An Evaluation of Feedback and Model Predictive Control for the Rejection of Weather Transients in Renewable High-Temperature Solar Thermal Chemical Processing on a Novel Hybrid Solar-Electric Reactor

Concentrated solar power, which has been commercialized for electrical generation, is now of interest as a means of directly driving chemical processes. Investigations have included the use of solar heat to produce metals from oxide ores, generate renewable hydrogen for use as fuel, and to gasify biomass. These processes often operate at high temperatures where reactors and receivers are built partly or wholly from ceramics, ceramics that are vulnerable to thermal shock and fatigue. Weather transients, which can deprive a facility of solar heat within seconds, must be controlled to insure the longevity of solar-driven chemical process equipment. Supplementary electrical and/or gas heaters can be used to provide auxiliary heat when transient cloud cover would otherwise interrupt solar thermal chemical production. We evaluate different control schemes on a new solar-electric reactor prototype at the University of Colorado Boulder. These include both feedback and model predictive control paradigms, whose performance was measured by integral squared absolute error over the course of programmed weather transients on a 45kWe high-flux solar simulator. Implications for the implementation of commercial solar thermal processes will be discussed.