Multi-objective techno-economic optimisation of a Carnot battery application in a parabolic trough concentrating solar power plant
L.G. Redelinghuys, Craig McGregor
Abstract
This research focuses on conducting multi-objective optimisation (MOO) for a Carnot battery applied within a conventional parabolic trough concentrating solar power (CSP) plant. The technical model is formulated and validated in previous work; here, we focus on its economic modelling and techno-economic MOO. Our original contributions aim to provide fundamental (transferable) insight into what constitutes an optimum CSP Carnot battery design and why. We utilise the levelised costs of electricity and storage (LCOE and LCOS) and capacity factor (CF) as objective functions. Our design variables of interest include the thermal energy storage (TES) and heater capacities. Our main findings show that: (1) LCOE and CF dominate the complete MOO problem; (2) TES and heater capacities are positively correlated at Pareto optimality; (3) A narrow TES capacity range (3.4 h to 6.4 h) yields Pareto optimality between LCOE and LCOS; (4) The LCOE can be successfully used to generate a graphical solution method for the entire MOO problem; (5) All Pareto-optimal designs lie on the boundary of the graphical solution method, providing the best energy-cost trade-off; (6) The graphical solution method can be used to directly estimate a significant range of Pareto-optimal designs. • Multi-objective optimisation is conducted for a solar multiple of 2.6. • Economic optimum thermal energy storage (TES) capacities of 3.4–6.4 h are found. • Pareto-optimal resistive heater capacity and TES capacity correlate positively. • All Pareto-optimal solutions lie on the boundary of the graphical solution method. • This method fundamentally identifies a large portion of Pareto-optimal designs.