GSA Annual Meeting in Seattle, Washington, USA - 2017

Paper No. 193-4
Presentation Time: 9:05 AM

COMPARATIVE STUDY OF METHODS FOR DELINEATING WELLHEAD PROTECTION AREA IN UNCONFINED COASTAL AQUIFER


LIU, Yue, WEISBROD, Noam and YAKIREVICH, Alexander, Department of Environmental Hydrology & Microbiology, Zuckerberg Institute for Water Research, Jacob Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, Sede Boqer Campus, Midreshet Ben-Gurion, 84990, Israel, liuy@post.bgu.ac.il

Wellhead protection area (WHPA) delineation is an important step in protecting groundwater resources extracted by production wells. Various delineation methods have been developed, ranging from simple analytical methods to complex numerical models. From a practical point of view, a better WHPA delineation method would be simple to use, while still preserving regional hydrogeological characteristics. Numerical modeling is regarded as the most accurate delineation method. However, in spite of the powerful simulating efficiency, the uncertainty of input parameters is an obstacle for numerical modeling results. We performed a comparative study on different WHPA delineation methods in order to examine their performance, reveal the essential factors for WHPA delineation, and address delineation uncertainty by stochastic modeling.

Comparison and uncertainty analysis for WHPA delineation were performed at two pumping sites located on the unconfined coastal aquifer in Israel: a single well and a wellfield consisting of eight pumping wells. Five delineation methods were used, with each method belonging to one of three categories: analytical, semi-analytical or numerical. Numerical modeling with MODFLOW-MODPATH was selected as the reference method. The comparison showed that the semi-analytical method fits the reference WHPA best, while analytical methods overestimated WHPA when implemented in the unconfined aquifer. This indicates that water balance and regional flow characteristics are two essential factors for delineating an appropriate WHPA. Through comparing the results from the wellfield with results from the single well, interferences between the wellfield wells were identified. These interferences significantly reduced the accuracy of the analytical methods when they were applied to the wellfield. Stochastic modeling with Latin Hypercube Sampling was employed for analyzing the uncertainty of the numerical method. Results showed that a series of probabilistic WHPAs, instead of a deterministic protection area, can be generated by considering the uncertainty in the input parameters. This can provide a more nuanced paradigm for designing WHPAs in heterogeneous aquifers.