Vadose Zone Journal 2:270-275 (2003)
© 2003 Soil Science Society of America
NOTES
Flexible Time Domain Reflectometry Probe for Deep Vadose Zone Monitoring
Ofer Dahan*,a,
Eric V. McDonaldb and
Michael H. Youngc
a Desert Research Institute, Reno, NV, and The Jacob Blaustein Institute for Desert Research, Dep. of Environmental Hydrology & Microbiology, Ben-Gurion University of the Negev, Sede Boqer Campus, 84990, Israel
b Desert Research Institute, Reno, NV
c Desert Research Institute, Las Vegas, NV
* Corresponding author (odahan{at}bgumail.bgu.ac.il)
1 A patent is pending on the system described in this manuscript. 
Received 9 September 2002.
 |
ABSTRACT
|
|---|
Accurate determination of water content is an important aspect of most vadose zone monitoring programs. Real-time, continuous, in situ measurements of water content in relatively undisturbed conditions are usually limited to shallow soil horizons. We present a new methodology using time domain reflectometry (TDR) for water content monitoring in deep vadose zone horizons. The method uses flat, flexible, waveguides pressed against the wall of a borehole. The flexible TDR waveguides are attached to the outer side of a flexible sleeve filled with a liquid resin. The resin (e.g., a two-component urethane) generates hydrostatic pressure that forces the flexible waveguides against the borehole wall, ensuring a close fit to the irregular shape of the borehole walls. The probe can be used with either a standard TDR technique, which uses a cable tester (e.g., Tektronix 1520, Tektronix, Beaverton, OR) for collecting waveforms, or a water content reflectometer (e.g., model CS505, Campbell Scientific, Inc., Logan, UT), which provides a direct electrical output, which may be sampled using a data logger. Laboratory calibration experiments and a full-scale field experiment showed that the method is reliable and capable of providing accurate water content measurements in deep vadose zone horizons.
Abbreviations: PVC, polyvinyl chloride • TDR, time domain reflectometry
|
INTRODUCTION
|
|---|
ACCURATE DETERMINATION of water content is an important aspect of most vadose zone monitoring programs. Real-time, continuous, in situ measurements of water content in relatively undisturbed soil conditions are usually limited to the shallow horizons of the vadose zone, since measurements in deeper horizons present significant difficulties. Deep soil horizons usually are investigated by one-time measurement from a core sample, installation of monitoring equipment through the walls of a wide shaft (Zilberbrand and Gvirtzman, 1996) or by neutron thermalization through a permanently installed borehole (Hillel, 1980).
A common method for measuring soil water content in unsaturated field soils is TDR (Topp et al., 1980; Topp et al., 1984; Ledieu et al., 1986; Herkelrath et al., 1991). In this method, the TDR technique involves measurements of the propagation velocity of a high-frequency signal transmitted along a waveguide. The dielectric constant of water (approximately 80) is significantly higher than most soils (2–7); therefore, measurements of the velocity of propagation can be used to determine the volumetric soil water content. Time domain reflectometry probes are typically made with two or three metal rods (waveguides) placed parallel to each other (Dalton, 1992; Zegelin et al., 1992). One end of each rod is cast into a handle, where it is connected to a coaxial cable; the other end is disconnected so that an open electrical circuit is created. The rods are then inserted into the soil at the point of interest.
The use of TDR probes for measuring water content in the vadose zone often is limited by the probe installation technique. In shallow soil horizons, installation of standard TDR probes is reasonably straightforward. The probes can be inserted into the ground from the surface with only minor disturbance of the natural soil properties. The insertion of TDR waveguides into deep soil horizons, however, can be problematic if a natural undisturbed soil condition is required. Installation of standard TDR probes to deep levels usually involves (i) excavation of a deep trench followed by horizontal insertion of the probes through trench walls to the deep soil horizons (Schwartz and Schick, 1998) or (ii) drilling a large-diameter vertical borehole and using a special instrument to insert the TDR probe through the borehole wall into the surrounding soil (Murdoch et al., 1999).
