Source Scaling of Single-Fired and Delay-Fired Explosions Constrained by In-Mine and Regional Seismograms

This work quantifies seismic coupling as a function of charge weight for single-fired (simultaneously detonated) explosions observed in the mine and at regional distances. These single-fired explosions are contrasted to standard production shots from the same mine. Production shots include cast blasts to remove overburden and coal shots to fracture the resource. The blasts were conducted in a mine in NE Wyoming. The data sets consist of measurements in the mine (100-10000 m) and at regional distance (360 km) using a seismic array near Pinedale, Wyoming (PDAR). The single-fired explosions ranged in size from 5500 to 50000 lbs. Peak amplitude measured at a single element of the regional array are modeled by a power law dependent on explosive weight. Similar scaling constants were determined for each of the dominant regional phases: Pn – 0.84±0.14; Pg – 0.84±0.09; and Lg – 0.91±.08. Spectral ratios of observations from different explosions observed at the same station document the frequency dependence of the source scaling relations. Spectral ratios using in-mine observations are consistent with those using regional data. The empirical source ratio is matched with a Mueller-Murphy source model replicating both the increase in long period spectral level and the accompanying decrease in source corner frequency with increasing charge weight. The coal and cast blasts that make up the production shots ranged in size from 77,000 to 4,738,230 lbs. Peak amplitudes from the largest production shots are comparable to those from the largest single-fired explosion. This dramatic decrease in amplitude for production shots (factor of 60-300) can be explained in terms of destructive interference between waveforms from individual boreholes introduced by delay firing. Shot specific models of delay firing suggest that peak amplitudes for the production shots are insensitive to the total amount of explosives in the blast. Peak amplitude data supports this modeling result.