5.2 Borehole Testing
Borehole testing in karst systems differs from investigation in other aquifer types because connected conduit networks need to be identified and delineated. Tests used in other types of aquifers have some application in karst with respect to identifying strata that tend to contain conduits and control their connections, but additional tests are needed to specifically find conduits and evaluate their connectivity. These tests include: borehole geophysical techniques that acquire images of borehole walls and measure flow to/from specific zones; water tracing tests that determine connectivity of conduits over short and long distances; and aquifer tests (often using packers to isolate sections of the borehole) that provide information on the distribution of transmissivity and storativity in the aquifer.
Borehole Geophysical Tools Commonly Applied to Any Aquifer Type
Often borehole measurements at a single well provide minimal information in karst aquifer studies because randomly drilled boreholes are unlikely to intersect conduits that are important to karst systems, especially if the rock matrix porosity is small and conduits are sparse. Measurements of the transmissivity or permeability of fractures and solution openings intersecting boreholes represent formation properties that apply to, at most, a few borehole diameters around the measurement point. This is an inherent property of the radially convergent or divergent flow regime where most flow dissipation occurs in the immediate vicinity of the borehole wall. Therefore, geophysical well logs are mostly of use in characterizing the general hydrogeologic context of the formation containing karst features. In the example shown in Figure 47, conduits are unlikely to be detected; however traditional natural gamma logs from several wells are used to determine the subsurface dip of the sedimentary rocks. Borehole logs can indicate the continuity or offset of bedding plane horizons, or the pattern of fracture connectivity related to joints and large-scale stress patterns influencing the geometry of solution openings.
Borehole geophysics is generally used in groundwater and environmental investigations to: delineate hydrogeologic units; define water quality; and determine well construction and condition. These generalized applications of logs are discussed in a separate GW-Project book describing borehole geophysics.

Figure 47 – a) Horizontal alignment of gamma logs along an east west profile was used to identify stratigraphic dip in a dolomite aquifer in northern Illinois. The gamma-signal along with borehole lithologic logs are used to distinguish a geologic unit that has a similar gamma signal (marker). This marker is shown within the dashed lines and used to align the logs. b) Televiewer logs, stratigraphic correlation, and flowmeter information were used to identify continuous bedding planes and aquifer flow zones where arrows indicate direction of ambient flows measured in boreholes and locations of inflow and outflow (Paillet and Crowder, 1996). Inflow or outflow was associated with solutional-enlarged bedding plane openings. Many such bedding planes intersected each borehole, but only a few conducted most of the flow. The correlation of these conductive bedding planes was established over borehole separations of about a kilometer by correlating gamma logs. The gamma correlation established the strike and dip of bedding so that borehole elevation and the regional dip could be used to define the precise stratigraphic position of the bedding planes in each borehole. This structural correlation showed that sets of bedding planes served as regional conduits, but that the most transmissive bedding plane within groups of closely spaced bedding planes varied from one borehole to the next.
Common borehole geophysical logs include caliper, gamma, single-point resistance, spontaneous potential, normal resistivity, electromagnetic induction, fluid resistivity, temperature, flowmeter (ambient and pumping), television, and acoustic and optical televiewers provide continuous and in-situ data (Table 5). Multiple logs are typically collected to take advantage of their synergistic nature—much more can be learned by the analysis of a suite of logs as a group than by the analysis of the same logs individually.
Table 5 – Borehole geophysical tools that record properties with depth and what they measure.
Borehole Geophysical Tools Particularly Useful for Characterizing Karst Aquifers
Borehole geophysics provides several tools that can make significant contributions to characterization of karst systems. Image logs (Figure 16 and Figure 17) and flow logs provide detailed information about the nature of hydraulically active zones intersected by boreholes that can be used to infer how these zones fit into regional hydrogeology. Geometric correlation of logs from different boreholes can indicate connections of features between boreholes. Water-chemistry and hydraulic-head data derived from logs can be used to identify possible connections of flow paths in the subsurface. Cross-borehole flow experiments can be used to infer the properties of hydraulic connections among subsurface conduits. Geophysical measurements can be made at local, intermediate, and large scales to infer the relation between scale and hydraulic conductivity (Paillet, 2001).
