Techno-Economic Comparison of Solar-Driven SCO2 Brayton Cycles Using Component Cost Models Baselined With Vendor Data and Estimates

Abstract

Supercritical carbon dioxide (sCO2) Brayton power cycles have the potential to significantly improve the economic viability of concentrating solar power (CSP) plants by increasing the thermal to electric conversion efficiency from around 35% using high-temperature steam Rankine systems to above 45% depending on the cycle configuration. These systems are the most likely path toward achieving the Department of Energy’s (DOE) Office of Energy Efficiency and Renewable Energy (EERE) SunShot targets for CSP tower thermal to electric conversion efficiency above 50% with dry cooling to air at 40 °C and a power block cost of less than 900 $/kWe. Many studies have been conducted to optimize the performance of various sCO2 Brayton cycle configurations in order to achieve high efficiency, and a few have accounted for drivers of cost such as equipment size in the optimization, but complete techno-economic optimization has not been feasible because there are no validated models relating component performance and cost. Reasonably accurate component cost models exist from several sources for conventional equipment including turbines, compressors, and heat exchangers for use in rough order of magnitude cost estimates when assembling a system of conventional equipment. However, cost data from fabricated equipment relevant to sCO2 Brayton cycles is very limited in terms of both supplier variety and performance level with most existing data in the range of 1 MWe power cycles or smaller systems, a single completed system around 7 MWe by Echogen Power Systems, and numerous ROM estimates based on preliminary designs of equipment for 10 MWe systems. This data is highly proprietary as the publication of individual data by any single supplier would damage their market position by potentially allowing other vendors to undercut their stated price rather than competing on reduced manufacturing costs. This paper describes one approach to develop component cost models in order to enable the techno-economic optimization activities needed to guide further research and development while protecting commercially proprietary information from individual vendors. Existing cost models were taken from literature for each major component used in different sCO2 Brayton cycle configurations and adjusted for their magnitude to fit the limited vendor cost data and estimates available. A mean fit curve was developed for each component and used to calculate updated cost comparisons between previously-reviewed sCO2 Brayton cycle configurations for CSP applications including simple recuperated, recompression, cascaded, and mixed-gas combined bifurcation with intercooling cycles. These fitting curves represent an average of the assembled vendor data without revealing any individual vendor cost, and maintain the scaling behavior with performance expected from similar equipment found in literature.