3.1 Inter-well tracer tests

Once the tracer approach is chosen to investigate the speed and direction of groundwater flow, additional decisions still need to be made. Among the first is how the tracer will be introduced to the flow system. Depending on the specific questions being asked, tracers might be deployed on the surface, perhaps to examine infiltration or recharge rates, into sinkholes to determine groundwater flow in karst aquifers, or on the bed of lakes and streams to document groundwater-surface water interactions. It is probably safe to say that wells are the most common devices used to investigate flow systems in aquifers in part because they provide readily available conduits for tracer injections. Monitor wells are ubiquitous at sites undergoing hydrogeological investigations since they are used to obtain water level data that support Darcy’s Law calculations. A monitor well generally consists of a section of solid casing (a pipe with unperforated walls) and a screen (pipe with perforated or slotted walls). The pieces are assembled and positioned in a borehole with the screened portion on the bottom (Figure 5).

Schematic of a typical monitoring well
Figure 5 Schematic of a typical monitoring well (Bedient et al., 1994).

Monitor wells are commonly finished by filling the annulus space (the space between the pipe and the borehole walls) with permeable material around the screen (filter pack), and either backfill material or grout in the borehole above the screen. In some cases, the filter pack is separated from the upper annulus fill material by a seal of bentonite (clay) or cement to prevent contaminants from entering the screened portion of the well through the borehole. Wells can be designed to ‘see’ specific portions of an aquifer by tailoring the depth of installation and the screen length to the zone of interest. Screens vary in length from less than a meter long to nearly the entire depth of the well. Wells are an integral part of contaminated site investigations and so they are obvious tools for use in tracer studies. Perhaps the simplest experimental design involving tracers and wells is the injection of a tracer into an upstream well, and the monitoring of downstream well(s) for the subsequent appearance of the tracer (e.g., Clement et al., 1997) as shown in Figure 6. Such experiments are called interwell tracer tests, or sometimes natural gradient tracer tests when the groundwater flow is allowed to occur under naturally occurring conditions (i.e., without pumping).

Idealized conceptualization of an inter-well tracer test
Figure 6 Idealized conceptualization of an inter-well tracer test. A tracer (red) is injected at time zero, i.e., t = t0, where t0 = 0, and detected at a downstream well at a later time (t2). If a tracer is introduced as a pulse, the average velocity of the groundwater is found by timing the arrival of the middle of the pulse (i.e., the center of mass, which ideally coincides closely with the highest detected concentration) at the monitoring well. The arrival of the pulse at the downgradient monitoring well begins at a time between t1 and t2. The downgradient well may or may not be pumped. The center of mass of the pulse arrives at time t2. The entire breakthrough history of the pulse is recorded at some time after t2, as shown in the breakthrough curves above. Ideally, the seepage velocity can be calculated using time t2 (see equation).

The simplicity of inter-well tracer tests is offset by several problems related to the real-world complexity of aquifers. First, to ensure that the tracer will not be influenced by density driven flow or be entrained by small scale geologic features (strata, lenses), large dilute source volumes in the aquifer must be established; these may not be simple and inexpensive to design or create. Second, more than two wells are likely to be required at close spacings, which makes these tests potentially expensive. Third, if the flow system is not already reasonably well understood, the tracer may be carried along a path that misses even a closely spaced well network — for example by following a path beneath the wells due to unanticipated downward vertical flow, as illustrated for the spill depicted in Figure 7, or by breaking apart into disconnected plumes (Sudicky and Cherry, 1979). Finally, the time required for a tracer to move through the monitor well network may be many days. In extreme cases, many months may be needed for a test to run to completion. Throughout this time, water sampling and analysis is required to properly identify the tracer center of mass or peak arrival time. This requirement can also be expensive to satisfy, although recent in situ sensor developments promise to minimize these expenses in the future by automating the tracer monitoring task. Regardless, the time from the onset of a test to its completion may be quite long, delaying decisions that might avert risk.

Illustration showing a multi-well monitor network failing to intersect a plume due to inadequate well spacing and a sinking plume
Figure 7Illustration showing a multi-well monitor network failing to intersect a plume due to inadequate well spacing and a sinking plume. This scenario could apply to a spill, as illustrated, or tracer tests aimed at characterizing the aquifer — and failing.


Groundwater Velocity Copyright © 2020 by J.F. Devlin. All Rights Reserved.