Both the trench and large-diameter vertical borehole techniques may significantly disturb the natural soil properties in the near-field environment where the TDR probes are installed. This disturbance may create preferential flow paths for water to bypass the natural soil material and affect the representativeness of the TDR reading. Even if the excavated trench or large-diameter borehole is backfilled, generation of preferential flow paths can occur. Preferential flow paths may be generated either through the backfilled material, if the excavated material is used without industrial sealing material such as bentonite or concrete, or along the soil column due to tension release. Tension release in the soil column is a consequence of abrupt reduction of the lithostatic pressure of the soil near the trench or borehole walls. Moreover, standard TDR waveguides are susceptible to bending if installed into rocky or pebbly material. Since their installation usually involves forcing the probe rods into the soil their orientation may end up in a nonparallel position. Deformation of the waveguides away from the parallel configuration can bias the readings and create uncertainties about the accuracy of the water content measurements. Additionally their connecting coaxial cables are also somewhat vulnerable unless armored.
Selker et al. (1993) developed the use of noninvasive TDR probes where the probes are attached to the surface of the examined material without the need to penetrate into it. This probe was made as a stiff acrylic pad that allowed moisture measurements from flat surfaces without the need to penetrate into the material. Although this probe worked well in conditions with relatively smooth surfaces, contact with the soil can present difficulties if the soil surface is not smooth. Recently, E.S.I. Environmental Sensors Inc. (2002) developed a method in which stiff rods bearing multilevel TDR probes are pushed into the ground under very high load. The rods are built to be inserted to depths of up to 3 m and can work well in soft soils.
J.B. Sisson and J. Hubbell at the Idaho National Engineering and Environmental Laboratory (Wyatt et al., 1999) developed a method where a Campbell CS505 probe is encased in an insulation block, and the block is pressed against the borehole using a lever arm. In this method contact of the waveguide with the soil was questionable when used in a hard soil formation when the borehole walls were not smooth enough.
Limitations of existing TDR probe installation techniques in deep soils has led to our development of a new method for TDR probe installation. The new technique permits deep installation of TDR probes with minimal disturbance of the soil column properties. It allows multilevel installation to any desired depth or configuration and provides continuous, real-time, in situ measurements of the moisture profile.
|
MATERIALS AND METHODS
|
|---|
Figure 1 illustrates the methodological concept and general field installation setup.1 The new technique uses a probe consisting of flexible TDR waveguides made of stainless-steel foil attached to the outer side of a sleeve made of flexible liner (e.g., natural gum rubber, polyvinyl chloride [PVC] liner, or polyethylene). The assembly is inserted into a small-diameter (15–20 cm) borehole, drilled at a 45° angle from the vertical (Fig. 1 and 2). Once the flexible sleeve and TDR waveguides are inserted into the borehole, the sleeve is filled with liquid, two-component urethane. Other types of resin (e.g., epoxy, polyester) may be used if heat generation while curing is addressed. Before curing, the urethane generates hydrostatic pressure (urethane density = 1.15 g cm-3), causing the sleeve and waveguides to closely follow the contours of the borehole walls.
The system presented here uses small-diameter sloping boreholes for the installation of the flexible TDR probes. As opposed to vertical boreholes where the entire soil column is likely to be disturbed due to its nearness to the open borehole, the soil column above any given point along the sloping borehole is more likely to be undisturbed (Fig. 1). Nevertheless, the orientation of the borehole, vertical or sloping, is dependent on the experiment characteristics, and the system may be used in vertical boreholes as well as in sloping.
The flexibility of the waveguides considerably improves contact between the TDR waveguide and the borehole walls without the need to insert rods into undisturbed soil. The hydrostatic pressure generated by the liquid urethane in the flexible sleeve generates force normal to the borehole walls, thereby reducing tensional release and partly restoring the natural soil conditions. Although drilling in unconsolidated formations or soils can result in borehole walls with irregular diameters, the flexibility of the TDR waveguides attached to the outer side of the flexible sleeve improves the likelihood that the waveguides will have good contact with the borehole walls. For example, Fig. 3 shows a rubber sleeve with its attached flexible waveguide during excavation from a borehole. It is apparent that the borehole diameter is nonuniform as a result of the drilling procedure (solid stem auger) and that the sleeve and flexible waveguide conformed to the irregular shape of the borehole before the urethane solidified. Although some disturbance of the natural soil properties near the borehole walls (2–3 cm around the borehole) can be expected, the rest of the soil column above it is likely to remain undisturbed if a sloping drilling method is applied (Fig. 1).
