No abstract

No abstract

The results and interpretation of five induced‐gradient tracer tests performed at five different average interborehole fluid velocities in a single fracture in monzonitic gneiss are described. The experiments were conducted using radioactive 82Br and a fluorescent dye as conservative tracers where the tracers were pulse injected into radial convergent and injection‐withdrawal flow fields. The flow fields were established between straddle packers isolating the fracture in three boreholes over distances of 12.7–29.8 m. The tracer breakthrough curves were determined from samples of the withdrawn groundwater and were interpreted using residence time distribution (RTD) theory and two deterministic simulation models. The RTD curves of the tracer experiments were interpreted by fitting to the field data a simple advection‐dispersion model and an advection‐dispersion model with transient solute storage in immobile fluid zones. Both models consider the different flow field geometries associated with injection‐withdrawal and radial convergent tests. Comparison of the fits obtained by the simulation models suggest that the initial period of solute transport in single fractures is advection dominated and with increasing tracer residence time or decreasing fluid velocity, transport progresses toward more Fickian‐like behavior. During the advective‐dominated period, the transient solute storage model is shown to adequately describe the asymmetries and long tails characteristic of the fracture RTDs. Interpretation of the tracer experiments using both simulation models further suggests that induced‐gradient tracer experiments are likely to underestimate the dispersive characteristics of single fractures under natural flow conditions.

Ground water inflows to drifts ranging from 700 to 1615 m below ground surface at the Con Mine, Yellowknife, Northwest Territories, Canada, were used to study deep hydrogeological flow regimes in Shield terrain. Salinity trends are due to mixing between low‐TDS ground water and deep Ca(Na)‐C1 brines (>290 g/L) likely derived from Devonian sea water. C1‐−δ18O relationships demonstrate that all inflows are a mixture of three distinct components: modern meteoric ground water (δ18O ∼−18.9 ± 0.1%o), brine (δ18O ∼−10%o), and an isotopically depleted water (δ18O ∼−28%o). The origin of this third endmember is attributed to glacial melt water injected into the subsurface during ablation of the Laurentide Ice Sheet at ca. 10 ka. A mechanism is proposed where high hydrostatic pressure in the ablation zone imposes strong downward gradients beneath the ice sheet margin. Numerical simulation with the SWIFT II finite‐difference code recreates the observed salinity gradients within a modeled 50‐year interval, corresponding with the rate of retreat of the ice sheet across the landscape at this time. The persistence of this melt water in the subsurface for some 10,000 years following retreat of the ice and decay of the steep hydraulic gradients highlights the importance of gradient, in addition to permeability, as a major control on ground water flow and transport in deep crystalline settings.

A field example of measuring the dispersive properties of a single fracture in fractured plutonic rock is presented. The experimental technique involves injecting a slug of conservative tracer into a “steady” groundwater flow field established between a pumping and recharging borehole and monitoring the tracer breakthrough by sampling the withdrawal water directly. The breakthrough curves from two experiments were analyzed with a model which describes the flow field geometry either analytically or numerically and solves for hydrodynamic dispersion analytically. A longitudinal dispersivity of 1.40 m was estimated by fitting the model to each set of field data. The magnitude of the dispersion was determined to be independent of dispersive effects created by flow through the borehole instrumentation and thought to be purely hydrodynamic in nature.