperature at a specific depth, two 15-cm rubber membranes are placed on the top and bottom of the stainless-steel tube as packers to interrupt the vertical hydraulic connection. These rubber packers can bear, at most, 196.133 kPa of air pressure. The FBG piezometer, the thermometer, the stainless-steel tube, and two rubber packers are defined as a multifunction FBG sensor (MFS, Fig. 4). The MLMS is developed by assembling these MFS using an optical fiber and an air tube, which are covered by a galvanized soft metal tube (Fig. 4) to protect them from damage during the installation process and which can flexibly adapt to the well with PVC pipe twisted due to the long-term effect of lateral soil pressure. To install the MLMS system, the assembled MFS are first placed in the 2-inch well with fully-opened screens. Then an air-source compressor is used to inject air into the rubber packer in each MFS with constant pressure using an air tank. After these two simple steps, the MLMS is set and can measure depth-discrete groundwater pressure and temperature at the selected site. Note that the MLMS is reusable due to the removable rubber packers. This design decreases the risk of failure because repairing any damaged component of the system is possible. Furthermore, the sealing process inside a well is no longer required for this developed MLMS, making the MLMS’s installation process more efficient, economical, and simple, even if it needs to be installed in a relatively deep well with many monitoring zones.
Fig. 4. Schematic of the developed groundwater multi-level monitoring system installed in the study site
2.5 Field test design
2.5.1 Site description
A pollution-control site near an operational factory in northern Taiwan (Fig. 5) was selected to validate the capability of the MLMS system in the field site. 1,2-dichloroethane and vinyl chloride are two major contaminants. In this study, we consider the aquifer system from the ground surface to the depth of 34 m. The hydrogeological structure is complicated. It consists of an unconfined aquifer (0–16 m), a thin aquitard (16–20m), a thin leaky confined aquifer (20–25 m), and an aquitard (25–34 m) [54]. According to the hydraulic tests, the hydraulic conductivity ranges from $3.5 \times {10^{ - 5}}$ to $3.3 \times {10^{ - 2}}$ cm/s (Table 1) [54].
Table 1. Site drilling data and hydraulic property analysis
2.5.1 Water injection events
To validate the capability of the proposed MLMS, two field tests were conducted at the study site to examine the accuracy of FBG sensors and the capability of the developed MLMS, respectively. In the first test, an MFS and a reference electronic piezometer, Solinst Levelogger Model 3001 (range = 100 m; accuracy of 0.05% FS in pressure and 0.05°C in temperature), were placed into the well YEW3 at depths of 16 m and 15.5 m from the ground surface, respectively. The tap water controlled by a pressure valve from the nearby temple was injected into the well with a nearly constant flow rate (3 liters/min) using a rubber tube. The injection point was placed at 5 m below the land surface (the water table was 2.8 m below the land surface). The temperature and pressure were simultaneously measured by these two types of sensors, and their measurements are mutually compared to examine the accuracy of the FBG sensors.
In the second test, six MFS were separately installed in the wells YEW2 and YEW3 at depths of 14, 24, and 34 m. The injection point was placed at a depth of 34 m through a rubber tube, and the tap water was injected at a constant rate of 3 liters/min. All the FBG sensors simultaneously measured the temperature and pressure before, during, and after the test. All the packers were inflated while these sensors started to operate. Because wells YEW2 and YEW3 are full-screened wells without grout seal, the effect of the filter layer on the pressure and temperature variations of the first two monitoring zones can be observed.
3. Results and discussion
3.1 Calibration results for FBG sensors
The calibration curves for the piezometer and thermometer (i.e., one of the nine MFS) are presented in Fig. 6. They show that the wavelength shifts are linearly correlated with the water pressure and temperature. Their R2 values almost reach 1. The laboratory experiment suggests that the FBG piezometer can measure water pressure ranging from 0 to 400 kPa (i.e., 0 to 40.80 mH2O) with a resolution of 0.1 kPa (a change in pressure of 0.1 kPa corresponds to a 1 pm shifting of the FBG wavelength) and an accuracy of 0.2% full-scale (Fig. 6(a)). The FBG piezometer is also calibrated by the MHK-120AK temperature chamber (Fig. 6(b)), and the measured wavelength shift sourced from temperature variations is used to correct the pressure measurement error (see section 2.3). Furthermore, the FBG thermometer can measure temperatures ranging from 0 to 50 °C with a resolution of 0.1°C and an accuracy of 0.4% full-scale (Fig. 6(c)).
