Effect Of Immiscible Liquid Contaminants On P-Wave Transmission Through Natural Aquifer Samples

Abstract

We performed core-scale laboratory experiments to examine the effect of non-aqueous phase liquid (NAPL) contaminants on P-wave velocity and attenuation in heterogeneous media. This work is part of a larger project to develop crosswell seismic methods for minimally invasive NAPL detection. The test site is the former DOE Pinellas Plant in Florida, which has known NAPL contamination in the surficial aquifer. Field measurements revealed a zone of anomalously high seismic attenuation, which may be due to lithology and/or contaminants (NAPL or gas phase). Intact core was obtained from the field site, and P-wave transmission was measured by the pulse-transmission technique with a 500 kHz transducer. Two types of samples were tested: a clean fine sand from the upper portion of the surficial aquifer, and clayey-silty sand with shell fragments and phosphate nodules from the lower portion. Either NAPL trichloroethene or toluene was injected into the initially water-saturated sample. Maximum NAPL saturations ranged from 30 to 50% of the pore space. P-wave velocity varied by approximately 4% among the water-saturated samples, while velocities decreased by 5 to 9% in samples at maximum NAPL saturation compared to water-saturated conditions. The clay and silt fraction as well as the larger scatterers in the clayey-silty sands apparently caused greater P-wave attenuation compared to the clean sand. The presence of NAPLs caused a 34 to 54% decrease in amplitudes of the first arrival. The central frequency of the transmitted energy ranged from 85 to 200 kHz, and was sensitive to both grain texture and presence of NAPL. The results are consistent with previous trends observed in homogeneous sand packs. More data will be acquired to interpret P-wave tomograms from crosswell field measurements, determine the cause of high attenuation observed in the field data and evaluate the sensitivity of seismic methods for NAPL detection.