Evaluation of simple methods for estimating contaminant removal by flushing

The purpose of this work was to briefly outline and evaluate simple methods used for estimating the volume of water and the time required to remove contaminants from the surface by flushing (e.g., "pump-and-treat," "soil washing"). The methods are based on three general approaches to treating flow and transport: perfectly mixed flow, hydraulic, and advective-dispersive. Data obtained from a small field experiment designed to evaluate aquifer flushing was used to illustrate the performance of the estimation methods. The methods based on the hydraulic and ideal advective-dispersive transport approaches will generally provide estimates that underpredict the actual time. The magnitude of the underprediction will depend, in part, on the degree of nonideal behavior influencing contaminant removal (e.g., heterogeneity, rate-limited mass transfer) and on the ratio of target to initial contaminant concentration. The perfectly mixed flow reactor approach, because of its asymptotic removal curve, may be useful in providing rough approximations of required pore volumes and times. However, it cannot be overemphasized that all estimation technique are prone to failure as long as the conceptual models upon which they are based do not accurately represent field-scale contaminant transport.