Fig. 6. Calibration curves of the developed fiber Bragg grating sensors. (a) Calibration curve for the selected piezometer. (b) The relationship between the temperature variation and wavelength shift of the selected piezometer. This curve is the reference for the piezometer’s temperature compensation. (c) Calibration curve of the selected thermometer.
3.2 Field test results
3.2.1 Comparisons of the measurements between FBG and electronic sensors
In the first field test, we examine the accuracy of the FBG sensors. An electric sensor (Solinst Levelogger) is used as the reference to validate the FBG sensors because its accuracy is one order higher than that of FBG sensors.
Figure 7(a) shows the comparisons of the pressure variations between the FBG and the electric piezometer during the first field test. One can observe that these two pressure time series are almost identical, and the maximum difference between these two curves is approximately 5 cm, which may be caused by the difference in the response time and accuracy between these two types of sensors. More specifically, these two curves reveal that the response time of the FBG piezometer is 5 seconds (i.e., maximum value) slower than that of Solinst.
Figure 7(b) shows the comparisons of the temperature variations between the FBG and the electric thermometer. The results indicate that the two temperature time series are almost identical, and the maximum difference between these two curves is approximately 0.27 ${^\circ}{C}$. This difference may be caused by the difference in the response time and accuracy between these two types of sensors. Moreover, these two curves show that the response time of the FBG thermometer is 60 seconds faster than that of Solinst when the temperature rises or drops 1 ${^\circ}{C}$ in a short time. Overall, this comparison proves that the FBG piezometer and thermometer are practical and reliable.
3.2.2 Expression of groundwater pressure and temperature variations in the study aquifer
In the second test, a cross-well injection test is conducted to examine the performance of the FBG sensor array and packer system of MLMS. The pressure and temperature variations at all the monitoring zones are measured by the FBG sensors. The injection point is arranged at the third monitoring zone (34 m below the land surface) of YEW2. Figures 8(a) to 8(c) show the pressure variations of the three monitoring zones (i.e., at the depths of 14 m, 24 m, and 34 m) at well YEW2, and we observe that the temporal pressure patterns of these zones are similar. The pressures rise quickly at the beginning of the test, followed by a sudden drop, and then raise again (Fig. 8(c)). This type of pressure variation repeats triple times. According to the borehole records, the third monitoring zone locates in an aquitard (i.e., stratum mainly consists of clayey matters), and the pressure changes of the other two zones should not be significantly affected by the injection. However, the pressure patterns of these three zones are similar and proves that the pressure variations primarily caused by the injection can propagate to the other two zones through the filter layer. Besides, the sudden pressure drops in the early time may cause by the break of biofilm in the well screen of the third zone because bioremediation was conducted one year ago. This speculation still needs to be further examined. The pressure variation trend at the second zone, where the deposition mainly consists of fine sand, declines as time passes. This result reveals that some unknown pumping activity may have existed near the study site during the test period.
Fig. 8. Temporal distributions of groundwater pressure in well YEW2 at the depths of (a) 14 m, (b) 24 m, and (c) 34 m, and well YEW3 at the depths of (d) 14 m, (e) 24 m, and (f) 34 m during the second test.
The pressure variations observed in the monitoring zones in well YEW3 (Figs. 8(d)–8(f)) show that their variations do not correlate to the injection event. These results reveal that the third monitoring zone of YEW2 is blocked by the clayey matters. Thus, the pressures from the injection cannot propagate to the monitoring zones of YEW3.
