Green energy mix modeling under supply uncertainty: Hybrid system dynamics and adaptive PSO approach

Reducing global warming is crucial for sustainability. Carbon emissions primarily stem from energy supply, prompting a shift from Non-Renewable Energy Supply (NRES) to Renewable Energy Supply (RES). However, transitioning to RES entails substantial investment and faces supply uncertainty due to weather dependency. Therefore, the transition should be done gradually, requiring a reliable approach to energy mix modeling. This study proposes a System Dynamics framework integrated with Adaptive Particle Swarm Optimization (APSO) and Machine Learning to optimize the energy mix under supply uncertainty. Due to energy system dynamicity, the proposed framework considers not only supply, but also demand, energy storage, electric vehicle, and emission subsystems. The experiment has been conducted by taking the United States as a case under various scenarios namely to minimize the total system cost, total carbon emissions, and both, accounting for the static and dynamic cost of RES. Results of this study reveal four main points: (i) A 38% reduction in total system cost is achievable by decreasing the RES Ratio to 6%, but total emissions will rise by 8%; (ii) A 55% reduction in total emissions is possible by directly transitioning to 100% RES, but total system cost increases by 68%; (iii) Both objective functions can be significantly minimized at a time by increasing the RES ratio; (iv) Dynamic cost offers a better opportunity for reducing costs and emissions than static cost.