A statistical-coupled model for organic-rich shale gas transport

2018 
Abstract Computation of shale permeability is challenging owing to the complicated physics of gas flow and the small-scale heterogeneity in nanometer pores. In this study, a statistical-coupled model (SCM), based on the combination of FIB/SEM and SEM imaging measurements and statistical analysis is proposed to bridge the nanopore-scale with the organic representative elementary volume (oREV)-scale. FIB/SEM imaging is adopted to get the nanometer pores properties in organic matter (OM), and SEM imaging is used to get the OM content distribution. With the analysis of nanopores of a Longmaxi shale sample from the Chongqing Province, China, it is demonstrated that the pore size distribution obtained from FIB/SEM images of typical samples is representative considering the principal parts of the pore radii are similar comparing with 2D SEM image. Then, the SCM is constructed based on the combination of the statistical method, the series-parallel model and the equivalent model for microstructures, and the rationality of SCM are also investigated. The obtained characteristic parameters show an excellent performance in calculating the SCM element permeability with a small deviation of less than 3% and a significantly faster computation speed comparing with the previous literature by approximately 400 times. Using the SCM, a method for the construction of oREV and the determination of oREV-scale permeability for the organic-rich shale is presented. Finally, the sensitivity analyses of oREV-scale permeability are conducted and the results show that the permeability is sensitive to the OM content distribution, the district of shale sample and the permeability of IOM. The influence of OM permeability on macroscale is also analyzed. The new model can advance the understanding of the multiscale phenomena and establish a relationship between microscale properties and macroscale behavior.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    39
    References
    11
    Citations
    NaN
    KQI
    []