INSPIRED BY TGUESS: A PROCESS-BASED MODEL FOR CALCULATING HYDRAULIC CONDUCTIVITY FROM DRILLER-PERFORMED WELL COMPLETION TESTS

Over his long and storied career, Ken Bradbury’s research interests spanned a variety of topics in hydrogeology, including the estimation of hydraulic conductivity (K). These data are notoriously difficult to obtain at a high spatial resolution, despite their importance in hydrogeologic modeling. Reliance on limited datasets and generalized estimates can lead to inaccurate models and, consequently, inaccurate results. To improve the accuracy and resolution of K data in Ohio, the Ohio Department of Natural Resources (ODNR), Division of Geological Survey created a hydraulic conductivity geodatabase of data collected from pumping tests, slug tests, groundwater flow models, and well completion tests performed by drillers. This geodatabase will aid in an ongoing project to map statewide aquifer hydraulic conductivities and yields.

Hydraulic conductivity data collected from pumping tests and slug tests are of high quality, but do not have high spatial resolution. To improve the spatial resolution of the dataset, Survey staff created a Python model inspired by TGUESS, a computer program written by Ken Bradbury and Edward Rothschild, to derive K estimates from ODNR water well logs. The model utilizes an iterative formulation of the Theis Equation to estimate transmissivity from well completion test data (test rate, test duration, and drawdown) and a process-based framework for converting drilling log formations into hydrogeologic parameters.

The Survey model makes many of the assumptions of and corrections to the Theis Equation central to TGUESS, such as partial aquifer penetration. However, unlike TGUESS, the Survey model also corrects for drawdown in unconfined wells and utilizes well construction and drilling log information to determine aquifer confinement and estimate minimum and maximum saturated aquifer thickness, resulting in an estimated minimum and maximum K for each well based on the calculated transmissivity.

The Survey model establishes a generalized framework for obtaining hydraulic conductivity estimates from commonly available well logs. The model, final aquifer yield and hydraulic conductivity maps, and digital hydraulic conductivity dataset will be made publicly available once the project is completed in 2024. These data will provide users with unprecedented high-resolution aquifer hydraulic conductivity information for Ohio.