MULTIVARIABLE FUZZY CONTROL FOR MONITORING AND CORRECTING PHYSICOCHEMICAL PROPERTIES OF WATER-BASED DRILLING FLUIDS

During the drilling stage of oil production, the use of drilling fluids is essential to the success of the entire operation. These fluids exert countless properties on drilling operations, but those linked to rheological properties and maintaining the hydrostatic pressure gradient inside the oil well stand out. Since drilling fluids are subject to fluid-rock interactions, contamination by salts from the geological formation can become a problem during the exploration stage of the well, and this additional salt content can impact the fluid's rheological characteristics and density. The increased content of salts in water-based drilling fluids is characterized by an increase in the fluid's electrical conductivity. However, monitoring and controlling this property are yet to be explored. To face such challenges, this study proposes the development of a multivariable fuzzy controller to oversee the properties of electrical conductivity, apparent viscosity, and density in water-based drilling fluids. The technique was evaluated using an experimental unit for preparing and monitoring drilling fluids, where servo and regulatory tests were carried out to assess the controller's capacity. Servo tests are conducted by changing the setpoint of each variable individually, while the controller should maintain the other properties at their setpoints. Regulatory tests evaluated the controller's performance in rejecting a water kick-like disturbance imposed by the sudden addition of 10 liters of water into the system and another disturbance caused by adding 0,5 kg of salt to the fluid tank. The multivariable fuzzy control strategy developed in this work proved efficient and robust in keeping the physicochemical properties of water-based drilling fluids at their setpoints, emerging as a potential preventive and corrective tool for monitoring this process.