A knowledge of the variation of horizontal hydraulic conductivity with vertical position, K(z), is important in understanding the transport and dispersive properties of aquifers. Using an impeller meter to measure the discharge distribution in a screened well while pumping at a constant rate is a promising technique for obtaining the K(z) function. Such an application is described herein, and the resulting K(z) functions are compared with those obtained previously using tracer tests and multilevel slug tests. Impeller meter data were the most convenient to obtain, and tracer data the most difficult. The K(z) functions obtained by the three methods were not identical but quite similar overall. This similarity between both borehole tests and the larger-scale tracer test showed that nonstationary hydraulic conductivity trends, in a stochastic hydrologic sense, exist in the test aquifer. The impeller meter method was better able to detect the higher K layers than was the multilevel slug approach. Overall, the results suggest that a practical strategy for "fitting" impeller meter, tracer, or multilevel slug test data to a given aquifer is to use the selected testing procedure to obtain a dimensionless K/• distribution and then a standard pumping test to measure •. Combining both types of information enables dimensional values for K(z) to be calculated. In low permeability aquifers or near the bottom of a test well the fluid velocity due to pumping may be below the stall velocity of an impeller. Thus there is a definite need for the commercial development of more sensitive flow-measuring devices such as heat pulse flowmeters (Hess, 1986), which will extend the resolution of this field method. 1677

The U.S. Geological Survey used a recently developed heat-pulse flowmeter to measure very slow borehole axial water velocities in granitic rock at a site near Lac du Bonnet, Manitoba, Canada. The flowmeter was used with other geophysical measurements to locate and identify hydraulically conducting fractures contributing to the very slow vertical water flow in the two boreholes selected for study. The heat-pulse flowmeter has no moving parts and operates on the tag–trace principle. It is an improved version of the flowmeter developed by the Water Research Centre in England in 1975. The U.S. Geological Survey's heat-pulse flowmeter has a flow-measuring range in water of 0.06–6 m/min, and can resolve velocity differences as slow as 0.01 m/min. This is an order of magnitude slower than the stall speed of spinner flowmeters. The flowmeter is 1.16 m long and 44 mm in diameter. It was calibrated in columns of 76 and 152 mm diameter, to correspond to the boreholes studied. The heat-pulse flowmeter system is evaluated, and problems peculiar to the measurement of very slow axial water velocities in boreholes are discussed. Key words: flowmeter, borehole flow, low flow, borehole geophysics.

The distribution of fracture permeability in granitic rocks was investigated by measuring the distribution of vertical flow in boreholes during periods of steady pumping. Pumping tests were conducted at two sites chosen to provide examples of moderately fractured rocks near Mirror Lake, New Hampshire and intensely fractured rocks near Oracle, Arizona. A sensitive heat‐pulse flowmeter was used for accurate measurements of vertical flow as low as 0.2 liter per minute. Although boreholes were spaced at intervals ranging from 10 to 50 meters, acoustic televiewer logs showed little direct continuity of individual fractures from borehole to borehole in either the moderately fractured rocks or intensely fractured rocks. Results indicated that nearly all inflow and outflow to boreholes occurred by means of one or two discrete fractures in both cases. These fractures did not appear very different from other prominent fractures indicated on televiewer and resistivity logs for these boreholes. Hydraulic connections between boreholes apparently were composed of conduits formed by the most permeable portions of intersecting fractures. Most flow in the moderately fractured rocks occurred at isolated fractures at a depth of about 45 meters indicating a nearly horizontal zone of fracture permeability composed of orthogonal, steeply dipping fractures. Previous studies have identified a zone of horizontal permeability in the lower part of the boreholes in the intensely fractured rocks, but flowmeter tests indicated that flow also entered and exited individual boreholes by means of one or two steeply dipping fractures. These results indicate zones of fracture permeability in crystalline rocks are composed of irregular conduits that cannot be approximated by planar fractures of uniform aperture, and that the orientation of permeability zones may be unrelated to the orientation of individual fractures within those zones.

The U.S. Geological Survey has developed an all electronic, thermal-pulse flowmeter for measuring slow axial-velocities of water in boreholes. The flowmeter has no moving parts, but senses the movement of water (or any other fluid) by a thermal-tag/trace-time technique. In the configuration shown in this report the flowmeter can measure water velocities ranging between about 0.1 to 20 feet per minute, can resolve water velocity differences as small as 0.03 foot per minute, and distinguishes between upward and downward flow. Two diameters of interchangeable flow sensors are described; the smaller sensor is 1.75 inches in diameter including the collapsed centralizer and the larger is 2.75 inches in diameter with collapsed centralizer. The flowmeter probe is 48 inches long and has been tested and used in tubes and boreholes with diameters ranging from 2 to 10 inches. The flowmeter also should be useful in measuring slow velocity flow in larger diameter boreholes, though for best accuracy it should to be calibrated for flow at the diameter of the borehole in which it is to be used. The flowmeter probe has been designed to withstand a water pressure depth of 10,000 feet. The report includes a description of the operation of the flowmeter, functional diagrams, mechanical drawings, and electronic schematics for both the flowmeter probe and surface electronics. Lists of parts and materials, and instructions for fabrication, assembly, and calibration also are included. These diagrams and lists should be adequate to permit the construction and use of a thermal-pulse flowmeter system by those skilled in machining and fabrication, electronics fabrication, and borehole metrology.