With the continuous integration of high-proportion renewable energy into islanded microgrids, these systems face the dual challenges of fluctuations in wind and solar power output and seasonal imbalances in energy demand. In isolated grid environments, scalable energy storage technology has become a critical solution to address renewable energy curtailment and insufficient system regulation capacity. However, a single storage technology struggles to economically accommodate the full spectrum of temporal fluctuations, ranging from intra-day variations to seasonal imbalances. To overcome this limitation, this study proposes a temporally driven collaborative planning method for electricity-ammonia-thermal energy storage capacity. By improving the Seasonal-Trend decomposition using Loess (STL) algorithm, the source-load time series curves are decoupled into three independent components: trend, seasonal, and random fluctuation components. A scenario-fitting model is established based on temporal characteristics: seasonal ammonia storage integrated with electric heating serves as the main entity for inter-seasonal energy transfer; battery storage is employed to smooth out daily fluctuations; and a molten salt thermal storage system recovers process waste heat to construct a thermal inertia buffering layer. Through a two-stage progressive optimization process, the device capacities are collaboratively optimized to achieve optimal economic allocation while ensuring system stability. Case validation demonstrates that this method reduces the source-load dynamic mismatch index in island microgrids to 0.69%, while the renewable energy curtailment rate and supply shortage risk are decreased by 28.13% and 18.1%, respectively. These results verify the technical and economical dual optimization characteristics of the “temporal decoupling-scenario adaptation-multi-energy collaboration” planning paradigm in enhancing both system stability and operational efficiency.
Yang et al. (Thu,) studied this question.