SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. The DPIL
  5. Field assessment
  6. Conclusions
  7. Acknowledgments
  8. References

Information on vertical variations in hydraulic conductivity (K) can often shed much light on how a contaminant will move in the subsurface. The direct-push injection logger has been developed to rapidly obtain such information in shallow unconsolidated settings. This small-diameter tool consists of a short screen located just behind a drive point. The tool is advanced into the subsurface while water is injected through the screen to keep it clear. Upon reaching a depth at which information about K is desired, advancement ceases and the injection rate and pressure are measured on the land surface. The rate and pressure values are used in a ratio that serves as a proxy for K. A vertical profile of this ratio can be transformed into a K profile through regressions with K estimates determined using other techniques. The viability of the approach was assessed at an extensively studied field site in eastern Germany. The assessment demonstrated that this tool can rapidly identify zones that may serve as conduits for or barriers to contaminant movement.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. The DPIL
  5. Field assessment
  6. Conclusions
  7. Acknowledgments
  8. References

One of the major challenges facing investigators of sites of ground water contamination is how to assess the threat posed by the contamination. Without reliable means to perform such assessments, it is difficult to effectively use resources for remediation activities. As geologic strata tend to exhibit considerably more continuity in the lateral direction than in the vertical, information on vertical variations in K can often shed much insight into how a contaminant will move at a site. Butler (2005) describes the major methods for obtaining vertical profiles of K, the most common of which is the borehole flowmeter (Molz and Young 1993; Crisman et al. 2001). This method requires a well that is screened across the units of interest. At many sites, such wells are not common because of regulatory concerns about vertical movement of contaminants within the screened interval. In shallow (less than 30 m) unconsolidated settings, direct-push (DP) technology (McCall et al. 2005; Dietrich and Leven 2006) provides an effective vehicle for obtaining information on vertical variations in K in the absence of wells. Butler et al. (2007) have recently described a promising DP method for obtaining K estimates at a resolution and accuracy that has not previously been possible. This approach, however, takes time, as 10–15 min are required to test one relatively permeable interval (K > 1 m/d). A more rapid approach is needed to provide K information in support of site-screening and risk-assessment activities.

The direct-push injection logger (DPIL), the subject of this note, was developed to obtain rapid information about vertical variations in K in shallow unconsolidated settings. The primary objectives of the note are to describe this tool and evaluate the information it can provide. The note begins with a basic description of the procedure followed by an overview of the method used to process the acquired data. The viability of the approach is then assessed in a relatively controlled field setting. The note concludes with a discussion of the major advantages and limitations of the tool.

The DPIL

  1. Top of page
  2. Abstract
  3. Introduction
  4. The DPIL
  5. Field assessment
  6. Conclusions
  7. Acknowledgments
  8. References

The DPIL is a small-diameter tool with a short screen (Figure 1) that is attached to the lower end of a pipe string and advanced into the subsurface with hammer-assisted DP technology. As the tool is advanced, water is continually injected through the screen at a relatively high rate (up to a few L/min in sands and gravels) to keep the screen clean. Upon reaching a depth at which information about K is desired, advancement ceases and the water pressure in the injection tubing is measured at different injection rates using a pressure transducer and flow controller on the surface (Figure 1). Pitkin and Rossi (2000) describe a somewhat similar approach in which cessation of advancement is not required. Although information can be obtained more rapidly with that approach, more uncertainty is introduced because it is not possible to separate the pressures induced by injection from those induced by tool advancement. Cessation of advancement enables that source of uncertainty to be eliminated in permeable formations.

image

Figure 1. Schematic of the DPIL (a screen with a radius [outer] and length of 0.035 and 0.025 m, respectively, and an injection tube with an inner diameter of 0.008 m were used in this work).

Download figure to PowerPoint

The total resistance for the water injection (Rtotal) is the ratio of the injection pressure (pinj) over the injection rate (Q):

  • image(1)

Note that this resistance term is the inverse of the specific capacity parameter commonly computed in well performance tests. The injection pressure consists of the following components:

  • image(2)

where ptrans is the pressure measured at the transducer on the surface and inline image is the pressure exerted by a water column of length i (Figure 1). The hydrostatic pressure of the water column between the test interval and the water table inline image cancels out and thus does not appear in Equation 2.

