Optimal Management of An Aquifer Storage and Recovery (ASR) Program in a Subsidence Area, Lancaster, California
A ground-water flow model of the Lancaster region was developed and linked to linear optimization tools to identify optimal injection and extraction timing and distribution among 16 existing and 13 proposed wells. Previous GPS surveys and InSAR results showed that northern Lancaster is most susceptible to subsidence despite greater historical drawdowns in the southern area. The objective of the optimization was to maximize minimum head, which effectively minimized loss of well production capacity in the southern area. Constraints were imposed on maximum and minimum heads (a surrogate for subsidence). Injection and extraction capacities, availability of injection water, groundwater demand, and the distribution of stresses between model layers were also constrained.
The ASR program management scenarios optimized ranged from maintaining present practices (no injection) to allowing six months of injection per year in all 29 wells. Simulation/optimization model results suggest that optimal use of existing wells cannot maintain water levels in the southern area; phased installation of the proposed wells would be needed within a ten-year period to maintain production capacity and avoid triggering additional subsidence. Results also suggest that continued use of wells in the subsidence area, made possible by injection, would be needed to maintain extraction capacities in the southern area.