View larger version (140K):
[in this window]
[in a new window]
|
Fig. 3. The flexible TDR probe in the borehole revealed by excavation of the probe from the ground. The picture demonstrates the attachment of the flexible TDR waveguides to the irregular shape of the borehole walls.
|
|
Probe Design and Installation Setup
The flexible TDR waveguides are made of three 1.25 by 30 cm stainless-steel foil strips, glued parallel to each other with 2.54-cm spacing, onto a flexible sleeve made of 1.59-mm-thick gum rubber. The probe dimension presented here are similar to those of the standard TDR probe Campbell CS605. However, any other configuration of length and spacing may be used. The flexible TDR waveguide is connected to a coaxial cable, similar to the standard, three-rod TDR probe, where the center waveguide is soldered to the inner conductor of the cable and the two outer waveguides are soldered to the shield (Zegelin et al., 1989). Following the analysis made by Ferré et al. (1998) the sampling area of the probes presented above penetrates to about 2.5 cm into the surrounding soil (90% of the sampling area).
The flexible sleeve to which the TDR waveguides are attached (Fig. 2) is fabricated to a diameter larger then the estimated maximum borehole diameter. In this way, the sleeve can fit the irregularities and diameter variations of the borehole. The flexible sleeve is as long as the borehole depth, and the TDR waveguides are aligned longitudinally at the required spacing and to the depth of interest. A PVC pipe is placed inside the sleeve to add rigidity during installation. The sleeve assembly is inserted into the borehole and rotationally aligned to ensure that the waveguides face the uppermost part of the borehole. (Fig. 1 and 2). Before inserting the sleeve into the borehole, the integrity of the sleeve is pressure tested using compressed air to check for leaks and ensure that no resin will leak into the borehole annular space. The space between the PVC casing and rubber sleeve is filled with a two-part low viscosity liquid urethane. As the liquid urethane is poured into the sleeve, it generates hydrostatic pressure and the sleeve expands to fill the irregular shape of the borehole wall. If the borehole is drilled to the depth of groundwater, the inner PVC pipe may be used for sampling groundwater and for measuring water table fluctuations. Placing temperature sensors next to each flexible TDR waveguide provides complementary information regarding the flow regime, as the temperature profile can be used as a tracer for estimating flow velocities (Constantz and Thomas, 1996). The method described here may be limited if applied in the following conditions. Gravel or pebble sediments may prevent proper contact of the flexible waveguides to the borehole walls. If the probe is installed to depth shallower than 1 m, the hydrostatic pressure may be too low to assure proper contact of the waveguides to the soil material. Installation in an unstable formation may cause the borehole to collapse before the probe assembly is inserted.
|
TESTING AND RESULTS
|
|---|
Laboratory Calibration Tests
Laboratory experiments were conducted to test and refine the design and to establish calibration curves for converting probe signals to soil water content. In this design, the waveguides are only in contact with the soil surface along one side of the flexible sleeve, yet the measured bulk dielectric coefficient is affected by both the soil and the materials used to construct the flexible sleeve and its filling material. Since the dielectric properties of the construction material (PVC, urethane, and rubber) are small (1.5–2.5) and remain constant with time, the calibration curve isolates the response of the soil moisture content and is affected by the probe material.