In addition to the pressure variations, we also observe that the temperature variations of the three monitoring zones of YEW2 are similar, and time lags can be observed in Figs. 9(a) to 9(c). The time lag may correlate to the travel distance between the injection point and the monitoring zones through the filter layer. Figures 9(d) to 9(e) show that the temperature variations of the three monitoring zones at YEW3 are different from those of YEW2, which means the temperature variations of YEW3 are not affected by the injection. This is because the distance between the injection point and YEW3 is over 8.2 m, and the injected water is difficult to reach well YEW3 within 50 min because the low soil permeability (Table 1) and low water pressure gradient cause a relatively low groundwater flow velocity.
Fig. 9. Temporal distributions of groundwater temperature in well YEW2 at the depths of (a) 14 m, (b) 24 m, and (c) 34 m, and well YEW3 at the depths of (d) 14 m, (e) 24 m, and (f) 34 m during the second test.
According to the two field test results, we confirm that the measurements from the FBG MLMS are accurate and reasonable. These results support that this new FBG MLMS is practical and has the potential for field groundwater monitoring.
4. Conclusion
This study successfully employed FBG technology to develop a multilevel monitoring system (MLMS) to monitor the spatiotemporal groundwater pressure and temperature in the subsurface environment. This MLMS is more sustainable, efficient, and economical for data collection in a subsurface environment than the traditional well cluster and nested well. In particular, the FBG MLMS can be installed in a 2-inch well without a sealing process. The number of multi-function FBG sensors (MFS) in MLMS can be flexibly adjusted to meet the monitoring requirements without the issues such as waterproofing, signal loss, electromagnetic interference. The laboratory and field tests indicate that the FBG sensors are accurate and practical. The MLMS easily observed the depth-discrete groundwater pressure and temperature in the selected wells during the two injection events. By further analyzing these collected spatiotemporal data and their corresponding stimuli, we obtain a preliminary understanding of the possible subsurface hydraulic connections in the aquifer system in the study site. These results demonstrate that the proposed MLMS provides a reliable subsurface monitoring system that considers cost, measurement accuracy, data collection efficiency, and environmental sustainability. Accordingly, this FBG MLMS has the potential to be the next-generation groundwater monitoring system.
Funding
Ministry of Science and Technology, Taiwan (MOST 108-2116-M-002 -029 -MY3, MOST 108-2811-M-002 -642, MOST 109-2621-M-002 -016); Environmental Protection Administration, Executive Yuan, R.O.C. Taiwan (EPA. 108C003200).
Acknowledgments
The authors thank the support from Apollo Technology CO., LTD, for the field data sharing and DPI injection. The first author thanks the support from CIPTO TECH, for the installation of MLMS. We also wish to thank the anonymous reviewers for their constructive comments and suggestions, which much improved this manuscript.
Disclosures
The authors declare no conflicts of interest.
References
1. G. Ferguson, H. Beltrami, and A. D. Woodbury, “Perturbation of ground surface temperature reconstructions by groundwater flow?” Geophysical research letters 33(13), doi:10.1029/2006GL026634 (2006).
2. C. T. Hsiao, L. C. Chang, J. P. Tsai, and Y. C. Chen, “Features of spatiotemporal groundwater head variation using independent component analysis,” J. Hydrol. 547, 623–637 (2017). [CrossRef]
3. X. Li and F. T.-C. Tsai, “Bayesian model averaging for groundwater head prediction and uncertainty analysis using multimodel and multimethod,” Water resources research 45(9), doi:10.1029/2008WR007488 (2009).
4. C. R. Tiedeman, D. M. Ely, M. C. Hill, and G. M. O’Brien, “A method for evaluating the importance of system state observations to model predictions, with application to the Death Valley regional groundwater flow system,” Water Resources Research 40(12), doi:10.1029/2004WR003313 (2004).
5. A. C. Johnson, T. J. Besien, C. L. Bhardwaj, A. Dixon, D. C. Gooddy, A. H. Haria, and C. White, “Penetration of herbicides to groundwater in an unconfined chalk aquifer following normal soil applications,” J. Contam. Hydrol. 53(1-2), 101–117 (2001). [CrossRef]
6. H. H. Seo, “Modeling the response of groundwater levels in wells to changes in barometric pressure,” Retrospective Theses and Dissertations 12482, doi:10.31274/rtd-180813-13750 (1999).
