(747f) Simultaneous Working Medium Selection, Process and Control System Design for Organic Rankine Cycles

Organic Rankine cycles (ORC) are efficient processes for the recovery of relatively low enthalpy waste heat from industrial streams and exhaust gas streams in combustion engines and its conversion to electricity. The decisions involved in the design of an ORC range from the selection of a suitable working medium for the available temperature levels of the hot and cold sources to the unit operation sizing and extending to the associated controller design. It is well known that mixtures of components exhibit a superior performance as they allow the reduction of the exergy losses in the evaporator and the condenser in the ORC. The closer tracking of the hot source temperature curves is the primary reason for such a behavior. Papadopoulos et al (2013) offered a comprehensive investigation of a large number of potential mixtures as working media in an ORC that also considered variability in the operating conditions. Later, Zarogiannis et al (2017) studied the influence the working medium has on the achieved control performance. However, no study has ever performed a simultaneous working medium selection with process and control system design.

The present work encompasses the design of the process equipment in an ORC with the control system design for a set of potentially promising working media under a unified framework. Working media selection involves the identification of the components forming the mixture and the associated concentration of each component. The set of mixtures that possess the desired properties and have the potential for a highly performing ORC system have been identified using a computer aided molecular design (CAMD) procedure (Papapdopoulos et al., 2013). The CAMD procedure targets specific mixture properties but it does not consider the dynamic behavior of the system under disturbance influence or variability in the operating conditions. Since CAMD is based on multi-objective optimization it identifies all mixtures that are included in the Pareto optimal front.

The integrated working medium selection with the process and control system design aims to investigate the optimal combination of the working mixture with the ORC configuration that satisfy an economic criterion under steady-state and closed loop conditions. Therefore, for a selected working mixture the steady-state design of the ORC is performed based on economic optimization, where the optimal configuration and sizing of the equipment is obtained. In addition, a set of disturbance scenarios is constructed that are representative of the most commonly occurring operating conditions for the process. A model predictive control (MPC) scheme is employed for the satisfaction of the process goals under process variability and disturbance influence. The use of a MPC enables the best handling of the interactions among the various processes in the cyclic process. The integrated design framework enables the determination of the most suitable input-output structure for the controller by allowing the selection of the controlled and manipulated variables from a pool of screened options. The performance of the MPC under the specified disturbance scenarios is calculated through dynamic simulations for the selected working medium and process flowsheet configuration, in each optimization iteration. The dynamic performance is converted to an economic equivalent term so that it can be combined with the steady-state design economic term to generate the overall objective function value.

Novel and conventional hydrocarbon- and halogenated hydrocarbon-based mixtures are considered for the integrated ORC process and control system design under the proposed unified framework. Variation in the hot and cold source streams is considered in the system where the employed MPC aims to maintain the evaporator temperature at the desired level and achieve the maximum work in the expander. The simultaneous problem of working mixture selection and process design under static and closed loop conditions is solved using a stochastic optimization technique. The results provide significant insights in the behavior of the ORC system under closed loop conditions. The ability of the ORC system with the MPC to maintain a high performance under variability is found to be influenced by the sensitivity of the working medium physical properties within the range of variation that is transferred through the system.

Cited References

Papadopoulos A.I., M.Z. Stijepovic, P. Linke, P. Seferlis, and S.S. Voutetakis, “Towards Optimum Working Fluid Mixtures for Organic Rankine Cycles using Molecular Design and Sensitivity Analysis”, Industrial and Engineering Chemistry Research, 52(34),12116–12133, 2013.

Zarogiannis T., A. I. Papadopoulos, P. Seferlis, and P. Linke, “The Impact of Novel and Conventional Working Fluids on the Control Performance in Organic Rankine Cycles”, Computer Aided Chemical Engineering, 40, 2443-2448, 2017.