To avoid the problems associated with wells, filter packs, and well screens, techniques that allow sensors to be in direct contact with an aquifer matrix were developed. Two of the direct-contact in situ probe types that have received attention on the strength of their performances in field tests are the In Situ Permeable Flow Sensor (ISPFS), later renamed Hydrotechnics™ groundwater velocity sensors, and the point velocity probe (PVP) as shown in Figure 13a and b.

In Situ Permeable Flow Sensor

The ISPFS consists of a nearly one-meter long cylinder with a surface that is uniformly heated and monitored for temperature with a dense network of sensors as shown in Figure 13a (Ballard, 1996). The instrument is installed in dedicated boreholes established in unconsolidated, non-cohesive porous media. The disturbed sediments collapse against the probe leaving no open channels for flow in the borehole annulus. Groundwater flows up to and against the probe, cooling the upgradient side compared to the downgradient side. Patterns of temperature on the overall surface are then related mathematically to groundwater velocity (i.e., speed and direction). In practice, the sophistication and associated cost of the instrument restricts its use to large, well-funded projects. The length of the probe limits its ability to discern fine hydrogeologic and plume features, and is best suited for use in relatively thick, homogeneous aquifers or strata. Since the probe uses heat as a tracer, it is sensitive to density-induced flow and variations in thermal conductivity in the surrounding formation. The technology has been associated with spurious detections of vertical flow — particularly when installed with a filter pack or when installed across sediment interfaces with different thermal conductivities (Ballard et al., 1996; Su et al., 2006).

Point Velocity Probe (PVP)

The PVP, like the ISPFS, is a cylindrical probe installed in a dedicated borehole. However, the PVP senses water movement by recording the breakthrough curve of a tracer rather than a temperature distribution. In principle, the PVP tracer could be a radioisotope, a dye, heat, a conductive solution, or even deionized water — anything that could be detected in situ with a small sensor (Labaky et al., 2007). To date, the tracer most commonly used has been a dilute (<1 g/L) saline solution, which provides an electrical conductivity signal well above background in freshwater. The probe operates by releasing a small volume (<1 mL) of the tracer, which is carried by groundwater over the probe’s surface to conductivity detectors (Figure 13b). With at least two such detectors on the probe, both speed and direction of the horizontal water movement can be calculated at the centimeter scale. If detectors are placed above and below the tracer release point, vertical flow can also be quantified (Gibson and Devlin, 2018). The probe can be mounted in a multilevel arrangement, permitting as many as 7 probes to instrument a single borehole. The probes have performed well in both sandy, and glacial outwash aquifers. In deposits with elevated clay content, the collapse of the borehole against the probe body tends to be incomplete and reliable data cannot be collected.

Groundwater Variability Probe

A variation on the heat pulse flowmeter design, for use in near surface (<1 m depth), dedicated, uncased boreholes is the Groundwater Variability Probe (GVP) (Crawford et al., 2016). This device is installed in direct contact with aquifer material, and like the HPF is fabricated with several sensors arranged in a ring around a central tracer release point. In principle, data interpretation is simple and based on the time of arrival of the tracer at the sensors, with the affected sensors indicating the flow direction. However, the device is susceptible to biases if the sediment between the sensors is disturbed during installation.