7. R. A. Freeze and J. A. Cherry, “Groundwater. Prentice-Hall,” Inc., Englewood Cliffs, NJ (1979).
8. P. F. Hudak, “Principles of hydrogeology, “ CRC Press, (2004).
9. K. M. Hiscock, “Hydrogeology: principles and practice,” John Wiley & Sons, (2009).
10. J. A. Cherry and P. E. Johnson, “A Mulitilevel Device For Monitoring In Fractured Rock,” Groundwater Monitoring & Remediation 2(3), 41–44 (1982). [CrossRef]
11. T. C. J. Yeh and S. Liu, “Hydraulic tomography: Development of a new aquifer test method,” Water Resour. Res. 36(8), 2095–2105 (2000). [CrossRef]
12. S. J. Berg and W. A. Illman, “Three-dimensional transient hydraulic tomography in a highly heterogeneous glaciofluvial aquifer-aquitard system,” Water Resour. Res. 47(10), doi:10.1029/2011WR010616 (2011).
13. T. L. Johnson, “A comparison of well nests vs single-well completions,” Ground Water Monitoring And Remediation 3(1), 76–78 (1983). [CrossRef]
14. M. D. Einarson, “Multilevel ground-water monitoring,” Practical handbook of environmental site characterization and ground-water monitoring (CRC Press, 2006), pp. 808–845.
15. J. R. Meyer, B. L. Parker, and J. A. Cherry, “Detailed hydraulic head profiles as essential data for defining hydrogeologic units in layered fractured sedimentary rock,” Environ Geol 56(1), 27–44 (2008). [CrossRef]
16. J. A. Cherry, B. L. Parker, and C. Keller, “A new depth-discrete multilevel monitoring approach for fractured rock,” Groundwater Monitoring & Remediation 27(2), 57–70 (2007). [CrossRef]
17. D. A. Krohn, T. MacDougall, and A. Mendez, “Fiber optic sensors: fundamentals and applications. Bellingham,” WA: Spie Press, (2014).
18. A. B. Huang, J. T. Lee, Y. T. Ho, Y. F. Chiu, and S. Y. Cheng, “Stability monitoring of rainfall-induced deep landslides through pore pressure profile measurements,” Soils and Foundations 52(4), 737–747 (2012). [CrossRef]
19. Y. T. Ho, A. B. Huang, and J. T. Lee, “Development of a fibre Bragg grating sensored ground movement monitoring system,” Meas. Sci. Technol. 17(7), 1733–1740 (2006). [CrossRef]
20. M. Cardiff, W. Barrash, and P.K. Kitanidis, “A field proof-of-concept of aquifer imaging using 3D transient hydraulic tomography with temporarily-emplaced equipment,” Water Resources Research 48(5), doi: 10.1029/2011WR011704 (2012).
21. M. Cardiff, W. Barrash, and P. K. Kitanidis, “Hydraulic conductivity imaging from 3-D transient hydraulic tomography at several pumping/observation densities,” Water Resour. Res. 49(11), 7311–7326 (2013). [CrossRef]
22. D. L. Hochstetler, W. Barrash, C. Leven, M. Cardiff, F. Chidichimo, and P. K. Kitanidis, “Hydraulic Tomography: Continuity and Discontinuity of High-K and Low-K Zones,” Groundwater 54(2), 171–185 (2016). [CrossRef]
23. M. F. Liang, X. Q. Fang, G. Wu, G. Z. Xue, and A. H. W. Li, “A fiber Bragg grating pressure sensor with temperature compensation based on diaphragm-cantilever structure,” Optik 145, 503–512 (2017). [CrossRef]
24. P. Jia and K. Lv, “Optical Fiber Bragg Grating Ground Settlement Monitoring System Based on 4G,” 2015 8th International Symposium on Computational Intelligence and Design (ISCID) 1, IEEE, 169–172 (2015).
