The impact of kerogen properties on shale gas production: A reservoir simulation sensitivity analysis

2017 
Abstract Gas shales are complex materials with transport occurring over many length scales, from relatively large systems such as natural and induced fractures to relatively small systems including kerogen-hosted pores of nearly molecular size. Here we present results of a sensitivity analysis performed using reservoir simulation designed to test how these different pore space length scales impact gas production. The shale is modeled as a triple porosity system comprising: first, natural and induced fractures; second, kerogen-hosted pores of approximately 25 nm diameter (as are typically observed in SEM images); and third, kerogen-hosted pores of approximately 1 nm diameter (which are below the size detectable by SEM but are typically observed in X-ray diffraction and gas adsorption experiments). In these simulations, the smaller kerogen-hosted pores act as a source term, releasing primarily adsorbed gas. The larger kerogen-hosted pores contribute to storage and transport, serving both as a source of gas on their own and a system through which the gas originating in the smaller pores can flow. As a result, the volume of the smaller pores impacts mainly the later production, when the reservoir pressure is near or below the Langmuir pressure. The volume of the larger pores impacts mainly the early stage of production, when mainly free gas is being produced. The radius of the smaller pores impacts the late stage of production, as smaller pores have greater surface area and therefore larger Langmuir volume. The radius of the larger pores has little impact on the ultimate recovery but instead impacts the production rate, as larger pores correlate with greater permeability. The results suggest that measurements of the variation of kerogen properties—performed potentially using cuttings, cores, or logs—can be used to refine parameters in reservoir simulations.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    43
    References
    13
    Citations
    NaN
    KQI
    []