Interpreting Slug Tests with Large Data Sets

Abstract

Abstract Pressure transducers are frequently used to monitor slug tests, and to collect a data set for height Z of water column versus time t. Direct application of the data to draw a velocity graph usually produces a wide scatter, thus limiting the usefulness of the velocity graph method. This paper proposes to interpret the data set in five steps. First, a Z(t) plot is drawn to assess the uncertainty in the Z data. Second, a limited number of data points are selected to reduce the uncertainty to about ±10 % in the velocity graph. Third, the velocity graph is drawn to obtain the hydraulic conductivity k and the piezometric error H0, if any, giving the piezometric level for the test. Fourth, the graph of ln(Z−H0) versus t is drawn to verify its linearity and check the k value of the velocity graph. Fifth, the graph of Z versus log10(t) can be used in a curve fitting process to yield directly both H0 and k, thus providing a second method to interpret the data set, which may be the only one possible when the Z(t) data are inaccurate. Examples are provided to illustrate the five steps and common problems.