model can be used to predict changes in direction and velocity of flow in the UFA, and to evaluate the possible consequences of these changes on existing wells (see Bradbury and Krohelski, 1999 for a similar approach). The aggregate head buildup is needed also to predict areas where more detailed studies of the potential for hydraulic fracturing (see chapter 4) should be done. An elevated potentiometric surface that could induce fracturing and upward migration of recharged water could result from the superposition of head buildups of multiple ASR wells.
The regional or telescoped models of sub-regions can be used to compare and evaluate alternative ASR system designs. For example, regional effects of an ASR system that is concentrated in a small geographical area, such as the planned deployment of 100 ASR wells along the northern half of Lake Okeechobee, could be compared to a system of similar volume that is more geographically dispersed. Various recharge and withdrawal scenarios can be easily simulated with the regional model, and the model results can be used to optimize the ASR system while minimizing deleterious impacts (e.g., using the approach described by Wagner, 1995).
The regional study should include development of a formal procedure for the siting of ASR wells. For the few exploratory ASR wells included in the pilot project, preliminary siting was based solely on location of a water source for recharge and proximity of a water body for receipt and conveyance of drilling and testing fluids. However, for full implementation of the ASR proposal, well siting should include evaluation of other factors (e.g., anticipated well capacity, proximity to other users, etc.) as well. It is important that the site selection process be explicit, repeatable, and based on best available data and selection methods.
The proposed Phase 1 siting analyses include simple overlay of GIS data to evaluate site suitability. This approach may be adequate for the local pilots, but a more formal, multi-criteria siting approach should be designed and tested during the pilot studies that would facilitate the much larger and more complex problem of siting many wells over the region. Promising approaches range from structured scoring methods such as Analytical Hierarchy Process and Knowledge Bases (Reynolds et al., 1999) to more complex optimization models using mathematical programming or heuristic search procedures (e.g. Haight et al., 2000; Matthews et al., 1998; Miller, 1996; Murray et al., 1998).
In conclusion, the pilot projects provide a valuable means for acquiring detailed information on ASR performance at a few specific sites. However, even if all the sites tested prove successful, they will not by themselves demonstrate the feasibility of ASR implementation regionally at the scale of 1.7 billion gpd (6.3 million m3/day). A regional study involving construction of a regional flow model is an invaluable and indispensable tool to assess feasibility of ASR at the proposed scale.