ESTIMATION OF HYDRAULIC CONDUCTIVITY FROM SLUG TESTS

Through comparisons of hydraulic conductivities estimated by Cooper-Bredehoeft-Papadopulos, Hvorslev, and Bouwer and Rice slug test methods to the actual conductivities used in numerical simulations, it is found that storativity may have more significant effect on the accuracy of conductivity estimation in a slug test than traditionally realized. In addition, when storativity and/or conductivity are relatively high in an aquifer, particularly in an unconfined aquifer, where specific yield ranges from 0.1 to 0.3, a curved or “multi- segment” line exist in the traditional semi-log H/Ho vs. time plot for a slug test. The curved H/Ho vs. time plot makes it difficult to extract an accurate conductivity value using the Hvorslev , and Bouwer and Rice methods. In a medium to high conductivity aquifer(>5 meters/day), correct extraction of the conductivities from well data will require multiple water level data collection shortly after the slug injection, preferably during the first second after slug injection. This requirement will be difficult to realize in a field slug test and might account for the lower conductivity estimate from slug tests than estimates from pumping tests. Traditional approach of using the “late segment” of H/Ho vs. time plot for extracting conductivity in a high conductivity aquifer might result in an unrealistically low conductivity. Data from prior field studies of a Texas site, where multi-type tests for extracting hydraulic conductivity were conducted in an alluvial aquifer, were used to examine some of these conclusions. Based on the relationship between conductivities extracted by Hvorslev, and Bouwer and Rice methods and the actual conductivities used in the numerical simulation of a slug test in a fully penetrating well, the extracted field conductivities by those methods can be corrected if the storativity of an aquifer can be obtained in a fully penetrating well.