The invention provides a
methane occurrence state quantitative
analysis method for accurate resource evaluation of deep
coal bed gas, which comprises the following steps: firstly, collecting a two-dimensional
nuclear magnetic resonance spectrogram of a
coal sample adsorption test, carrying out gridding
processing on two-dimensional
nuclear magnetic resonance spectrogram data of the
coal sample adsorption test, and establishing a two-dimensional
signal matrix; then, according to the distribution characteristics of the
longitudinal relaxation time T1 and the
transverse relaxation time T2, dividing the
spectrogram into four regions, namely a single-layer adsorption state, a filling adsorption state, an inter-pore
free state and an inter-cavity
free state; and finally, carrying out point-by-point summation on the signals of each region, and converting a value obtained after
signal intensity summation into a gas
content model so as to obtain the quantitative content of
methane in different occurrence states. According to the method, clear distinguishing between single-layer adsorption and filling adsorption can be achieved, the law that the occurrence state of adsorption and free
methane evolves along with pressure can be revealed, and the technical problem of accurately determining the in-situ occurrence state of deep coal
bed gas, which puzzles the deep coal
bed gas industry for a long time, is solved.