Optimal Design of Modular Cogeneration Plants for Hospital Facilities and Robustness Evaluation of the Results

The widespread adoption of combined heat and power generation is widely recognized as a strategic goal to achieve significant primary energy savings and lower carbon dioxide emissions. In this context, the purpose of this research is to evaluate the potential of cogeneration based on reciprocating gas engines for some Italian hospital buildings. Comparative analyses have been conducted based on the load profiles of two specific hospital facilities and through the study of the cogeneration system-user interaction. To this end, a specific methodology has been set up by coupling a specifically developed calculation algorithm to a genetic optimization algorithm, and a multi-objective approach has been adopted. The results from the optimization problem highlight a clear trade-off between total primary energy savings (TPES) and simple payback period (SPB). Optimized plant configurations and management strategies show TPES exceeding 18% for the reference hospital facilities and multi–gas engine solutions along with a minimum SPB of approximately three years, thereby justifying the European regulation promoting cogeneration. However, designing a CHP plant for a specific energetic, legislative or market scenario does not guarantee good performance when these scenarios change. For this reason, the proposed methodology has been enhanced in order to focus on some innovative aspects. In particular, this study proposes an uncommon and effective approach to identify the most stable plant solutions through a multi-objective robust design optimization. In particular, the sensitivity of the expected results to possible difficulties in finding commercially available CHP gas engines with sizes reasonably close to the optimal numerical solutions has been estimated. The results indicate that the economic sensitivity is often higher than the energetic sensitivity for most of the optimal solutions, with standard deviation accounting up to 7% of its mean value for the SPB, whereas that percentage is always under 3% for the TPES. Furthermore, the research highlights how the expected results obtained through a deterministic definition of the input decision variables could be overestimated compared to the robust design approach. The proposed research also highlights how optimized CHP plants can be characterized by reasonable levels of energetic and economic sensitivity to changes in the following variable quantities: selling price of electricity, reference efficiency of the Italian thermoelectric generation and selling price of the energy efficiency certificates recognized by the Italian legislation. Indeed, Pareto optimal solutions indicate that the standard deviation for the SPB is always less than 3.5% of its mean value, while this percentage is always under 7% for the TPES