Mechanistic controls on permeability evolution in thermally-upgraded low-maturity oil shales: Application of machine learning outputs

In-situ thermal upgrading aids recovery from low-maturity oil shales where low permeability is the rate-limiting feature. We use machine leaning classified pore and pore network morphological descriptions recovered at elevated temperatures to define the dynamic thermal evolution of permeability. These descriptions define key factors influencing permeability evolution, in particular the development of anisotropy and its implication for recovery. Heating enhances permeability by increasing both the number and total cross-sectional area (SEM) of pores. Fractal dimensions indicate that the pore microstructure is anisotropic in the bedding-parallel and bedding-perpendicular directions and is upgraded by elevated temperature. The permeability anisotropy endures throughout the entire heating process and fluctuates at elevated temperature, quantified by the index called anisotropic coefficient of permeability. A newly proposed “aggregation degree” indexes the relative contribution of minority pores to overall permeability – 50 % of the permeability (K) is sourced from <8 % pores in the SEM section. Increased temperature elicits increased permeability – thus, the temperature applied at the injection well defines the flow limiting permeability threshold at the production wells and thus controls flow rates from the entire heated reservoir. This work provides fresh insights in defining the thermal permeability response of low-maturity oil shales and guides fluids recovery.