Extraction and identification of spectrum characteristics of coal and rock hydraulic fracturing and uniaxial compression signals

Microseismic (MS) events generated during coal and rock hydraulic fracturing (HF) include wet events caused by fracturing fluid injection, in addition to dry events caused by stress perturbations. The mixture of these two events makes effective fracturing MS events pickup difficult. This study is based on physical experiments of different coal and rock HF and uniaxial compression. The differences of waveform characteristic parameters of various coal and rock ruptures were analyzed using the Hilbert–Huang transform, leading to some useful conclusions. The phase characteristics of the acoustic emission (AE) energy differed significantly and responded well to the pumping pressure curve. The AE waveforms of HF exhibit similar energy and frequency distribution characteristics after Empirical mode decomposition. The main frequency bands for coal, sandstone, and shale samples are 100–300 kHz, while the mudstone sample is in the range of 50–150 kHz. The decay ratios for coal, sandstone, shale and mudstone samples are 0.78, 0.83, 0.67 and 0.85, respectively. When compared to the uniaxial compression test, the main frequency bands of HF were reduced for coal, sandstone and mudstone samples, whereas shale remained essentially unchanged. The duration, instantaneous energy, and total energy of the HF waveform are smaller than those of uniaxial compression, while the decay ratio is greater, especially for the mudstone samples. The waveform characteristic parameters, trained using the multilayer perceptron neural network, can effectively identify HF and uniaxial compression events with an accuracy of 96%.