Nonlinear Estimation and Control with Application to Upstream Processes

Subsea development and production of hydrocarbons is challenging due to remote and harsh conditions. Recent technology development with high speed communication to subsea and downhole equipment has created a new opportunity to both monitor and control abnormal or undesirable events with a proactive and preventative approach rather than a reactive approach. Two specific technology developments are high speed, long-distance fiber optic sensing for production and completion systems and wired pipe for drilling communications. Both of these communication systems offer unprecedented high speed and accurate sensing of equipment and processes that are susceptible to uncontrolled well situations, leaks, issues with flow assurance, structural integrity, and platform stability, as well as other critical monitoring and control issues. The scope of this dissertation is to design monitoring and control systems with new theoretical developments and practical applications. For estimators, a novel `1-norm method is proposed that is less sensitive to data with outliers, noise, and drift in recovering the true value of unmeasured parameters. For controllers, a similar `1-norm strategy is used to design optimal control strategies that utilize a comprehensive design with multivariate control and nonlinear dynamic optimization. A framework for solving large scale dynamic optimization problems with differential and algebraic equations is detailed for estimation and control. A first area of application is in fiber optic sensing and automation for subsea equipment. A post-installable fiber optic clamp is used to transmit structural information for a tension leg platform. A proposed controller automatically performs ballast operations that both stabilize the floating structure and minimize fatigue damage to the tendons that hold the structure in place. A second area of application is with managed pressure drilling with moving horizon estimation and nonlinear model predictive control. The purpose of this application is to maximize rate of drilling penetration, maintain pressure in the borehole, respond to unexpected gas influx, detect cuttings loading and pack-off, and better manage abnormal events with the drilling process through automation. The benefit of high speed data accessibility is quantified as well as the potential benefit from a combined control strategy versus separate controllers.