Using spatial-temporal flexibility of data center building in energy management of distribution grid coupled with multi-energy hubs and energy storage

Due to the progress made and the huge production of data, the need to create data center buildings is increasing. Therefore, data center buildings are one of the most important consumption loads of the distribution network in the coming years. Data center buildings can transfer some of the loads on them to other data center buildings or other operating times, which is the spatial-temporal flexibility of data center buildings. Considering that data center buildings have many processors that generate heat by processing data, the heat produced by these centers can be used in hubs. In this article, a formulation for energy management of the distribution network with multi-energy hubs in the presence of data center buildings and energy storages is presented. First, the formulation of the distribution network with the existence of hubs and the modeling of the equipment inside the hubs including energy storages are presented. Then, by modeling the data center building and defining the spatial-temporal flexibility of this unit, its aggregation has been done in the problem of energy management of the distribution network. Finally, due to the existence of uncertainties in the problem, the final formulation of the problem has been done using the robust optimization method. Considering the distribution network of 33 buses on which 4 hubs are placed, the simulation has been done in two modes of equal distribution of load between data center buildings and distribution based on spatial-temporal flexibility. The amount of losses has increased from 4488 to 5039 kW, and the amount of gas purchased has increased from 40,800 to 41,377 kWh. The objective function value has decreased from 48266 to 43184, and the peak power received from the sub-distribution substation has decreased from 4022 to 3524 kVA. Although in the second case, the amount of losses and purchased gas have increased by about 12 % and 1 %, respectively, but the amount of costs and the peak power received from the sub-distribution substation have decreased by about 10 % and 12 %, respectively. Considering the importance of economic issues and the need to reduce the peak load of the network, it is possible to ignore the increase in losses in contrast to the reduction of these indicators and show that the spatial-temporal flexibility method of data center buildings has improved the scheduling conditions.