No Detail Found!

Emergency Auxiliary Services: A Bi-Directional Mutual Beneficial Framework for Power Systems and Data Centers

Author(s)
M. Jawad Electrical Engineering, COMSATS University Islamabad, Lahore Campus,Pakistan
S. M. Ali Electrical Engineering, COMSATS University Islamabad, Abbottabad Campus,Pakistan
Z. Ullah Electrical Engineering Department, UMT Lahore, Sialkot Campus,Pakistan
M. U. S. Khan Department of Computer Science, COMSATS University Islamabad, Abbottabad Campus,Pakistan
C. A. Mehmood Electrical Engineering, COMSATS University Islamabad, Abbottabad Campus,Pakistan
B. Khan Electrical Engineering, COMSATS University Islamabad, Abbottabad Campus,Pakistan
A. Fayyaz Electrical Engineering, COMSATS University Islamabad, Abbottabad Campus,Pakistan
Abstract
The power systems are facing a gradual increase in electrical load at different hours of the day and on different buses of the network. This situation can create the problem of demand/supply mismanagement in the power system. Therefore, the power systems need fast auxiliary services to keep power management, stability, and reliability in the network. Conventionally, power systems have own dedicated computing facility for executing auxiliary services, however, data centers are among the largest energy consumption clients for the power systems and have the capability to provide enough computational resources to the power system when required. This paper proposes an Emergency Auxiliary Services (EAS) model for power systems and data centers to work combinedly with mutual benefits. A dynamic Service Level Agreement (SLA) is introduced along with an EAS job scheduling algorithm that motivates data center to run power system jobs on priority and effectively during emergency conditions and maintain data center revenue. The EAS includes Optimal Power Flow (OPF) analysis, bus centrality index, and transmission line centrality index. The simulations are performed on real workload of a data center integrated with the IEEE 30-bus system to assess the performance of the proposed model. The results illustrate that the priority execution EAS on data centers has a minimal impact on overall energy consumption and on other cloud computing jobs' time of execution. Moreover, the dynamic SLA compensates the data center revenue loss due to prior execution of the EAS. Therefore, the SLA encourages the data center operators to execute EAS on priority.
Publication Details
Page(s) 52-63
DOI DOI not available
Published Journal: Pakistan Journal of Engineering and Applied Sciences, Volume: 25, Issue: 0, Year: 2019

Request the full text from PASTIC

Keywords
optimization, Data center, Scheduling, Smart grid, Load flow
0

Citations

0

Downloads

105

Views