25. T. H. Chan, L. Yu, H. Y. Tam, Y. Q. Ni, S. Y. Liu, W. H. Chung, and L. K. Cheng, “Fiber Bragg grating sensors for structural health monitoring of Tsing Ma bridge: Background and experimental observation,” Engineering Structures 28(5), 648–659 (2006). [CrossRef]
26. D. Kinet, P. Mégret, K. W. Goossen, L. Qiu, D. Heider, and C. Caucheteur, “Fiber Bragg grating sensors toward structural health monitoring in composite materials: Challenges and solutions,” Sensors 14(4), 7394–7419 (2014). [CrossRef]
27. C. Caucheteur, T. Guo, and J. Albert, “Review of plasmonic fiber optic biochemical sensors: improving the limit of detection,” Anal. Bioanal. Chem. 407(14), 3883–3897 (2015). [CrossRef]
28. Y. Q. Liu, Z. Y. Guo, Z. G. Liu, D. H. Zhao, C. F. Ge, and X. Y. Dong, “Fiber grating sensor with enhanced pressure and temperature sensitivity,” Chin. Phys. Lett. 17(2), 115–116 (2000). [CrossRef]
29. T. Liu, S. P. Li, H. L. Kou, W. L. Chai, and G. L. Wei, “Excess pore pressure observation in marine sediment based on Fiber Bragg Grating pressure sensor,” Mar. Georesour. Geotechnol. 37(7), 775–782 (2019). [CrossRef]
30. A. B. Huang, J. T. Lee, T. S. Chuang, Y. T. Ho, and C. C. Wang, “Integrated pore water pressure and ground subsidence profile monitoring,” in AGSSEA Conference (2013).
31. L. Schenato, J. P. A. López, A. Galtarossa, A. Pasuto, T. Bogaard, and L. Palmieri, “Design and field testing of a fiber optic pressure sensor for underground water level monitoring,” Proc. SPIE 11199, 11199OJ( 2019). [CrossRef]
32. H. Ahmad, S. W. Harun, W. Y. Chong, M. Z. Zulkifli, M. M. M. Thant, Z. Yusof, and P. Poopalan, “High-sensitivity pressure sensor using a polymer-embeddedFBG,” Microw. Opt. Technol. Lett. 50(1), 60–61 (2008). [CrossRef]
33. B. J. Peng, Y. Zhao, J. Yang, and M. Zhao, “Pressure sensor based on a free elastic cylinder and birefringence effect on an FBG withtemperature-compensation,” Measurement 38(2), 176–180 (2005). [CrossRef]
34. Q. P. Liu, X. G. Qiao, Z. A. Jia, X. Y. Wang, and L. I. Ting, “FBG sensor for stress based on double cantilever beam,” Acta Photonica Sin. 36(9), 1645–1647 (2007).
35. Y. J. Rao, “In-fibre Bragg grating sensors,” Meas. Sci. Technol. 8(4), 355–375 (1997). [CrossRef]
36. J. P. Dakin, “Distributed optical fiber sensors,” Distributed and Multiplexed Fiber Optic Sensors II, 76-108 (1797), International Society for Optics and Photonics, (1993).
37. G. Bolognini, J. Park, M. A. Soto, N. Park, and F. Di Pasquale, “Analysis of distributed temperature sensing based on Raman scattering using OTDR coding and discrete Raman amplification,” Meas. Sci. Technol. 18(10), 3211–3218 (2007). [CrossRef]
38. M. B. Hausner, F. Suárez, K. E. Glander, N. V. D. Giesen, J. S. Selker, and S. W. Tyler, “Calibrating single-ended fiber-optic Raman spectra distributed temperature sensing data,” Sensors 11(11), 10859–10879 (2011). [CrossRef]
39. J. S. Selker, L. Thevenaz, H. Huwald, A. Mallet, W. Luxemburg, N. Van De Giesen, M. Stejskal, J. Zeman, M. Westhoff, and M. B. Parlange, “Distributed fiber-optic temperature sensing for hydrologic systems,” Water Resources Research 42(12), doi:10.1029/2006WR005326 (2006).
