Energy hub-based optimal planning framework for user-level integrated energy systems: Considering synergistic effects under multiple uncertainties.

The design of user-level integrated energy systems is challenged by a variety of combinations of energy converters, complex cascade utilization of multiple energy flows, as well as sequential matching between stochastic energy resources and periodic energy demands. These issues can be addressed by an optimal energy-hub based planning framework considering the synergistic effects under multiple uncertainties. The energy hub model is extended to analyze energy-level matching and source–load balance with time-varying coupling factors representing part-load characteristics. The optimization problem is formulated as a bi-level planning model with uncertainties evaluated by a two-stage global sensitivity analysis. The bi-level planning model determines the system structure and component sizing at the upper level and identifies the optimal operation strategy at the lower level by employing piecewise linearization of part-load characteristics of components involved. The global sensitivity analysis reduces model size with the elementary effect method, identifies the most influential uncertain parameters with a variance-based method. A case study in Beijing is demonstrated for the proposed methodology. The results show that the proposed method can effectively plan the integrated energy system considering sequential source–load matching with the rational scheduling strategy of components. The demand-side response influences the system configuration and renewable energy penetration. Integrating the components' part-load characteristics help avoid the mismatch between the component capacity and energy demand, reducing 6.8% cost compared to the scheme with constant energy efficiency. The three most influential factors identified among 551 uncertain parameters are natural gas price, valley electricity price and nominal efficiency of the gas turbine. • Energy planning framework considering synergistic effects under uncertainties. • Extending energy hub model for energy-level match and source–load balance. • Bi-level linear programming considering components' part-load characteristics. • Global sensitivity analysis performed to analyze multiple uncertainties. • Identifying the most influential factors among 551 uncertain parameters. [ABSTRACT FROM AUTHOR]