Design and modeling of a micro gas turbine with a bottoming Organic Rankine Cycle: steady-state and dynamic approach

This PhD dissertation deals with the integration of a bottoming Organic Rankine Cycle (ORC) cycle to recover heat from the flue gases of a commercial micro gas turbine (mGT) for the production of electricity. The off-design behavior is modelled with a steady-state approach and, in addition, the transient response is analyzed. The application of this system aims to instantaneously satisfy the electricity demand of a set of 150 houses in Cologne (Germany). First, the design of the energy system under study (i.e. mGT coupled with the bottoming ORC) is performed. A commercial mGT (on-design electric power output equal to about 100 kWe) is selected for this application and a thermodynamic optimization of the design of the bottoming ORC is performed aiming to maximize the electric power output. The most promising working fluid for the ORC is an environmentally friendly one, the R1233zd. The total electric output is equal to 119.7 kW at ambient temperature of15°C, split in 93.9 kW from the mGT and 25.8 kW from the ORC. The ORC equipment is sized subsequently after the complete design. Second, the part-load performance of the designed energy system is studied through steady-state models developed in Matlab®. The off-design study aims to analyze how the system behaves when the operating conditions (i.e ambient temperature, power demand) differ from the nominal ones. In addition, the results obtained with the steady-state part-load approach are used to validate the dynamic model results once steady-state conditions are reached. Third, the dynamic response of the energy system is analyzed when the ambient temperature and the electricity demand vary along the two most critical days of the year (i.e. maximum and minimum load). For that purpose, dynamic models are developed in Modelica (in Dymola environment) and validated with the match of the results with the steady-state part-load performance. A control system is designed aiming to guarantee good operational conditions, controlling the electric output, the TOT of the mGT and the TIT of the ORC. The dynamic simulations are performed considering a time step of one second, however, two types of electricity-demand profiles are imposed: real demand conditions and one-minute averaged demand conditions. After analyzing the transient response, it is concluded that the proposed system formed by the mGT and ORC cannot work in stand-alone conditions to satisfy the demand instantaneously, as there is a difference in the demanded electricity and the given electricity. Therefore, a fast, parallel system (e.g. a battery or the electric grid) should be coupled to fulfill instantaneously the demand. The results show that taking an averaged profile helps to get considerably smoother results, therefore, the equipment would suffer less as the temperature gradients are lower.