The total resistance is a function of conditions at the screen and within the injection tube. In order to evaluate conditions at the screen, the influence of the hydraulic behavior within the tube must be removed. The total resistance can be expressed as a sum of the resistance due to conditions at the screen (Rscreen) and the resistance due to flow in the tube (Rtube):

  • image(3)

The resistance at the screen is inversely related to the hydraulic conductivity of the strata adjacent to the screen. Thus, using this resistance, a ratio can be calculated to serve as a proxy for K:

  • image(4)

Note that this ratio is different from the index of hydraulic conductivity that is used in the Waterloo Profiler (Pitkin and Rossi 2000) and defined as Q/ptrans (Equation 4 without the the inline imageterm in Equation 2 and the Rtube term).

The calculation of the KDPIL ratio requires that the tube resistance be known. In the case of laminar flow within the tube, the resistance can be calculated using the Hagen-Poiseuille law (Vennard and Street 1975):

  • image(5)

where L is the length of the tube between the pressure transducer and the screen, r is the radius of the tube, and νis the kinematic viscosity of the injected fluid. In the case of turbulent flow within the tube, the resistance depends also on the flow rate and the roughness of the tube wall. For this work, a linearization of the relation between tube resistance and flow rate is used for turbulent conditions:

  • image(6)

where a and b are parameters that can be determined from a regression analysis of flow and pressure data obtained from experiments with the DPIL equipment. In those experiments, pressures are measured for different flow rates while the injection screen is placed at the same distance above land surface as the pressure transducer. The validity of Equation 6 and the appropriate values for parameters a and b should be assessed at each site.

Using Equations 5 and 6, the tube resistance can be described as a function of flow rate. The transition point between laminar and turbulent flow is characterized by the critical Reynolds number (Vennard and Street 1975). The critical Reynolds number for flow in tubes is approximately 2100; so, using the definition of the Reynolds number (=2Q/rπν), the flow rate (Qcr) at the transition point can be written as follows:

  • image(7)

Because tube resistance in the vicinity of the transition from laminar to turbulent flow is very difficult to characterize, flow rates for the DPIL should be chosen to be considerably above or below the critical rate defined in Equation 7.

In summary, the DPIL procedure consists of measuring the injection rate and pressure at a given vertical position and then calculating the KDPIL ratio from the acquired data. This calculation involves three steps: (1) using Equation 1 to calculate the total resistance from field measurements; (2) using Equation 5 or 6 to calculate the tube resistance; and (3) using Equation 4 to remove the tube resistance from the total resistance. A vertical profile of the KDPIL ratio can thus be generated from the results of a series of sequential advances and injection tests.

Field assessment

  1. Top of page
  2. Abstract
  3. Introduction
  4. The DPIL
  5. Field assessment
  6. Conclusions
  7. Acknowledgments
  8. References

The viability of the DPIL procedure was assessed at the Nauen test site, which is located approximately 40 km west of Berlin, Germany. This site was established by the German Federal Institute for Geosciences and Natural Resources in collaboration with the Department of Applied Geophysics of the Technical University of Berlin for the development and evaluation of methods for hydrogeophysical characterization (Goldbeck 2002; Yaramanci et al. 2002). The results from the previous work, in particular, data from a continuously cored borehole and recent DP investigations (Butler et al. 2007), create a relatively controlled framework for method evaluation. The work described here was done in the upper 20 m of the unconsolidated sequence at the site, which consists primarily of fine- to medium-grained sands. Two DPIL profiles were obtained in the vicinity of an existing well at which a continuous core had been collected and within 2 m of sites of DP slug tests (Butler et al. 2002; Butler 2002) and a direct-push permeameter (DPP) profile (Butler 2005; Butler et al. 2007). Data from the DP slug tests, DPP, and the continuous core will be used here to assess the viability of the DPIL approach.

In the first profile, DPIL1, measurements were obtained at 24 levels with a vertical spacing of approximately 0.5 m. The profile was terminated at 15.49 m, just past the maximum depth of the nearby DPP profile (14.94 m). At each level, measurements were obtained at three flow rates to assess the dependence of tube resistance on flow rate. The injection rates ranged from 42 to 100 L/h. In all cases, the flow was turbulent, considerably above the critical flow rate defined in Equation 7. In Figure 2, the ratio of the injection rate over the pressure measured at the transducer (Q/ptrans) is calculated following the approach used by Pitkin and Rossi (2000). In Figure 3, the KDPIL ratio is calculated using the procedure described here. The coincidence of KDPIL profiles in Figure 3 demonstrates the importance of including tube resistance and pressure inline image for obtaining defensible data with the DPIL.

image

Figure 2. Comparison of DPIL profiles for different injection rates. The plotted ratio Q/ptrans neglects Rtube and the pressure of the water column between the pressure transducer and the water table.

