EXAMPLES OF THE EFFECT OF SOURCE AND RECORDING SITE PARAMETERS ON THE SEISMIC RESPONSE OBSERVED FROM PLOWSHARE PROJECTS, GASBUGGY AND RULISON

Abstract

The amplitude and frequency composition of the seismic motions observed over a wide geographic area from Gasbuggy and Rulison, two Plowshare detonations, illustrates the effect which source and recording site parameters have on the resultant ground motion. Gasbuggy (29 kt) and Rulison (40 kt) were detonated, respectively, in the San Juan and Piceance Creek sedimentary basins of New Mexico and Colorado. Because both detonations were emplaced in sedimentary formations to stimulate the flow of natural gas, these detonations were placed at a greater depth of burial than typically required for containment. The effect of source parameters (device depth of burial and energy release) was simulated by seismic scaling theory. This theory (based on an extension of Sharpe’s problem) predicted a smaller elastic radius and, consequently, higher dominant frequency of generation for the elastic waves than would be expected on the basis of Nevada Test Site experience with typically contained events of 29 and 40 kt. Observed effects of the source variables were displayed in the Gasbuggy and Rulison seismic data: (1) a shift of the frequency of maximum spectral response to the high frequency end of the spectrum and (2) enhancement of the peak vector surface particle accelerations and velocities and a decrease in the peak vector surface particle displacements. Some of the Rulison recording sites were located on thin (50 ft thick or less) layers of alluvium which, on the basis of refraction surveys, exhibited a fairly significant contrast in acoustic impedance relative to that of the underlying sedimentary rocks. The effect of the lowvelocity alluvium layers at recording sites of interest was simulated by amplitude amplification modeling (based on the Haskell‐Thompson matrix formulation) using interpreted data from refraction surveys to define the layer physical parameters. These calculations indicated that significant local frequency‐dependent amplification would occur at some locations in the Piceance Creek Basin, a prediction which was verified by the seismic data.