An intelligent hydraulic fracture optimization model: A new tool to stimulate low permeable oil reservoir

The increasing demand for hydraulic fracturing to stimulate low-permeable reservoirs has lead to the use of various fracture optimization techniques to find optimum values of controllable design parameters. In case of inappropriate values of these parameters, many unexpected complexities may arise during treatments that lead to high treatment cost and low productivity. While recognizing the role of incompatible fracturing fluids and inappropriate execution of treatments, the role of inappropriate treatment optimization is not well investigated. To avoid complexities, one needs to solve a constrained model to screen out potentially problematic designs with the aid of an intelligent optimization algorithm, in contrast to the widely-followed conventional approach of unconstrained parametric optimization. This paper presents an integrated constrained model for treatment optimization to maximize production. The Model has incorporated many design constraints based on hydraulic fracture mechanics and industry experience, and the global optimization scheme is driven by an intelligent moving object algorithm. The design constraints relate fracture geometry and its growth, surface operational parameters, non-dimensional fracture conductivity, proppant number and proppant fall rate in the fracture. Successful viscosity scheduling of the fracturing fluid is particularly influenced by power-law parameters and shear rate in the fracture, which can be controlled by fluid preparation and controlled injection. These parameters are included in the free design variables in optimization work for the first time. A combined objective function is formulated to maximize production while simultaneously minimizing the treatment cost, and trade-offs between these two conflicting design objectives have been investigated. While applied this model to a low permeable oil reservoir to demonstrate its merits, an optimum treatment designed by this model has offered an incremental production by 100% and a treatment cost saving by 20% of their initial cost. Sensitivity results of various parameters also indicate the implication of design other than optimum.