Microgrids have been extensively studied topic in power systems, and are a common solution for remote industrial locations, such as mining sites and factories; or in geographically isolated locations which don’t allow for the connection to a larger power system, such as islands. The operation of these isolated energy grids is facing a significant challenge from the pressing needs for decarbonisation and sustainable energy generation and use. These challenges are supported by advances in grid-scale energy storage and renewable energy generation. A key component in enabling the decarbonisation of microgrids lies in energy management systems (EMS) which target the operational optimisation of these grids, ensuring these small and isolated grids remain stable while still reducing its overall carbon emissions.
Assessing EMS performance in microgrids is a significant challenge due to the unique operational constraints and combination of assets in microgrids. Moreover, optimal performance depends on the context in which a microgrid is operated: Industrial microgrids, such as the ones developed to support mining or factories, can be measured in terms of operational reliability and costs of energy dispatch. In parallel, distribution microgrids in islands, can be measured in terms of sustainability, renewable energy penetration, or energy availability. Finally, these microgrid performance indicators are often influenced by external factors, such as fuel costs or grid faults caused by extreme weather events.
This work proposes a methodology, entitled an “Optimisation Key Performance Indicator (KPI)”, to evaluate the performance of an energy management system by comparing the operational performance of a microgrid against its theoretical optimal performance. This approach eliminates the influence of external factors in the performance assessment. This qualifies the Optimisation KPI as a measure of energy management system performance that allows for benchmarking the performance of different EMS for a single microgrid; or to benchmark the performance of one EMS in different contexts. The methodology also enables the consideration of different parameters, such as energy production price, renewable penetration, carbon emissions or reliability aspects such as SAIFI or CAIDI as key components of the performance, or whichever parameters are most relevant for the context in which the EMS is deployed.
The methodology was developed as an internal project and was based on both synthetic simulation data and real-world data gathered from partners and customers. The results presented in the paper are gathered with a simulated dataset that represents typical microgrid operations in different contexts and scales.