An investigation into the cold heavy oil production with sand process using synthetic cores and designed experiments

Abstract

This study deals with the development of a methodology for making low compressive strength cores used in an experimental investigation of the Cold Heavy Oil Production with Sand (CHOPS) process. An experimental setup was designed and built to investigate the effect of rock compressive strength as well as flow parameters, such as oil viscosity, net and total confining pressure, and injection rate, on core permeability. The approach was to optimize the value of a response variable by changing the values of the affecting factors. Sand blends were prepared by varying the ratios of aggregate, cementing material and water to prepare synthetic cores. An experimental unit was built to simulate wormhole propagation during the CHOPS process, where oil, at an ambient temperature, was injected into 2-inch × 4-inch cores at varying rates of 0.5–10 ml/min under differential confining pressures of 500 and 1000 psia. The pressure drop across the core was monitored and recorded throughout the process. When non-swelling clay is used as a cementing material compared with actual cement to make synthetic core, the compressive strength of the samples falls dramatically by 64%. Two factors were considered in the coreflood experiments: Oil Viscosity (370 and 690 cp) and Injection Rates (0.5 and 3 ml/min) at a net confining pressure of 500 psia, below the compressive strength of the core. It is hypothesized that injecting oil of different viscosities at different rates affects the internal structure of the core in different ways (there is fluid-rock interaction) and thus, at lower pore volumes of injection, the permeability of the core for high viscosity oil is almost 11.2% greater than for low viscosity oil. Also, design of experiment approach was used and regression model was developed for permeability of core based on values recorded at specific pore volumes injected for different injection rates and oil viscosities. It was found that at a constant confining pressure for all rates and at lower pore volumes injected, 99.8% of the variance in permeability can be explained by oil viscosity, injection rate and their interaction. At higher pore volumes injected, the variance in permeability that can be explained by oil viscosity, the injection rate and their interaction is only 40.63%.