Computility-Power Coupled Scheduling of Data Centers Toward Grid Resilience Enhancement

Youzhen Wu,Jianxiao Wang,Qi An,Zhaohao Ding,Jing Dai
DOI: https://doi.org/10.1109/icpst61417.2024.10602063
2024-01-01
Abstract:In light of escalating natural disasters, the resilience of power systems, particularly concerning data centers, has become an increasingly important area of study. This paper addresses a notable gap in the existing literature by examining the impact of these disasters on data centers and their responses. We introduce innovative workload management strategies for data centers aimed at bolstering grid stability during such crises. Our approach features a two-stage resilience model coupled with a workload scheduling framework, specifically designed to fortify the resilience of crucial data centers against N-K failures. A case study is presented to illustrate the effectiveness of these strategies. The results show a noteworthy enhancement in both resilience and economic efficiency. Notably, our flexible workload scheduling approach led to a reduction in load-shedding by 7517.66 kW and generated an estimated cost saving of $225,528.
What problem does this paper attempt to address?