The comprehension of the mechanisms governing spontaneous water imbibition in gas-water systems plays a significant role in the operation of hydraulic fracturing and the development of coalbed methane (CBM). In this study, nuclear magnetic resonance (NMR) techniques were used to investigate the pore structure and fluid behavior of different rank coal samples during spontaneous imbibition. Analyses were conducted on the 1D NMR T2 spectrum, 2D NMR T1-T2 spectrum, and layer division T2 spectra to achieve accurate and detailed information about the internal pore structure of coal and the characteristics of fluid transport during spontaneous imbibition. Low-rank coal exhibits good pore connectivity and favorable pore sorting characteristics, indicating favorable reservoir conditions. High-rank coal has larger pore spaces and highly developed micropores, which are highly beneficial for gas adsorption. However, its poor pore sorting characteristics and connectivity limit the migration and diffusion of fluids within the reservoir. The imbibition capacity follows a specific order of contribution, with small pores (10-50 nm) having the most significant role, followed by micropores (2-10 nm), ultramicropores (r < 2 nm), mesopores (50-1000 nm), macropores (1000-10,000 nm), and microfractures (r ≥ 10,000 nm). Low-rank coal stands out due to the restricted development of ultramicropores and small pores, leading to a different contribution of imbibition capacity compared to other samples, where macropores and microfractures dominate over all pore types. The coal reservoirs with favorable pore sorting characteristics and pore connectivity tend to exhibit a tendency toward rapid saturation and attainment of a prompt stable state during the hydraulic fracturing process. Finally, the mechanism of the late retreat effect of imbibition and the laws governing different coal ranks, pore structures, and fluid transport were discussed. This study offers comprehensive analyses of the mechanism of coal spontaneous imbibition and the characteristic laws of fluid seepage, providing insights into the optimization of CBM recovery and reservoir management.