SMART PUMP AND TREAT

Lawrence Livermore National Laboratory (LLNL) is approaching the final phase of the Superfund decision-making process for site restoration a nd will soon initiate full-scale clean-up. Despite some well-publicize d perceived failings of the pump and treat approach, we have concluded that intelligent application of this strategy is the best choice for ground water restoration at LLNL. Our proposed approach differs suffic iently from the pump-and-treat methods implemented at other sites that we call it ''smart'' pump and treat. Smart pump and treat consists of four distinct, but interrelated, elements: three pre-remediation stra tegies and one applying to the active management of the extraction pro cess. Together, these techniques constitute an integrated program that embodies crucial aspects of contaminant flow and transport to speed u p the remediation of contaminated aquifers. The four elements are: (1) a spatially detailed site characterization, linked with regional hydr ogeologic simulations; (2) directed extraction, where the extraction a nd recharge locations are controlled by field-determined hydrogeologic parameters; (3) field-validated simulations that match the complexity of the collected data; and (4) adaptive pumping and reinjection where spatial positions and rates vary with time. Together, these technique s minimize the cost and the time to reach regulatory-directed cleanup goals and maximize the rate of contaminant removal.