Download figure to PowerPoint

image

Figure 3. Comparison of profiles of the KDPIL ratio calculated using the procedures outlined in text with the K profile obtained using the DPP.

Download figure to PowerPoint

The results of the second profile, DPIL2, are also shown in Figure 3. DPIL2 was performed to evaluate the repeatability of the approach at a nearby location (1.23 m from DPIL1). Measurements were obtained at 35 levels with a vertical spacing of approximately 0.5 m. At each level, a single flow rate (43 to 48 L/h) was used. The profile was continued until the tool could no longer be advanced (maximum depth of 18.44 m). As shown in Figure 3, the two closely spaced DPIL profiles are, in general, in very good agreement, demonstrating the lateral continuity of site hydrostratigraphy over this short separation distance. The biggest difference between the two profiles is in the vicinity of the apparent low K unit at a depth of 9 m. This discrepancy may largely be a result of small differences in vertical position coupled with the short (0.025 m) screen length. In both profiles, an injection rate of 3 to 4 L/min (180 to 240 L/h) was used during advancement, except in the vicinity of the apparent low K unit, to keep the screen clean.

The vertical variations displayed in the KDPIL profiles of Figure 3 suggest that the DPIL approach can provide valuable information about vertical variations in K. The results of a nearby (distance of 1.16 and 1.65 m from DPIL1 and DPIL2, respectively) DPP profile are also plotted on Figure 3 to substantiate that suggestion. The comparison of the two DPIL profiles with the DPP profile is quite good, demonstrating that the DPIL can provide important information about relative variations in K. Note that the DPP tool used in this assessment can be advanced only by pushing (no hammering) because of the fragility of the pressure transducers incorporated in that tool (Butler et al. 2007). The DPIL2 profile indicates the existence of a zone of apparent low K just below the depth of maximum penetration of the DPP. This zone may be an interval of more compacted material, which would increase the resistance to pushing and thus prevent the DPP from being advanced to greater depths.

Figure 3 demonstrates that the DPIL can provide valuable information about relative variations in K. In an attempt to get more quantitative information, slug test data from nearby DP installations (separation distances of 0.27 to 1.84 m) were used in a regression analysis. The regression was performed between KDPIL ratios and slug test K estimates obtained at the same depths using the KDPIL ratios from both profiles. Although only four slug test K estimates were available, the K range spanned by those values and the large R2 (0.958) indicate the possibility of a strong correlation between the KDPIL ratio and K (Figure 4). The regression equation was therefore used to transform the KDPIL ratios into K estimates. Figure 5 compares the K values calculated with the regression to the K estimates obtained from both the DPP and slug tests. In addition, the dominant grain-size intervals determined from a sedimentologic analysis of the core samples are shown. The comparison in Figure 5 demonstrates that the DPIL approach can identify zones of different hydraulic conductivity and can, when coupled with nearby K data, yield semiquantitative estimates of the hydraulic conductivity of the logged interval.

image

Figure 4. Regression analysis of KDPIL ratio vs. K values from DP slug tests (KST). Four KDPIL values are available for each depth at which a slug test was performed.

Download figure to PowerPoint

image

Figure 5. Comparison of DPIL K profiles obtained using the regression of Figure 4 with K values from DP slug tests and the DPP. Dominant grain-size intervals based on the German Institute for Standardization (Deutsches Institut für Normung) classification system are also plotted. The DPIL1 K profile is based on the average KDPIL ratio of the three profiles shown in Figure 3.

Download figure to PowerPoint

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. The DPIL
  5. Field assessment
  6. Conclusions
  7. Acknowledgments
  8. References

The DPIL is a promising tool for delineating vertical variations in hydraulic conductivity in shallow unconsolidated settings. The major advantages of the approach are its speed and mechanical robustness. In units of moderate or higher K (1 m/d or higher), testing at a single level with two flow rates can be completed within 2 to 3 min. This is in contrast to the 10 to 15 min needed to complete testing of a single level with the DPP and the 1 to 2 h needed for DP slug tests (Sellwood et al. 2005; Butler et al. 2007). The robustness of the tool allows it to be used with hammer-assisted DP technology to reach greater depth than possible with a push-only advancement. In addition, the KDPIL ratio can be transformed into semiquantitative K estimates through regressions with K values obtained using other techniques at similar depth intervals.

