Modelling and optimisation of microgrid configuration for green data centres: A metaheuristic approach

Abstract The power load demand of a data centre is in many ways different from that of a residential area. There are differences in the load profiles, the impact of the environmental parameters, the number of peak load demand occurrences in a day as well as peak and off-peak demand. The residential load demand is a complex function of human behaviour which varies based on the geographical, economical, and social characteristics of the consumers. In contrast, estimating the power load of a data centre is rather straightforward. Due to the specific nature of the power load demand of a data centre, finding the optimal configuration of the supplying microgrids can be achieved using off-the-shelf optimisers once an optimisation model has been established. This paper presents a model for finding high-quality microgrid configurations for an enterprise-size green data centres. We present an optimisation model that considers the costs and greenhouse gas emissions associated with all components of the microgrid system, as well as their interactions. The capital, operational, and degradation costs are calculated based on a lifetime of 20 years for the system. Our application of this model to a real scenario of a data centre with a given load demand in a specified environment demonstrates that the model can compose good-quality microgrid configurations for different tradeoffs of cost and sustainability.