(706c) Dynamic Optimization to Leverage Flexible Heat Integration within a Hybrid CSP Plant

A previous study presented the concept of flexible heat integration (FHI) into a solar central receiver-Rankine cycle (SCR-RC) power plant hybridized with natural gas and integrated with additional thermal energy storage (TES) tanks. FHI works by independently dispatching harnessed solar energy, via heated molten salt, to a steam superheater, steam generator, and feedwater preheater based upon solar conditions rather than in sequence as is typical in an SCR-RC power plant. With previous studies, operation of this FHI plant is carried out by having pre-determined decisions variables and logic operations mapped out based upon a known solar-activity schedule. These decision variables and logic operators include variables such as collector temperature, storage discharge rates, TES charge/discharge rates, and discharge times. To implement FHI in real-time, more sophisticated methods are needed. First, machine learning techniques can be employed to create an accurate forecast of solar activity and weather which would govern the future control decisions. Then, using dynamic optimization, real-time decisions can be made to automate control of the power plant utilizing FHI which can lead to further improvement of solar energy utilization. This study highlights implementation of dynamic optimization to leverage the FHI design and details the dynamics and performance of the plant SCR-RC when optimized versus a non-optimized case.