Optimisation model of fuzzy ball fracture diverting fluid based on multiple quadratic regression method

Abstract

Fuzzy ball working fluid is a novel self-matching, high strength and low damage plugging material, which is ideal as the fracture diverting agent in petroleum engineering. The optimisation model of fuzzy ball fracture diverting fluid in terms of viscosity, density and stable time was developed in the present paper based on the multiple quadratic regression method. Twenty groups of experimental data were used to establish the forward model, relating viscosity, density and the stable time of fuzzy ball fracture diverting fluid with the composition of the fluid. The other four groups of experimental data were use to validate the model. The maximum model error was less than 5%. The inverse model was developed afterwards based on the forward model to optimise the composition of the fuzzy ball fracture diverting fluid at given viscosity, density and stable time. The validated model was then applied to find the eligible compositions of fracture diverting fluid having the viscosity of 120 cP, density of 0·87 g cm−3 and stable time of 12 h. The fluid was compounded according to the predicted compositions and the maximum relative error in terms of viscosity, density and stable time was found to be 7·5%. The model can provide an effective tool to optimise the fuzzy ball fracture diverting fluid and enhance the efficiency of preparing working fluid.