Wellskins and slug tests: whereʹs the bias?

Pumping tests in an outwash sand at the Camp Dodge Site give hydraulic conductivities (K) approximately seven times greater than conventional slug tests in the same wells. To determine if this difference is caused by skin bias, we slug tested three sets of wells, each in a progressively greater stage of development. Results were analyzed with both the conventional Bouwer–Rice method and the deconvolution method, which quantifies the skin and eliminates its effects. In 12 undeveloped wells the average skin is +4.0, causing underestimation of conventional slug-test K (Bouwer–Rice method) by approximately a factor of 2 relative to the deconvolution method. In seven nominally developed wells the skin averages just +0.34, and the Bouwer–Rice method gives K within 10% of that calculated with the deconvolution method. The Bouwer–Rice K in this group is also within 5% of that measured by natural-gradient tracer tests at the same site. In 12 intensely developed wells the average skin is <−0.82, consistent with an average skin of −1.7 measured during single-well pumping tests. At this site the maximum possible skin bias is much smaller than the difference between slug and pumping-test Ks. Moreover, the difference in K persists even in intensely developed wells with negative skins. Therefore, positive wellskins do not cause the difference in K between pumping and slug tests at this site.