Tie-line Power Fluctuation Smoothing Algorithm Based on Data Center Demand Response

The exploit of renewable energy can provide an effective solution to solve the problem of enormous energy consumption and operation cost in datacenters. While comparing with the traditional power supply mode, renewable energy has its unique characteristics, such as intermittency and randomness. Thus, when renewable energy supplies power to the datacenter industrial park, this kind of power supply mode not only has bad effects on the normal operation of equipment in datacenters and the power supply reliability of the industrial park, but also it would impact the stability of the main gird operation. After analyzing the characteristics of datacenter structure and different kinds of user requests, we acknowledge that the server cluster is a controllable workload. Furthermore, with the occupation of the redundant uninterruptible power system（UPS）, a new data center tie-line power control method was proposed. A server cluster work load tuning model was proposed based on task delay mechanism. A demand side response controllable work load model of server clusters and a demand side response controllable workload model of UPS were established. Based on the models a novel datacenter industrial park tie-line power control method was proposed.. On the premise of ensuring service-level agreement（SLA） and power supply reliability, this method can regulate the tie-line power fluctuations between the datacenter industrial park and main grid effectively by dynamic adjusting the server cluster workload and the state of charge of UPS energy batteries. With the experiments, the proposed control method was evaluated; furthermore multiple parameters which would influence theeffect of smoothing the power fluctuation in tie-line were analyzed as well.