An evaluation of borehole flowmeters used to measure horizontal ground-water flow in limestones of Indiana, Kentucky, and Tennessee, 1999

Three borehole flowmeters and hydrophysical logging were used to measure ground-water flow in carbonate bedrock at sites in southeastern Indiana and on the west-central border of Kentucky and Tennessee. The three flowmeters make point measurements of the direction and magnitude of horizontal flow, and hydrophysical logging measures the magnitude of horizontal flowover an interval. The directional flowmeters evaluated include a horizontal heat-pulse flowmeter, an acoustic Doppler velocimeter, and a colloidal borescope flowmeter. Each method was used to measure flow in selected zones where previous geophysical logging had indicated water-producing beds, bedding planes, or other permeable features that made conditions favorable for horizontal-flow measurements.

Background geophysical logging indicated that ground-water production from the Indiana test wells was characterized by inflow from a single, 20-foot-thick limestone bed. The Kentucky/Tennessee test wells produced water from one or more bedding planes where geophysical logs indicated the bedding planes had been enlarged by dissolution. Two of the three test wells at the latter site contained measurable vertical flow between two or more bedding planes under ambient hydraulic head conditions.

Field measurements and data analyses for each flow-measurement technique were completed by a developer of the technology or by a contractor with extensive experience in the application of that specific technology. Comparison of the horizontal-flow measurements indicated that the three point-measurement techniques rarely measured the same velocities and flow directions at the same measurement stations. Repeat measurements at selected depth stations also failed to consistently reproduce either flow direction, flow magnitude, or both. At a few test stations, two of the techniques provided similar flow magnitude or direction but usually not both. Some of this variability may be attributed to naturally occurring changes in hydraulic conditions during the 1-month study period in August and September 1999. The actual velocities and flow directions are unknown; therefore, it is uncertain which technique provided the most accurate measurements of horizontal flow in the boreholes and which measurements were most representative of flow in the aquifers.

The horizontal heat-pulse flowmeter consistently yielded flow magnitudes considerably less than those provided by the acoustic Doppler velocimeter and colloidal borescope. The design of the horizontal heat-pulse flowmeter compensates for the local acceleration of ground-water velocity in the open borehole. The magnitude of the velocities estimated from the hydrophysical logging were comparable to those of the horizontal heat-pulse flowmeter, presumably because the hydrophysical logging also effectively compensates for the effect of the borehole on the flow field and averages velocity over a length of borehole rather than at a point. The acoustic Doppler velocimeter and colloidal borescope have discrete sampling points that allow for measuring preferential flow velocities that can be substantially higher than the average velocity through a length of borehole. The acoustic Doppler velocimeter and colloidal borescope also measure flow at the center of the borehole where the acceleration of the flow field should be greatest.

Of the three techniques capable of measuring direction and magnitude of horizontal flow, only the acoustic Doppler velocimeter measured vertical flow. The acoustic Doppler velocimeter consistently measured downward velocity in all test wells. This apparent downward flow was attributed, in part, to particles falling through the water column as a result of mechanical disturbance during logging. Hydrophysical logging yielded estimates of vertical flow in the Kentucky/Tennessee test wells. In two of the test wells, the hydrophysical logging involved deliberate isolation of water-producing bedding planes with a packer to ensure that small horizontal flow could be quantified without the presence of vertical flow. The presence of vertical flow in the Kentucky/Tennessee test wells may preclude the definitive measurement of horizontal flow without the use of effective packer devices. None of the point-measurement techniques used a packer, but each technique used baffle devices to help suppress the vertical flow. The effectiveness of these baffle devices is not known; therefore, the effect of vertical flow on the measurements cannot be quantified.

The general lack of agreement among the point-measurement techniques in this study highlights the difficulty of using measurements at a single depth point in a borehole to characterize the average horizontal flow in a heterogeneous aquifer. The effective measurement of horizontal flow may depend on the precise depth at which measurements are made, and the measurements at a given depth may vary over time as hydraulic head conditions change. The various measurements also demonstrate that the magnitude and possibly the direction of horizontal flow are affected by the presence of the open borehole. Although there is a lack of agreement among the measurement techniques, these results could mean that effective characterization of horizontal flow in heterogeneous aquifers might be possible if data from many depth stations and from repeat measurements can be averaged over an extended time period. Complications related to vertical flow in the borehole highlights the importance of using background logging methods like vertical flowmeters or hydrophysical logging to characterize the borehole environment before horizontal-flow measurements are attempted. If vertical flow is present, a packer device may be needed to acquire definitive measurements of horizontal flow.

Because hydrophysical logging provides a complete depth profile of the borehole, a strength of this technique is in identifying horizontal- and vertical-flow zones in a well. Hydrophysical logging may be most applicable as a screening method. Horizontal- flow zones identified with the hydrophysical logging then could be evaluated with one of the point-measurement techniques for quantifying preferential flow zones and flow directions.

Additional research is needed to determine how measurements of flow in boreholes relate to flow in bedrock aquifers. The flowmeters may need to be evaluated under controlled laboratory conditions to determine which of the methods accurately measure ground-water velocities and flow directions. Additional research also is needed to investigate variations in flow direction with time, daily changes in velocity, velocity corrections for fractured bedrock aquifers and unconsolidated aquifers, and directional differences in individual wells for hydraulically separated flow zones.