Hydraulic Tomography: Estimating 3D Hydraulic Conductivity, Fracture Network, and Connectivity

Tiedeman, Claire

Hydraulic Tomography: Estimating 3D Hydraulic Conductivity, Fracture Network, and Connectivity

Monday, September 23, 2019: 9:20 a.m.

Claire Tiedeman , Water Mission Area, U.S. Geological Survey

Warren Barrash , Boise State University, Boise, ID

Hydraulic tomography was performed in fractured mudstones at the former Naval Air Warfare Center, West Trenton, New Jersey, by conducting 47 pumping tests in packer-isolated intervals of seven closely-spaced wells. During each test, drawdown data were collected from 30-40 intervals throughout the wellfield. The 3D hydraulic conductivity (K) distribution was estimated at high resolution (K blocks < 1 m³) using the drawdown data, a heterogeneous continuum forward model, and geostatistical inversion. The 3D K estimate ranges from ~0.1 m/s (highest-K fractures) to ~10^-13 m/s (unfractured mudstone). Important estimated features include: (a) a highly fractured zone (HFZ) consisting of a sequence of high-K bedding-plane fractures; (b) a low-K zone that disrupts the HFZ; (c) several secondary fractures of limited extent; and (d) large regions of very low-K rock matrix, particularly at depth. The 3D K estimate explains highly complex drawdown behavior observed in the field. Model fit to the drawdown data is best in the shallower part of the wellfield, with high density of observations and tests. The simulated drawdown progression during the tests, and particle tracking simulations in several directions across the wellfield, reveal a 3D fracture network within the estimated K distribution, and connectivity routes through the network.

Claire Tiedeman, Water Mission Area, U.S. Geological Survey

Claire Tiedeman is a Research Hydrologist at the U.S. Geological Survey, where her work involves characterizing and modeling flow and transport in fractured rock aquifers, calibrating and evaluating models of complex groundwater flow systems, and developing methods to evaluate prediction uncertainty.

Warren Barrash, Boise State University, Boise, ID

Warren Barrash received an AB (1970) in English from Dartmouth College, then worked in public health in Malaysia and Ethiopia as a Peace Corps volunteer between 1970 and 1974. He received MS (1978) and PhD (1986) degrees in Geology (emphasis in Hydrogeology) from the University of Idaho. Since 1993 he has been a Research Professor with the Center for Geophysical Investigation of the Shallow Subsurface (CGISS) and the Department of Geosciences at Boise State University. Current research interests include measuring and modeling heterogeneity in aquifers using hydrogeologic, geophysical, and geostatistical methods, and modeling subsurface and stream-aquifer flow and transport.