Many remote communities and industrial sites across Canada — including northern mines — still depend on diesel for their electricity supply, with high costs and significant environmental impact. Their transition to renewables faces intermittent generation and extreme climate conditions that demand flawless reliability.
This LIREI program develops sizing and management methods for hybrid microgrids integrating renewable generation and energy storage (batteries, hydrogen, thermal): optimal capacity planning algorithms, operating strategies that minimize both emissions and costs, and consideration of demand flexibility and energy recovery mechanisms — for a viable energy transition of remote communities and the mining sector.
Main objective
Develop distributed thermo-energy management strategies for Canadian autonomous grids with high penetration of intermittent renewables, combining planning through Dynamic Operational Envelopes (DOEs) with real-time distributed optimization via Multi-Agent Reinforcement Learning (MARL).
Methodology
The CRSNG strand runs in five phases: (i) data-driven electrothermal modeling with stochastic scenario generation; (ii) a DOE planning framework computed by stochastic optimization with fair allocation mechanisms; (iii) MARL distributed control algorithms operating within the envelopes; (iv) techno-economic and social analysis, with robust co-optimization of hybrid storage sizing and cooperative-game business models; (v) validation through integrated co-simulation against technical, economic and equity indicators.