TRACING TRANSPORT AND WATER QUALITY CHANGES AT A GROUNDWATER BANKING FACILITY IN CALIFORNIA'S CENTRAL VALLEY

Subsurface storage of excess winter/spring flows or imported water in overdrafted groundwater basins is playing a growing role in the management of a strained water supply in California. Artificial recharge facilities typically consist of spreading ponds, temporary dams on streams, or injection wells, where almost a billion cubic meters of water is recharged per year statewide. We used an inert dissolved gas, sulfur hexafluoride, to tag about 125,000 m³ of surface water at a groundwater banking facility near Stockton, California, and trace its movement in 11 monitoring wells and 4 production wells. The objective of the study was to identify and quantify changes in water quality during recharge and subsurface transport. Applied water in this study is very low TDS river water from Sierra Nevada runoff, and the facility has only been in operation for about three years. In contrast, large artificial recharge operations in southern California apply relatively high TDS water at long-established recharge areas. Analysis of tritium-helium groundwater age, fraction pre-modern, O and H isotopes in water, and the SF₆ tracer allow estimation of the bulk flow rate, dilution of recent recharge with ambient groundwater, and travel time of the most recently recharged water. Water quality is examined using measurements of major ions, trace metals, trace-level volatile and semi-volatile organic compounds, dissolved oxygen, and dissolved carbon in both surface water and groundwater. A number of trends appear in the geochemical parameter data that are indicative of recharge water. Both Cl^- and DIC, which follows a clear mixing trend between low DIC, high d¹³C recharge water and high DIC, low d¹³C ambient groundwater, serve as recharge water tracers. Exchange of Ca²⁺ and Mg²⁺ for Na⁺ and K⁺ is evident in groundwater samples collected from wells that the Cl^-, DIC, and SF₆ tracer data indicate are effected by recharge water. Moreover, the concentrations of some trace elements (e.g., arsenic) are elevated in these samples in comparison to background. PHREEQC simulations employing an ion exchanger and an active hydrous ferric oxide surface provide some insights into the possible mechanisms responsible for the observed trends.