Push-Pull Tests for Site Characterization

There is a continuing need for quantitative information about the subsurface environment. Applications include developing and managing new or existing water supplies, storm water management, energy storage and recovery, environmental remediation, artificial recharge, gas and oil production, carbon sequestration, and many others. The spectrum of processes that might be of interest include essentially all the physical, chemical, and biological processes that are operational in any natural or engineered environment. Physical processes might include fluid flow by advection and mixing/spreading processes such as dispersion, conduction, and convection that apply to transport of matter and energy. Chemical processes might include the wide range of chemical reactions that are possible in water/mineral/atmosphere systems such as acid/base, precipitation/dissolution, sorption, surface complexation, and oxidation/reduction reactions. Biological processes might include any of the diverse metabolic activities displayed by indigenous microorganisms as they obtain energy and nutrients for growth and include processes related to nutrient cycling, detoxification of contaminants, and many others. Of course all of these processes are coupled so that physical properties will influence fluid flow, which controls the flow of substrates that in turn influence microbial growth, or solutes that participate in chemical reactions. There is no satisfactory term to describe the coupling among all these processes; “biogeochemical processes” will be used here with the explicit understanding that physical processes are also included.

Push-pull tests can be conducted using any facility or device that makes it possible to inject and extract pore fluids from the formation (Fig. 1.2). Thus, tests may be conducted in open boreholes, screened intervals of conventional monitoring wells, sampling ports of multi-level monitoring wells, drive-points wells, piezometers, or through drive points inserted into the sidewalls of an open excavation. Tests may be conducted above or below the water table, at any depth, and in any type of geologic formation. Tests may be conducted in terrestrial subsurface environments or in saturated sediments that lie beneath lakes, rivers, estuaries, the sea floor, etc. Push-pull tests are most suitable for tests conducted in porous media where flow is laminar but push-pull tests have also been conducted in deep lakes and other surface water bodies where weak turbulent mixing limits dilution losses of injected test solutions. Small-scale push-pull tests have been successfully conducted using the simplest equipment, such as plastic syringes injecting test solutions through “well screens” formed of syringe needles manually inserted into saturated or submerged sediments.

Research related to what we now call push-pull tests began in the late 1960s although significant development of the method has occurred since the late 1990s. A type of single-well injection/extraction test was first used in a qualitative way by Sternau et al. (1967) to study the degree of mixing of injected water with background groundwater. The theory of dispersive transport in the radial flow field near an injection well was first presented by Hoopes and Harleman (1967) who were investigating processes occurring during the recharge and disposal of injected liquid wastes and wastewater. They solved the advection-dispersion equation including linear sorption to predict tracer concentrations as a function of radial distance and time from the injection well. They also conducted a large-scale (~6 m) laboratory push-pull test in a semi-cylindrical sand box to compare experimental results with model predictions.

Leap and Kaplan (1988) and Hall et al. (1991) presented a type of push-pull test for determining regional groundwater velocity and effective porosity if the hydraulic conductivity of the aquifer and local the hydraulic gradient are known. The hydraulic conductivity could be determined by a pumping test conducted in the same well at the same time as the push-pull tracer test described here; the hydraulic gradient could be determined from water level measurements in a set of nearby wells surrounding and including the push-pull test well. The procedure involves injecting a constant-concentration test solution containing a nonreactive tracer into the aquifer using a single well, allowing the test solution to drift downgradient with the regional groundwater flow, and then extracting the tracer solution/groundwater mixture from the same well by continuous pumping to determine the temporal displacement of the tracer center of mass. The basic equations (using the notation of Hall et al. 1991) are: where q is the apparent groundwater (Darcy) velocity, n is effective porosity, Q is the extraction pumping rate, t is the time elapsed from the start of extraction pumping until the centroid of the tracer mass has been extracted, b is the aquifer saturated thickness, d is the elapsed time from the end of tracer injection until the centroid of the tracer mass is extracted (drift time + t), K is the saturated hydraulic conductivity, and I is the local hydraulic gradient. Obviously, uncertainties in computed values of q and n will reflect uncertainties in values of K and I.