Development of a non-linear state estimator for advanced control of an ORC test rig for geothermal application. (December 2020)

• Record Type: Journal Article
• Authors: Pili, R.; Eyerer, S.; Dawo, F.; Wieland, C.; Spliethoff, H.
• Is Part Of: Renewable energy. Volume 161(2020)
• Page Start: 676
• Page End: 690
• Publication Date: 2020-12
• Subjects: Organic rankine cycle -- Control -- State observer -- Estimator -- Geothermal; Renewable energy sources -- Periodicals; Power resources -- Periodicals; Énergies renouvelables -- Périodiques; Ressources énergétiques -- Périodiques; 333.794
• DOI: 10.1016/j.renene.2020.07.121
• Languages: English
• ISSNs: 0960-1481

Abstract

Abstract: Organic Rankine Cycle systems are increasingly installed in low-enthalpy geothermal plants to supply efficiently heat and power. The variation in heat demand and the possible contribution to grid ancillary services result in part-load operation and flexibility challenges for the ORC. A suitable control system, able to maximize the power output and guarantee high profits, can be achieved by state-based multivariable advanced controllers. These are based not only on the error on the controlled variables, but also on the information of the system state, which is often non-fully measurable. Therefore, this work focuses on the development of a non-linear discrete state estimator to estimate the state of the compound evaporator/expander of the ORC. First, a dynamic model of the full ORC test rig in Dymola is developed and validated against experimental data, with relative root mean squared error for the major variables below 3.5%. A non-linear state estimator (Unscented Kalman Filter) using a finite volume discretized evaporator coupled with a twin-screw expander is designed and tested on a benchmark case in Simulink. Both the dynamic and the observer model are in good agreement with the experimental data. The observer converges to the evaporator wall temperature with sufficient speed and high accuracy. Highlights: Validation of full dynamic ORC model from experimental data with good agreement. Development of simplified evaporator model for state estimation . Development of Unscented Kalman Filter for evaporator state estimation. Good agreement between measurements, full dynamic model and state observer.