[home] [Personal Program] [Help]

---

Session: Parallel Session: Simulation and Design Tools
Session starts: Thursday 22 September, 11:10
Presentation starts: 11:30
Room: Auditorium

---

Vincent Lemort (University of Liège)
Sylvain Quoilin (University of Liège)
Sébastien Declaye (University of Liège)
Assaad Zoughaib (Ecole des Mines de Paris)

Abstract:
The behaviour of the ORC has been extensively studied theoretically and experimentally in the past few years. Most of the proposed ORC models are steady-state models, accounting for stabilized working conditions. However, some ORC applications are used with inherently variable heat source inputs. The case of heat recovery on internal combustion engines is a good example of varying heat source flow rate and temperature in very short periods of time. ORC cycles coupled to solar applications can also be highly dynamic in the absence of storage since direct radiation can fluctuate quickly depending on the climatic conditions. Steady-state models are not able to predict the transient behaviour of the cycle with such heat sources, nor can they simulate a proper cycle control strategy during part-load operation or start & stop procedures. This paper proposes a dynamic model of an ORC by focusing specifically on the dynamic behaviour of the heat exchangers, the dynamics of the other components being of minor importance. The model is developed under the Modelica environment using the TILMedia library for the computation of the working fluid thermodynamic properties. This model is specifically designed to ensure the robustness and the speed of the simulation algorithms: - Initialization strategies are developed by simplifying the problem at time zero in order to avoid too complex non-linear initial systems of equations. - Simplified heat transfer laws are proposed based on heat transfer correlations available in the scientific literature. - Numerical issues such as division by zero are avoided by linearising or limiting to a finite value some thermophysical equations or properties. The studied system is a small-scale waste heat recovery system using a scroll machine as expansion device. The heat exchangers are plate heat exchangers and the pump is a diaphragm pump. In order to control the system under transient conditions, the set point of the evaporating temperature is optimized using a steady state model and implemented into the control unit, where it is continuously re-adjusted to the optimal value. The two control variables are the superheating at the expander inlet and the evaporating temperature. Two types of controllers are implemented and compared in terms of performance (seasonal performance of the system), safety (achievement of set points), and robustness: a feedback control strategy based on two PID controllers and a model predictive control strategy (MPC). The potential of the later controller over classical PID is discussed.