Development of an expert system to remotely build and control drilling fluids

Abstract Drilling processes are viable only if drilling fluid is used. Drilling fluids are responsible for a number of operations, including cooling the system, carrying drill cuttings to surface, transmitting pressure and signals throughout the well and others. Therefore inefficiency during control of the drilling fluid parameters may cause costly and even deadly problems. In a previous work, researchers developed a sensor skid that was capable of reading, in real-time, physicochemical data of drilling fluids. This work expands the concepts and results obtained from that study, allowing automatic control of apparent viscosity and density. Using software, researchers have developed a pilot production plant. The production plant was coupled to the previously fluid monitoring skid, forming a closed hydraulic-electric loop. Some uncertainties had to be overcome, so researchers developed a controller based on Boolean logic and heuristic rules; this controller accounts for the kinetic behavior of the viscosifier, the constant changes of solids manufacturers, and the need to control multiple devices concomitantly. The expert system incorporates human knowledge instead of spending time and resources on calibrating the required classic control parameters. Researchers performed tests of set point tracking and disturbance rejection. The results demonstrated that the proposed control technique was adequate and could be applied as a control strategy to automate the quality control of drilling fluid.