Load Management for Multiple Datacenters towards Demand Response in the Smart Grid Integrating Renewable Energy

Peicong Luo,Xiaoying Wang,Hailong Jin,Yuling Li,Xuejiao Yang

DOI: https://doi.org/10.1145/3297156.3297197

2018-01-01

Abstract:Recently, as the renewable and distributed power sources boost, many such resources are integrated into the smart grid as clean energy input. However, since the generation of solar power is intermittent and unstable, the smart grid needs to regulate the load to keep the stability after integrating the renewable energy source. At the same time, with the development of the cloud computing, large-scale datacenters are becoming good potential controllable loads for the smart grid due to their high energy consumption. In this paper, we propose an appropriate approach to dynamically adjust the datacenter load to balance the unstable solar input into the grid. This could meet the demand response requirements by taking the advantage of the variable power consumption of datacenters. We examine the scenarios of one or more datacenters integrated into the grid and use a stochastic algorithm to solve the problem we established. From the experimental results, we found that the dynamic load management of multiple datacenters can help the smart grid to reduce loss and thus save operational costs. Besides, we also analyzed the impact of the flexibility and the delay of datacenter actions, which can apply to more general scenarios in realistic environments.