Frequency-constrained autonomous microgrid planning for mining industry applications

This paper presents a new frequency-constrained microgrid (MG) planning methodology for mining industry with a high penetration of renewable energy sources (RES). The proposed model is formulated as a multi-objective bi-level optimisation problem in which the upper-level (UL) problem aims to minimise the total net present cost (NPC) of the MG, mitigate the greenhouse gas (GHG) emissions, and improve system reliability. In the lower-level (LL) problem, the operation of the MG is simulated considering unit commitment, battery operation, and frequency stability constraints. The non-dominated sorting genetic algorithm II (NSGA-II) is employed to solve the UL multi-objective optimisation problem, generating candidate solutions that define the MG’s energy source capacities. Each candidate solution is evaluated by solving the LL problem, formulated as a mixed-integer linear programming (MILP) problem and solved using Gurobi solver. This iterative process ensures accurate evaluation of operational costs and emissions while exploring trade-offs between objectives. The fuzzy decision-making method is then applied to the Pareto optimal solutions to obtain the final optimal plan. Through this model, optimal capacities of RES units, battery energy storage systems (BESSs), and back-up fossil fuel generators are obtained to meet the MG demand. Moreover, the proposed approach ensures that the frequency stability requirements, including rate of change of frequency (RoCoF), minimum/maximum frequency and steady-state frequency remain within acceptable thresholds after considering any imbalance events. Finally, the simulation study is conducted to validate the effectiveness of the proposed model using a real-world MG in a remote underground mine in Australia utilising historical load, expected RES generation, and commodity/capital price data. The results demonstrate that the proposed model effectively ensures compliance with frequency stability requirements–always maintaining frequency above 49.5 Hz and RoCoF below 0.5 Hz/s–while achieving a balanced trade-off between cost and emissions. Compared to the conventional approach, the proposed solution results in a slightly higher NPC (2.8 %) and GHG emissions (4.5 %), but significantly reduces RES curtailment (13 %) and eliminates severe frequency deviations, which in the conventional case occurred 60 % of the time with drops as low as 48.9 Hz and RoCoF as high as 0.875 Hz/s. This highlights the critical role of BESS in enhancing both economic performance and frequency stability in off-grid mining MGs.