Master's thesis and Internship : Efficient and cost-effective electricity generation in tower CSP plants

Abstract

Conventional concentrating solar power with thermal storage can deliver dispatchable renewable electricity, but it remains more expensive than photovoltaics, which motivates John Cockerill to develop a modular tower concept based on a large number of small solar units feeding a shared power block. The central design question is the choice of power block, as it sets the plant efficiency and therefore the number of solar units required per megawatt of electrical output, making it a key driver of the levelised cost of energy. This work determines whether any alternative power block can achieve a lower levelised cost of energy than the reference subcritical steam Rankine cycle fed by Solar Salt at 565 °C.

Steady-state thermodynamic models were built with the open-source LaboThapPy library for the steam Rankine cycle, the supercritical carbon dioxide recompression cycle, the recuperated air Brayton cycle with intercooling and reheating, and the Air+Steam and Air+organic Rankine cycle combined cycles. These models were used to perform parametric studies from 565 to 1000 °C. A dedicated model of the molten salt piping network was developed to quantify its length, cost, and thermal losses as a function of plant layout. These results feed a techno-economic levelised cost of energy model that optimises each configuration over the solar multiple and the storage capacity at the reference site of Zamora, Spain. For the air-based configurations, no cost model existed prior to this work, and a complete cost structure covering the air receiver, the packed bed storage, the high-temperature heat exchanger, and the air power block was developed from scratch.

Under current market conditions, no alternative outperforms the reference. The 100 MWel steam Rankine cycle reaches the lowest levelised cost of energy at 60.68 €/MWh. The Air+Steam combined cycle and the Air-Salt engine follow at 66.78 and 68.10 €/MWh respectively. The Air-Salt engine reaches this result by eliminating the piping network and the heat exchanger between the heat transfer fluid and the working fluid, despite a low technology readiness level. The remaining air-based cycles range from 81.37 €/MWh for the standalone air Brayton cycle to 81.88 €/MWh for Air+organic Rankine cycle, while the supercritical carbon dioxide recompression cycle stays far behind at 148.59 to 156.31 €/MWh. The analysis shows that thermodynamic efficiency alone does not determine the levelised cost of energy, as the dominant factor is the specific cost of the power block at the small unit sizes imposed by the modular concept. A sensitivity study on power block cost and net efficiency identifies the air Brayton cycle and the Air-Salt engine as the most credible paths toward competitiveness with the reference.