The calibration curves were established for three different soils, a clay loam from an agricultural field where a field experiment was conducted, a silt loam from the Truckee River bank, and an engineered sand. The calibration tests were performed using containers filled with soils of homogeneous texture and uniform volumetric water content. To simulate the exact design and material composition of the field installation, the flexible TDR probe was covered with 22-mm (1 inch)-thick flexible urethane layer and then loaded by pressure of 3.85 KPa to achieve intimate contact between the waveguides and the soil surface. Figure 4 demonstrates two representative TDR waveforms that were retrieved by a cable tester connected to flexible TDR probe at two different water contents. These typical waveforms demonstrate the probe response to different water contents. Further analyzing of the waveforms by standard technique, such as the method described by Herkelrath et al. (1991), allows calculation of the soils bulk dielectric constant and the volumetric water content.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 4. Time domain reflectometry waveform retrieved by cable tester from the flexible probes system for two volumetric water contents measured in clay loam.
|
|
Figure 5 presents the calibration results in terms of the measured apparent waveguide length vs. the soil volumetric water content. It shows that TDR readings were affected only by changes in the water content, and no significant soil effect was detected on various soils that were used in this study. The response of the probes to changes in water content shows correlation (R2) of 0.99 to a second-order polynomial curve. To evaluate possible variation in probe accuracy due to manufacturing variation, several probes were manufactured and tested throughout calibration. Each probe was calibrated against both soils and through the full range of water contents to confirm that manufacturing variability was negligible. Additionally, TDR readings with double pressure load (7.7 KPa) were taken to ensure that there is no pressure effect. This test showed that all of the probes gave the same readings when used for the same soil conditions, and there were no significant effects due to individual probe, pressure load, or soil type.
The efficiency of the flexible TDR waveguides as compared with three-rod standard TDR probes was evaluated through calibration test where TDR readings from the same soil samples were taken by both standard and flexible waveguides. Soil was mixed with different amounts of water to get a homogeneous evenly mixed soil of constant water content. Then TDR readings were taken from the surface of the packed soil and from inside the soil by standard three-rod TDR probes. Figure 6 presents the relative apparent probe length (apparent length/length) of the standard TDR probe compared with the flexible probe. The figure clearly shows a linear correlation (R2 = 0.995) between the two probe types. Although the calibration curves showed that readings were independent of the soil types used in this research, we strongly recommend performing calibration curves for specific soils and probe construction material.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 6. TDR reading of evenly mixed soil of different water contents by flexible TDR waveguide vs. standard TDR probe.
|
|
Field Experiment
A full-scale field experiment using all components of the proposed methodology was conducted under an agricultural field located at the agricultural farm of the University of Nevada in Reno. The soil texture at the site is mostly fine-loam, and the main agricultural crop is alfalfa (Medicago sativa L.). The field is typically flood irrigated and the groundwater table is about 3.2 m below the surface. The purpose of the experiment was to test the probe fabrication and field installation procedures without investigating the site hydrological characteristics. A sloping borehole, 18 cm in diameter, was drilled at a 45° angle to a total length of 5 m (approximately 3.5 m vertical depth). The borehole penetrated the entire vadose zone from the land surface to the water table. A flexible TDR probe system was installed in the borehole following the construction and installation procedures described above. The flexible waveguides (0.3 m long each) were spaced 1 m apart along the sleeve, so that the vertical depths to the center of each waveguide were 0.8, 1.5, 2.2, and 2.9 m. The individual waveguides were connected to a digital cable tester (Tektronix, model 1502C) through a multiplexer (Campbell model SDM50), and the entire system was connected to a data logger (Campbell model CR10X) for data collection. Because this experiment was conducted as part of the development process, the entire TDR waveform was collected to allow examination of the waveform quality retrieved by the TDR system. Data were recorded every 3 h throughout the growing season (April–August 2001) and were analyzed following the procedure described by Herkelrath et al. (1991).
Figure 7 presents the volumetric water content as was measured at four depths (0.8, 1.5, 2.2, and 2.9 m) during the growing season (May–August). The waveforms indicate that significant noise occurred during the period of measurement. This noise was most likely due to interference from a nearby high voltage power line and was found to influence the system only after the installation was completed. This noise did not occur when the system was operated away from power lines, power motors, or generators. Nevertheless, the waveforms for all four waveguides clearly indicated that the probe recorded increases in water content immediately after irrigation. The deepest waveguide, located at 2.9 m depth, consistently displayed abnormally high volumetric water contents of 70% and higher. We hypothesize that the higher water content from the deepest waveguide is an artifact of improper positioning of the probe, as its end was installed under the water table. Installation under the water table could have prevented proper contact between the soil and waveguide, leaving a gap that could increase the water content reading significantly if it was filled with water. The water content fluctuation at the lower waveguide (2.9 m) reflects the water table rising as a consequence of the flood irrigation and does not represent the soil moisture.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 7. Volumetric water content as measured by the multilevel flexible TDR waveguides in four different depths. It clearly shows the effect of three irrigation flood events on the water content of the entire vadose zone.
