A NONLINEAR MODEL FOR ANALYSIS OF SLUG-TEST DATA

While doing slug tests in high permeability aquifers, we have consiste ntly seen deviations from the expected response of linear theoretical models. Normalized curves do not coincide for various initial heads, a s would be predicted by linear theories, and are shifted to larger tim es for higher initial hi:ads. We have developed a general nonlinear mo del based on the Navier-Stokes equation, nonlinear frictional loss, no n-Darcian flow, acceleration effects, radius changes in the well bore. and a Hvorslev model for the aquifer, which explains these data featu res. The model produces a very good fit for both oscillatory and nonos cillatory field data, using a single set of physical parameters to pre dict the field data for various initial displacements at a given well. This is in contrast to linear models which have a systematic lack of fit and indicate that hydraulic conductivity varies with the initial d isplacement. We recommend multiple slug tests with a considerable vari ation in initial head displacement to evaluate the possible presence o f nonlinear effects. Our conclusion is that the nonlinear model presen ted here is an excellent tool to analyze slug tests, covering the rang e from the underdamped region to the overdamped region.