We study infrasonic signals from large surface mining explosions in Wyoming. Detections at the Pinedale Infrasound Array (PDIAR) (obtained using a conventional array-processing technique) are automatically associated with ground-truth mining explosions at a range of 368 km. We then focus on four clear signals from mining explosions. By performing a detailed noise study and modeling the propagation of infrasound using a raytracing algorithm and ground-to-space (G2S) atmospheric models, we assess the factors that contribute to the detectability of mining explosions. We find that we can explain most of the observations by propagation and noise effects alone, but that there are at least two notable outliers. Because of high noise levels at the PDIAR array, which places significant constraints on the sizes of mining explosions that can be detected, these results are strongly biased and must be reassessed for lower-noise infrasound arrays.