Most geophysical logs provide precise information about the in-situ properties of subsurface formations in the form of measurements such as gamma activity or electrical conductivity that are indirectly related to the hydraulic properties of interest. Generally, the transmissivity of bedding planes, fractures, and solution openings cannot be inferred from the appearance of those features on borehole image logs or the apparent aperture of those features on caliper logs (Paillet, 1998). High-resolution flow logging equipment such as the heat-pulse (Hess, 1986) and electromagnetic (Molz et al., 1994) flowmeters add the important ability to tie borehole hydraulics to geophysical log data.
Paillet (1998) illustrates the application of logs to karst aquifers, by comparing gamma, short-normal resistivity, fluid column resistivity, caliper, and televiewer logs with a borehole flow profile obtained with a heat-pulse flowmeter during injection in a borehole that was drilled into fractured and bedded limestone in northern Arizona. The logs indicate the precise depths where water exits the borehole during steady injection. The outflow points can be associated with features on the other logs that represent the hydraulically conductive features in the vicinity of the borehole, including fractures, bedding planes, and a small cavern. Although the full set of logs from one well provides no information about how far these features extend away from the borehole, the flow log indicates where flow enters or exits the borehole. Patterns from all the logs associated with the flow feature may represent a useful vertical marker for comparison with borehole logs from adjacent wells. These marker patterns in specific units can be used to find similar zones in other wells where flow logs have not been recorded. Similar to the use of the gamma-log marker pattern in Figure 47. This is an important step beyond simply identifying the fractures and solution features that intersect a single borehole.
One possible approach to understanding how hydraulically conductive fractures, bedding planes, and solution openings identified in boreholes are connected to form conduit flow systems is to project these features into the regions between boreholes. This seems simple in principle, but is difficult in practice when there are many features that may be permeable in each borehole and boreholes are located far apart. Spatial correlation based on the appearance of features in image logs and their occurrence at similar depths is generally not effective. An effective approach is to locate permeable openings with respect to sedimentary structure and potential marker units (parts of the formation with a distinct geophysical log pattern that are associated with flow). The combination of structural correlation and flow log analysis can be useful in identifying how solution openings are organized into continuous flow paths even when large-scale aquifer-test data are not available.
Geophysical measurements made in boreholes apply to the immediate vicinity of the measurement point, however, hydraulic head measurements made at discrete points in a borehole can be related to the large-scale flow field in the formation and, therefore, to the possible presence and location of interconnected conduits supporting active flow. Such discrete head measurements have traditionally been made with straddle-packer equipment that isolate short sections of borehole to provide a direct measurement of hydraulic head in openings connected to that interval. Use of straddle packer technology is equipment- and time-intensive and is especially cumbersome in boreholes that are intersected by many potentially separate flow zones. For this reason, the US Geological Survey developed a wireline-operated packer system as a simple extension of a typical geophysical logging program (Paillet et al., 1990). The equipment was designed to measure the hydraulic head above and below a single packer dividing the borehole into two separate compartments, and then data analysis converted a series of such measurements into hydraulic head values for the intervals between individual packer stations.
An example of the borehole-packer head-measurement application is given in Figure 48 and Figure 49 for a karst site in eastern Illinois, USA. This study investigated a site where heavy-metal contamination was detected at shallow depths in bedded limestone overlying a municipal aquifer with production wells completed below 1000 feet (~300 m) in depth. In planning the study, it was assumed that a steep vertical hydraulic head gradient was present in the shallow subsurface. Heat-pulse borehole flowmeter logs indicated strong downward flow under ambient conditions in response to that gradient. However, flow logs indicated an unexpected reversal to upward flow near the bottom of the boreholes (Figure 48). The wireline packer was used to investigate the hydraulic head distribution along the borehole and showed that very little vertical head gradient was present. Instead, the packer analysis showed a nearly constant hydraulic head along the borehole with lower head in a single zone about 165 feet (~50 m) deep. Although no obvious karst features were indicated at that depth on the televiewer or caliper logs, the head data showed that flow in the shallow part of the formation was being conducted along one or more bedding planes at that depth, probably to a vertical conduit located at an undetermined distance.

