Slug tests in unconfined formations: An assessment of the bouwer and rice technique

Abstract

The slug test is one of the most common techniques for the in situ estimation of hydraulic conductivity in unconfined flow systems. Recently, a mathematical model describing the flow of ground water in response to a slug test in an unconfined flow system has been proposed. This model incorporates the effects of partial penetration, anisotropy, an upper constant-head boundary, and, in its most complete form, well skins of either higher or lower permeability than the formation itself. This model is useful in identifying conditions when conventional approaches (i.e., the Bouwer and Rice model) introduce large errors into parameter estimates. For slug tests performed in homogeneous, isotropic formations that would be classified as aquifers, the Bouwer and Rice model provides estimates within 30% of actual field values. In less-permeable, clay-rich formations, however, estimates may overpredict formation conductivity by more than 100%. The Bouwer and Rice model introduces the largest error (can easily exceed an order of magnitude) in the presence of a low-permeability skin. Uncertainty about anisotropy can also be the source of considerable error. The semianalytical solution to the mathematical model described here can be employed for parameter estimation under conditions when the Bouwer and Rice model introduces unacceptably large errors into parameter estimates. This solution can be rapidly evaluated, allowing easy incorporation into an automated well-test analysis package and/or ready generation of type curves.