Numerical modeling techniques are used to examine the hydrogeologic, design, and management factors governing the recovery efficiency of subsurface fresh-water storage. The modeling approach permitted many combinations of conditions to be studied. A sensitivity analysis was used that consisted of varying certain parameters while keeping constant as many other parameters or processes as possible. The results show that a loss of recovery efficiency resulted from: 1) processes causing mixing of injected fresh water with native saline water (hydrodynamic dispersion); 2) processes or conditions causing the irreversible displacement of the injected fresh water with respect to the well (buoyancy stratification and background hydraulic gradients); or 3) processes or procedures causing injection and withdrawal flow patterns to be dissimilar (dissimilar injection and withdrawal schedules in multiple-well systems). Other results indicated that recovery efficiency improved considerably during the first several successive cycles, provided that each recovery phase ended whgen the chloride concentration of withdrawn water exceeded established criteria for potability (usually 250 milligrams per liter). Other findings were that fresh water injected into highly permeable or highly saline aquifers would buoy rapidly with a deleterious effect on recovery efficiency. -Author