Figure 48 – Neutron, gamma, caliper, televiewer and heat pulse flowmeter logs for borehole T-6 located in eastern Illinois karst. Unit conversions: 100 feet ~30.5 m; 1 inch ~2.54 cm; 1 gallon per minute ~3.8 L/min (liters per minute). From Paillet and others (1990).

Figure 49 – Estimates of in situ hydraulic head made with an experimental wireline-operated packer for borehole T-6 in karst in eastern Illinois karst. Unit conversions: 100 feet ~30.5 m; 1 foot ~0.305 m. From Paillet and others (1990).
At the time of the wireline packer study, there was no alternative for interval head measurements. Since then, methods for estimating the hydraulic head directly from borehole flowmeter measurements have been developed (Paillet, 2000; Day-Lewis et al., 2002). The analysis starts where a pair of steady flow profiles obtained under two different conditions (usually ambient and pumping) are simultaneously modeled to estimate transmissivity and hydraulic head for each inflow or outflow zone. An example of this technique applied to a karst aquifer in north-central Tennessee is shown in Figure 50. The borehole site was located a fraction of a kilometer from a ravine with known karst discharge from a small cavern. The ambient flow profile showed downflow entering at about 128 feet (~39 m) and exiting at about 143 feet (~50 m). Pumping at about 1.3 gallon per minute changed the flow regime to upward flow entering at about 154 feet (~47 m) and exiting at 128 feet (~39 m). Flow log modeling showed three equally transmissive zones with the lowest hydraulic head at 143 feet (~50 m), highest head at 128 feet (~39 m), and an intermediate head at 154 feet (~47 m). This distribution indicated that the bedding plane at 143 feet (~50 m) was connected to the discharge point in the nearby ravine and that flow was directed downward towards that horizon from above, and upward from below.
It is important to note the significance of conducting both ambient and pumping flow measurements in this situation. The ambient and pumping flow profiles in Figure 50 show that the central flow conduit in this karst study would not have been identified if only a pumping profile had been obtained. The ambient flow log indicates, there is no flow above and below the fracture zone (from approximately 128-143 feet (~39-44 m) below top of casing) and all measurements indicate small downward flow (Figure 50). From the pumping flow log, there is still no flow at the bottom of the well, some flow is detected slightly below the main fracture and enters in the fracture zone (a constant and low upward flow rate), then just above the main fracture (note large opening in caliper log at approximately 129 feet below top of casing; the flow rate exhibits an abrupt increase and stabilizes. Thus, with only a pumping flow log a much longer vertical zone would appear to be contributing flow to the borehole. The combination of the ambient and pumping flow logs and the caliper log help identify the fracture where most of the water enters the borehole.

Figure 50 – Geophysical log composite for a karst well at Fort Campbell, Tennessee using flowmeter logging. Unit conversions: 100 feet ~30.5 m; 1 inch ~2.54 cm; 1 gallon per minute ~3.8 L/min (liters per minute). From Paillet (2000) and Day Lewis and others (2002).
Many hydrologists still rely on straddle packer measurements to determine hydraulic head distributions in karst formations. Televiewer and caliper logs often indicate several potential flow zones that would have to be treated with individual tests. Even when packer work is included in a study, the use of flow log profiling as part of the geophysical log suite can identify the specific flow zones that take part in flow to enhance the efficacy of packer work or piezometer installation. Also, once the initial pair of flow logs has been analyzed, zone transmissivity is determined, and subsequent monitoring of zone hydraulic head variation can be conducted with repeat ambient flow logs in an open borehole without use of packers or elaborate completions.