Numerical Evaluation of the Constant-Head Borehole Permeameter Method for Stormwater Infiltration Design

Abstract

The current standard of practice for sizing stormwater infiltration facilities typically relies on one-dimensional (1D) test methods that do not account for the full dynamics of groundwater flow, including lateral flow and capillary flow. Although some agencies allow methods that do account for lateral and capillary flow, these methods are relatively small-scale and may not replicate the effects of soil layering below a full-scale infiltration facility. The uncased and cased methods evaluated in this study account for lateral and capillary flow and can be used to evaluate both small-scale and large-scale infiltration tests in a broad range of test facilities, including excavated pits, uncased shallow boreholes, and deep cased wells. This study provides numerically calibrated shape factors for both glacially consolidated and normally consolidated soils that are generally considered suitable for stormwater infiltration [saturated hydraulic conductivity (Ks)>0.1 m/day]. Soil sorptive numbers (α*), which quantify the degree of soil capillarity, were also calculated for the 10 representative soils evaluated in this study. Using the α* estimates and calibrated shape factors developed for this study, these methods can provide estimates of Ks with a bias range of 0.87–1.13 and an average bias of 0.99. Bias is the calculated Ks based on the constant-head borehole permeameter method divided by the specified Ks used in the numerical model. As demonstrated in this study, the constant-head borehole permeameter methods are well-suited for predicting the flow capacity of full-scale infiltration facilities. Current methods for sizing stormwater infiltration facilities rely on either grain-size analyses or infiltration testing. Grain-size methods provide only an approximate estimate of soil permeability, and current methods for infiltration testing have significant limitations. For example, some methods assume vertical flow from the test facility and do not account for the full dynamics of groundwater flow, including lateral flow and capillary flow. Other methods are relatively small-scale and may not replicate the effects of soil layering below a full-scale infiltration facility. Finally, there are no generally accepted methods for predicting the capacity of drywells that are dominated by horizontal flow rather than vertical flow. This paper presents methods that address these shortcomings and provides estimates of hydraulic conductivity that are within 13% of specified hydraulic conductivity used in numerical simulations.