Bi-Level Optimisation Model for Harvesting Spatial-Temporal Load Shifting Flexibility of Data Centres Using Endogenously Formed Locational Price Signal

Abstract

With the rapid advancement of digital information technology, the demand for computational power continues to grow, driving the expansion of the data centre (DC) industry, leading to gigantic amount of energy consumption and greenhouse gas emission. DCs, serving as flexible and adjustable resources on the load side of power systems, are essential to the operation of modern energy systems. In this context, this paper proposes a coordinated and secure scheduling strategy for DCs and power system, utilising spatial-temporal flexibility. First, the architecture and operational characteristics of DCs are analysed, followed by the development of an power consumption model to quantify their spatial-temporal flexibility. Then, a bi-level optimisation model is developed to coordinate computational power and electricity, considering DCs' dual roles as independent operators and flexible resources that influence locational electricity prices. Finally, simulation analyses are performed on the modified IEEE 39-node system. The results demonstrate that the proposed strategy effectively utilises the spatial-temporal flexibility of DCs, alleviating system congestion. In terms of economic benefits, it enhances the cost efficiency of the DCs by 70% and reduces the system's operating costs by 46%. Regarding environmental impact, the strategy increases renewable energy consumption by 26.58% and significantly reduces carbon emissions by 63%.