|
|
Design Modification
To remove the need for an expensive cable tester and to reduce raw data processing, the standard TDR probe and traditional cable tester were replaced with the Fuel Moisture Sensor, model CS505 (Campbell Scientific). The CS505 electronic device is identical to the Water Content Reflectrometer CS615 (Campbell Scientific), but allows for factory-stock waveguides to be replaced with the customized waveguides. The CS505 does not require the cable tester for collecting propagation velocity data, but contains a bistable multivibrator that measures the travel period for an electromagnetic wave to and from the waveguide end. As with the CS615, the flexible TDR waveguides are connected to only two waveguides, instead of the three that are typical with standard TDR probes. Use of a CS505 with each flexible waveguide facilitates data analysis and eliminates the need to analyze full waveforms. Additionally, unlike the cable tester that is wired to the waveguides by coax cable, the CS505 is directly wired to the waveguides. Therefore, the signal loss that is typically associated with increasing coaxial cable lengths is avoided (Bilskie, 1999).
The performance of the modified flexible TDR waveguides with the CS505 was checked by developing new calibration curves for three types of soil (clay loam, silt loam, and a sandy loam from Yuma Proving Ground, AZ). The calibration results (Fig. 8) demonstrated a nearly linear relationship between the variables (R2 = 0.989). Recently (September 2001), a full-scale field experiment was conducted at Yuma Proving Ground, Arizona, where the flexible TDR probes system combined with CS505 instruments were installed. Three probes were installed in three boreholes that were drilled to sloping depth of 5 m at 35°. Each sampler consisted of four flexible TDR waveguides, spaced at 1-m increments along the borehole (mean depth of installation 1, 1.8, 2.7, and 3.5 m) in ephemeral washes of a desert terrain. Although the system has operated since September 2001, no significant precipitation event has occurred in the study area. Therefore the experimental results presented here only demonstrate the measuring system capability and the stability of the readings. Figure 9 presents the water content measurements in one of the three waveguides. The gradual loss of water from the upper section of the borehole represents the loss of the water that was added to the borehole during the drilling process. By December 2001 a steady equilibrium was achieved, followed by a slight increase in the water content of the upper section during the winter and spring, with no significant change in the water content readings from spring through fall of 2002.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 9. Volumetric water content variation as was measured by Flexible TDR system using a Campbell Scientific CS505 instrument in four depths at Yuma Proving Ground in Arizona.
|
|
|
CONCLUSIONS
|
|---|
Deep vadose zone monitoring requires a methodology capable of providing continuous, real-time, in situ measurements of water content. We present a new TDR application that extends the commonly used TDR technique for in situ, real-time, measurement of water content in deep vadose zone horizons, thus providing the capability to measure water content at any depth from the land surface to the water table. This TDR technique also provides soil water content readings for relatively undisturbed soil conditions, thus providing more representative results. Once the probes are installed into the ground, the system requires no maintenance. Throughout laboratory and field experiments, the method was found to be reliable and capable of providing water content measurements in deep vadose zone horizons.
The flexible TDR system is especially suitable for deep vadose zone monitoring in sites where undisturbed soil column conditions are required or where direct access from land is difficult. Those may include monitoring of the entire vadose cross section under landfills or waste ponds, monitoring the moisture profile under the streambed of ephemeral washes, and for the study of water percolation and groundwater recharge.
|
ACKNOWLEDGMENTS
|
|---|
Funding for development and testing of the flexible probe was provided through a Nevada Applied Research Initiative grant provided in support of United States Department of Defense (EPSCoR Grant 207035), the Desert Research Institute Center for Arid Lands Environmental Management, the Division of Hydrological Sciences, and The Vaadia-BARD Postdoctoral Fellowship (Fellowship FI-291-99). We thank Don Kennedy from the University of Nevada Agricultural Farm in Reno for the use of the farmland and facilities for the field experiment. We thank Todd Mihevc, Todd Caldwell, and Sergio Obregon for help during field testing. We would also like to thank the anonymous reviewers, who helped improve the manuscript through their constructive comments.
|
REFERENCES
|
|---|
- Bilskie, J. 1999. TDR manual. Campbell Scientific, Logan, UT.
