Renewables and Grid-Energy Storage Systems

Concurrent Optimization of Capital Cost and Expected O&M for CSP Central Receiver Plants

CSP technologies can utilize heat from concentrated sunlight from a field of tracking mirrors to generate electricity, reform fuel, provide process heat, or augment fossil plant heat sources. The most common CSP conversion systems generate electricity using conventional steam turbines in a Rankine cycle. Among the four major CSP technologies – Parabolic Trough, Linear Fresnel, Dish Stirling, and Power Tower – the lattermost also called a “Central Receiver” system has greatest potential for efficiency improvement and cost reduction. The first US commercial power tower facilities are only recently coming online with BrightSourceTM’s Ivanpah I-III facilities and SolarReserveTM’s Crescent Dunes facility. These facilities represent an important step for CSP in the US, but the relative scarcity of power tower facilities worldwide leaves a dearth of knowledge on operations and maintenance (O&M) costs, performance impacts, and operating strategies to minimize cost of energy.

This project develops a new and validated optimization model that balances component cost, expected operations and maintenance (O&M) cost, operations policies, and operational design targets. Project partners NREL, Northwestern University, Colorado School of Mines, Argonne National Laboratory, and SolarReserve develop detailed performance and cost models leveraging NREL’s System Advisor Model (SAM). These include thermal storage dispatch optimization subject to forecast uncertainty, heliostat and receiver stochastic degradation and failure, and O&M costs. The tool is implemented and disseminated via extensions to SAM and using Argonne National Laboratory’s MINOTAUR open-source optimization toolkit.

FUNDING: U.S. Department of Energy SuNLaMP

Publications

1. Wagner, Michael J., Alexandra M. Newman, William T. Hamilton, and Robert J. Braun. “Optimized
   dispatch in a first-principles concentrating solar power production model.” Applied Energy 203 (2017):
   959-971.