In partially penetrating slug tests, hydraulic conductivity (K) estimates might not necessarily be valid because of vertical flows in heterogeneous formations. We assess the error in hypothetical stratified formations by numerical sensitivity analysis, and propose an effective method for compensation by incorporating two types of casing configuration (piezometer and tube). The hypothetical stratified formation consists of completely horizontal layers, each 1 m thick; the permeability is different between, but not within, layers. In this study, conductivity estimates in the piezometer and tube methods are calculated by assigning various patterns of conductivity to the test, upper, and lower layers: K T , K U , and K L . The effect of vertical flow becomes significant when K T is small relative to K U or K L , and K L is more important than K U because the base of the borehole is open to the lower formation. The conductivity ratios (estimate over actual value) are treated as approximately linearly dependent on logarithms of K T /K U and K T /K L , so that conductivity estimates can be straightforwardly derived from one piezometer measurement and two tube measurements at the top and bottom of the screen. The linear relations are evaluated and constant parameters are determined under specific conditions. This study also recommends alternating piezometer and tube methods in the drilling procedure because the actual variation of K with depth is larger than that found using isolated measurements, as shown in a field study of alluvial fan gravel deposits in Sapporo, Japan.
KeywordsSlug tests, Partially penetrating well, Hydraulic conductivity, Sensitivity analysis, Numerical simulation,
Heterogeneity 1.IntroductionSlug tests are employed in both practical engineering and scientific studies to determine in situ hydraulic conductivity, K, by measuring the recovery of head in a single borehole after artificially raising or lowering the water level in the well. Although slug tests are sensitive to artificial and natural conditions in the well (e.g., Black, 2010), they are useful because they are simpler, cost less, and require less equipment than multiple pumping tests. It is also common for slug tests to have a logistical advantage, especially in groundwater contamination investigations. One reason is that slug tests can be conducted without removing or adding water, by lowering a solid piece of metal, called a slug, into a well, which does not disturb the site of suspected contamination. Another reason is that slug test results are related mainly to permeability on a small scale, and thus local information is directly relevant to preferential flow paths for contamination plumes. In addition, a series of slug tests at multiple depths aids in interpreting vertical variations of K.When such vertical profiles are interpolated among different locations, three dimensional distributions of K can be obtained for an individual site. Various previous studies have shown the spatial variability in K obtained from slug test an...