Nonlinear Model Predictive Control and Dynamic Modelling of a Small-Scale Organic Rankine Cycle for Low-Grade Waste Heat Recovery

This paper presents a comprehensive study on the dynamic modeling and control of a small-scale organic Rankine cycle (ORC) system for waste heat recovery in power generation applications. The primary objective is to develop an effective control strategy that optimizes the system performance under transient operating conditions, maximizing the expander power output. A high-fidelity dynamic model of the evaporator is developed using the finite volume method and validated against experimental data. A reduced-order control-oriented model is derived using the moving boundary method to facilitate the design of the controller. A nonlinear model predictive control (NMPC) strategy is proposed, utilizing the control-oriented model to predict future system behavior and optimize control actions while explicitly considering system constraints. An extended Kalman filter (EKF) is integrated with the NMPC controller to estimate unmeasured state variables. Experimental validation on an engine test bench demonstrates the controller’s ability to effectively track the reference superheating degree with an average error of ±2°C, compensate for disturbances within 5 seconds, and maintain safe expander operation under highly transient conditions. The NMPC approach achieved a 15% improvement in cycle efficiency compared to steady-state operation. Comparative analysis with traditional control methods highlights the superiority of the NMPC approach, reducing tracking error by 40% and response time by 50% compared to PID control, while effectively handling system constraints. The proposed dynamic modeling and control strategies provide a solid foundation for the effective utilization of small-scale ORC systems in waste heat recovery applications, contributing to the advancement of efficient and sustainable energy systems.