Examples of Variable-Head Field Permeability Tests Used in Books: Given Interpretations and Correct Interpretations

Abstract

When a monitoring well is tested for permeability, three methods, with three types of graphs, may be used to analyze the data of the water column Z(t) versus time t. The three graphs provide a clear diagnosis, previously proven to be user-independent. According to experience, there is usually a systematic error H0 on the Z(t) data, which has different origins. Statistically, most plots of log Z(t) versus t are curved upward, a few are curved downward, and very few yield a straight line. Positive or negative values of H0 yield upward or downward curvatures, whereas a null piezometric error yields a straight line. This article presents an analysis of 21 sets of slug test data found in textbooks with the three diagnostic graphs and obtain three new findings. First, the textbooks ignore the method already proven and implemented in other countries since the 1980s. Second, the books selected biased data because their plots of log Z(t) versus t are either curved upward or straight, but no plot is curved downward. Third, the data of the first test of the group 3 theory are abnormal and do not correspond to usual field data with good equipment. In addition, one book presents a test in an aquitard as an example of test in an aquifer. The H0 value was easily found by the optimization method for all tests, and the derivative graph for 19 of the 21 tests, two data sets being too inaccurate to yield a good derivative graph.