Townley, L.R., and Barr, A.D. (1992), Two-dimensional regional groundwater flow near Koongarra, IN: Alligator Rivers Analogue Project, 3rd Annual Report 1990-91, ANSTO, 165-193.

Modeling of groundwater flow in the Koongarra region has previously been described by a number of ARAP participants. Raffensperger and Garven (1989) modelled a 2 km cross-section through the No.1 orebody, using an assumed water table as the upper boundary, Koongarra Creek as the downgradient boundary, and a range of conductivity assumptions for distinct hydrogeological zones, including the Koongarra Fault. Lever and Morris (1991) presented a similar study, with emphasis on travel times from the orebody to the land surface boundary. Townley and Trefry (1991) extended the length of the vertical section to 26 km, to verify that Koongarra Creek must be an effective downgradient boundary, and then presented three- dimensional simulations beneath a 3 km square, again showing sensitivity to the conductivity of the Fault. The three-dimensional modelling was later repeated at a finer resolution, but subsequent discussions concerning the role of the Koongarra Fault have made that work obsolete.

Recent field measurements, and considerable discussion at the ARAP Hydrology Workshop in Tucson in October 1991, have resulted in consensus on a number of important issues. First, there is general agreement that most of the horizontal groundwater flow (with the exception of shallow flows in surficial sands and gravels) occurs at or below the bottom of the weathered zone, and that the system can be viewed as a confined aquifer with essentially horizontal flow below the base of weathering. Second, there is little evidence for significant flow across the Koongarra Fault. Third, there appears to be a highly conductive zone to the (Mine Grid) southeast of the No.1 orebody. Fourth, it appears that directions of flow are influenced by the orientation of planes of schistosity, or by fractures sub-parallel to these planes, thus inducing a significant degree of anisotropy in an effective hydraulic conductivity tensor. Fifth, there is evidence from standing water level measurements that the gradient in heads sometimes reverses towards the Fault near the orebody.

The purpose of this report is to describe attempts to take into account most of the features of the current consensus. Here we present the results of two- dimensional modelling in plan of the region near the orebody. Of the five issues listed above, all but the zone of enhanced conductivity (an issue of regional heterogeneity) have been addressed. The modelling has been performed using AQUIFEM-N (Townley, 1992), a finite element groundwater flow model. Figure 1 shows the 3 km square region which was used previously by Townley and Trefry (1991) for three-dimensional modelling, and identifies the boundary of the two-dimensional region for which a two- dimensional plan model has been developed. Before proceeding to present model results, we discuss preliminary analyses of depths and flows in Koongarra Creek and of regional groundwater levels, both of which affect the way we have set up our model.