- Constantz, J., and L.C. Thomas. 1996. The use of streambed temperature profiles to estimate the depth, duration, and rate of percolation beneath arroyos. Water Resour. Res. 32:3597–3602.
- Dalton, F.N. 1992. Development of time-domain reflectometry for measuring soil water content and bulk soil electrical conductivity. Soil Sci. Soc. Am. J. 30:143–167.
- E.S.I. Environmental Sensors, Inc. Data from web page. Available at http://www.esica.com/index.html (revised 10 Oct. 2002; accessed December 2002; verified 12 Mar. 2003.) E.S.I. Environmental Sensors, Inc., Victoria, BC, Canada.
- Ferré, P.A., H.J. Knight, L.D. Rudolph, and G.R. Kachanoski. 1998. The sample area of conventional and alternative time domain reflectometry probes. Water Resour. Res. 34:2971–2979.
- Herkelrath, W.N., S.P. Hamburg, and F. Murphy. 1991. Automatic, real-time monitoring of soil moisture in a remote field area with time domain reflectometry. Water Resour. Res. 27:857–864.
- Hillel, D. 1980. Application of soil physics. Academic Press, Orlando, FL.
- Ledieu, J., P. de Ridder, P. de Clerck, and S. Dautrebande. 1986. A method of measuring soil moisture by time-domain reflectometry. J. Hydrol. (Amsterdam) 88:319–328.
- Murdoch, L., B. Harrar, B. Nilsson, W. Slack, and R. Siegrist. 1999. In situ measurements in fractured till using sidewall sensors. Nord. Hydrol. 30:257–266.
- Schwartz, U., and A.P. Schick. 1998. Floods in Nahal Zin and their influence on the alluvial aquifer. p. 88. In Proc. Israel Geological Society annual meeting. Jerusalem, Israel. 22–24 Mar. 1998.
- Selker, J.S., L. Graff, and T. Steenhuis. 1993. Noninvasive time domain reflectometry moisture measurement probe. Soil Sci. Soc. Am. J. 57:934–936.[Abstract/Free Full Text]
- Topp, G.C., J.L. Davis, and A.P. Annan. 1980. Electromagnetic determination of soil water content: Measurement in coaxial transmition line. Water Resour. Res. 16:574–582.
- Topp, G.C., J.L. Davis, W.G. Bailey, and W.D. Zebchuk. 1984. The measurement of soil water content using a portable TDR hand probe. Can. J. Soil Sci. 64:313–321.
- Wyatt, D., D. Stephenson, B. Looney, J. Rossabi, J. Cook, H. Holmes-Burns, and J.B. Sisson. 1999. E-aAea vadose zone proposed plan (U). Document WSRC-RP-99–4039, Rev. 0. Westinghouse Savannah River Company, Aiken, SC.
- Zegelin, S.J., I. White, and D.R. Jenkins. 1989. Improved field probes for soil water content and electrical conductivity measurement using time domain reflectometry. Water Resour. Res. 25:2367–2376.
- Zegelin, S.J., I. White, and G.F. Russell. 1992. A critique of the time domain reflectometry technique for determining field soil-water content. Soil Sci. Soc. Am. J. 30:187–208.
- Zilberbrand, M., and H. Gvirtzman. 1996. Monitoring of water flow and solute transport through the unsaturated zone using a large-diameter borehole. Ground Water 34:57–65.[ISI]
This article has been cited by other articles:
|
|
|
|
 
E. D. Mattson, K. E. Baker, C. D. Palmer, C. R. Breckenridge, J. M. Svoboda, and R. W. Smith
A Flexible Water Content Probe for Unsaturated Soil Column Experiments
Vadose Zone J.,
May 26, 2006;
5(2):
805 - 808.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
