USGS

M.A. Scholl, 1997.
Use of a Numerical Model to Estimate the Effect of Permeability Heterogeneity on Biodegradation Rate Measurements in an Alluvial Aquifer: EOS, Transactions, American Geophysical Union 1997 Spring Meeting, v. 78, no. 17, p. S140.


A closed landfill near the city of Norman, Oklahoma, is sited on the floodplain of the Canadian River. Leachate is moving toward the river, through layered sediments ranging from sand to clay. A 2-D numerical model that simulates biodegradation processes in ground water (BIOMOC, Essaid et al., 1995) was used to design a sampling strategy to measure biodegradation rates of compounds in the leachate plume. Hydraulic conductivity (K) measurements were made at one-meter depth intervals through the aquifer along a transect parallel to flow direction. Sequential Gaussian simulation was used to generate K distributions, conditioned on 35 data points from the aquifer, over the range log K = -5 to -3.5 m/s (clean sand to silty sand). Layers of similar K were 1-3 m thick and up to 50 m long. A zero-order biodegradation rate was set for the contaminant, with no bacterial growth and uniform bacterial population. A simulation with homogeneous K at the mean value was used as the control case. Simulations were then run with different K realizations, and the modeled steady-state contaminant plume was sampled at various depths and distance intervals. Apparent biodegradation rates were calculated using the average flow velocity, the parameter that would be available in a field situation. For this example (sand-silty sand, disappearance of compound within 100 m), results of initial simulations show that apparent rates differ from the control case by up to 54% if the contaminant plume is sampled every 20 m at midpoint depth in the aquifer. Apparent biodegradation rates are slower if flow velocity is higher than average, and faster if velocity is lower. Estimates can be improved to within 10% of the control by averaging samples at 3 depths, but there is little further improvement with samples at 6 depths. Decreasing well spacing to 10 m apart does not substantially improve the rate estimates. Further simulations will examine the effects of different K ranges and bacterial growth on biodegradation rate measurements.


[ USGS, Oklahoma District Home Page | Index of Abstracts | Norman Landfill Home Page ]
http://ok.water.usgs.gov/norlan/abstract/ab3.html
For comments and questions about this web page, contact:
email iconJason Masoner