The major limitation of the DPIL is that clogging of the injection screen can be misinterpreted as a decrease in K. Changes in the effective screen length due to clogging can have a significant effect on the KDPIL ratio, so relatively high injection rates (3 to 4 L/min in this work) should be used during tool advancement to minimize that possibility.

The field assessment reported here demonstrates that the DPIL has considerable potential for rapid delineation of vertical variations in K in support of contaminant site investigations. The tool should prove of particular value for the rapid identification of zones that may serve as conduits for or barriers to contaminant movement. Knowledge of the existence of such zones is critical for realistic risk assessments and the design of effective remediation schemes.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. The DPIL
  5. Field assessment
  6. Conclusions
  7. Acknowledgments
  8. References

This work was supported by the Centre of Applied Geoscience of the University of Tübingen, the DAAD (German Academic Exchange Service) Study Visit Grant Program, and the sabbatical leave program of the University of Kansas. The authors gratefully acknowledge the Department of Applied Geophysics of the Technical University of Berlin for access to the Nauen Test Site and for freely sharing the data from that site. Furthermore, we want to thank Bob Sterrett, Dave Hart, and an anonymous reviewer for the technical review provided.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. The DPIL
  5. Field assessment
  6. Conclusions
  7. Acknowledgments
  8. References
  • Butler, J.J. Jr. 2005. Hydrogeological methods for estimation of hydraulic conductivity. In Hydrogeophysics, ed. Y.Rubin and S.Hubbard, 2358. Dordrecht, The Netherlands: Springer.
  • Butler, J.J. Jr. 2002. A simple correction for slug tests in small-diameter wells. Ground Water 40, no. 3: 303307.
  • Butler, J.J. Jr., P. Dietrich, V. Wittig, and T. Christy. 2007. Characterizing hydraulic conductivity with the direct-push permeameter. Ground Water 45, no. 4: 409419.
  • Butler, J.J. Jr., J.M. Healey, G.W. McCall, E.J. Garnett, and S.P. Loheide II. 2002. Hydraulic tests with direct-push equipment. Ground Water 40, no. 1: 2536.
  • Crisman, S.A., F.J. Molz, D.L. Dunn, and F.C. Sappington. 2001. Application procedures for the electromagnetic borehole flowmeter in shallow unconfined aquifers. Ground Water Monitoring and Remediation 21, no. 4: 96100.
  • Dietrich, P., and C. Leven. 2006. Direct push technologies. In Groundwater Geophysics, ed. R.Kirsch, 321340. Berlin, Heidelberg, Germany: Springer.
  • Goldbeck, J. 2002. Hydrogeophysical methods at test site Nauen—Evaluation and optimization. Master thesis at the Technical University of Berlin, Institute of Applied Geoscience, Berlin.
  • McCall, W., D.M. Nielsen, S. Farrington, and T.C. Christy. 2005. Use of direct-push technologies in environmental site characterization and ground-water monitoring. In The Practical Handbook of Environmental Site Characterization and Ground-Water Monitoring, 2nd ed., ed. D.M.Nielsen, 345472. Boca Raton, Florida: CRC Press.
  • Molz, F.J., and S.C. Young. 1993. Development and application of borehole flowmeters for environmental assessment. Log Analyst 34, no. 1: 1323.
  • Pitkin, S.E., and M.D. Rossi. 2000. A real time indicator of hydraulic conductivity distribution used to select groundwater sampling depths. Eos 81, no. 19: S239.
  • Sellwood, S.M., J.M. Healey, S. Birk, and J.J. Butler Jr. 2005. Direct-push hydrostratigraphic profiling: Coupling electrical logging and slug tests. Ground Water 43, no. 1: 1929.
  • Vennard, J.K., and R.L. Street. 1975. Elementary Fluid Mechanics, 5th ed. New York: John Wiley and Sons.
  • Yaramanci, U., G. Lange, and M. Hertrich. 2002. Aquifer characterisation with surface-NMR and other geophysical techniques at the test site Nauen/Berlin. Journal of Applied Geophysics, Special Issue on Surface-NMR 50, no. 1–2, 4765.