40. J. Chesnoy, (Ed.), “Undersea fiber communication systems,” Academic press, (2015).
41. S. Drusová, W. Bakx, A. D. Wexler, and H. L. Offerhaus, “Possibilities for Groundwater Flow Sensing with Fiber Bragg Grating Sensors,” Sensors 19(7), 1730 (2019). [CrossRef]
42. T. I. Coleman, B. L. Parker, C. H. Maldaner, and M. J. Mondanos, “Groundwater flow characterization in a fractured bedrock aquifer using active DTS tests in sealed boreholes,” J. Hydrol. 528, 449–462 (2015). [CrossRef]
43. C. S. Lowry, J. F. Walker, R. J. Hunt, and M. P. Anderson, “Identifying spatial variability of groundwater discharge in a wetland stream using a distributed temperature sensor,” Water resources research 43(10), doi:0.1029/2007WR006145 (2007).
44. K. B. Moffett, S. W. Tyler, T. Torgersen, M. Menon, J. S. Selker, and S. M. Gorelick, “Processes controlling the thermal regime of saltmarsh channel beds,” Environ. Sci. Technol. 42(3), 671–676 (2008). [CrossRef]
45. F. Tauro, J. Selker, N. van de Giesen, T. Abrate, R. Uijlenhoet, M. Porfiri, S. Manfreda, K. Caylor, T. Moramarco, J. Benveniste, G. Ciraolo, L. Estes, A. Domeneghetti, M. T. Perks, C. Corbari, E. Rabiei, G. Ravazzani, H. Bogena, A. Harfouche, L. Brocca, A. Maltese, A. Wickert, A. Tarpanelli, S. Good, J. M. L. Alcala, A. Petroselli, C. Cudennec, T. Blume, R. Hut, and S. Grimaldi, “Measurements and Observations in the XXI century (MOXXI): innovation and multi-disciplinarity to sense the hydrological cycle,” Hydrol. Sci. J. 63(2), 169–196 (2018). [CrossRef]
46. S. W. Tyler, J. S. Selker, M. B. Hausner, C. E. Hatch, T. Torgersen, C. E. Thodal, and S. G. Schladow, “Environmental temperature sensing using Raman spectra DTS fiber-optic methods,” Water Resources Research 45(4), doi:10.1029/2008WR007052 (2009).
47. J. Iqbal, C. Metosh-Dickey, and R. J. Portier, “Temperature effects on bioremediation of PAHs and PCP contaminated south Louisiana soils: a laboratory mesocosm study,” J. Soils Sediments 7(3), 153–158 (2007). [CrossRef]
48. T. Hermans, F. Nguyen, T. Robert, and A. Revil, “Geophysical methods for monitoring temperature changes in shallow low enthalpy geothermal systems,” Energies 7(8), 5083–5118 (2014). [CrossRef]
49. A. McDaniel, D. Fratta, J. M. Tinjum, and D. J. Hart, “Long-term district-scale geothermal exchange borefield monitoring with fiber optic distributed temperature sensing,” Geothermics 72, 193–204 (2018). [CrossRef]
50. N. Diao, Q. Li, and Z. Fang, “Heat transfer in ground heat exchangers with groundwater advection,” Int. J. Therm. Sci. 43(12), 1203–1211 (2004). [CrossRef]
51. K. O. Hill, B. Malo, F. Bilodeau, D. C. Johnson, and J. Albert, “Bragg gratings fabricated in monomode photosensitive optical fiber by UV exposure through a phase mask,” Appl. Phys. Lett. 62(10), 1035–1037 (1993). [CrossRef]
52. G. Meltz, W. H. Glenn, and E. Snitzer. “Distributed, spatially resolving optical fiber strain gauge,” U.S. Patent No. 4,761,073. 2 Aug. (1988).
53. Y-J. Rao, “Fiber Bragg grating sensors: principles and applications,” Optical fiber sensor technology, Springer, Boston, MA, 355–379 (1998).
54. Groundwater Pollution Assessment and Investigation Plan for Zhonggang Basin, technical report, Environmental Protection Administration, Taiwan (2016).
