Dynamic data center load response to variability in private and public electricity costs

This paper assesses the potential cost-saving incentives for content distribution networks to shift traffic load among geographically distributed data centers in response to hourly variation in electricity prices. Such incentives are likely aligned with benefits to utilities and grid operators, which might take the form of peak-shaving or ancillary services. However, private cost savings are not strictly aligned with public benefits related to the avoidance of health and environmental damages from power plant emissions, so we compare private cost minimization with a strategy that minimizes these externalities. We find that feasible strategies exist to simultaneously realize public and private benefits and that load shifting can result in substantial cost savings and avoided damages in some circumstances. Concerns over increased latency and bandwidth costs can be mitigated with modifications to the model. However, the level of realized savings is dependent upon the specifics of a particular network operator and electricity rate schedule.

[1]  Hamidreza Zareipour,et al.  Data centres in the ancillary services market , 2012, 2012 International Green Computing Conference (IGCC).

[2]  Tajana Simunic,et al.  Providing regulation services and managing data center peak power budgets , 2014, 2014 Design, Automation & Test in Europe Conference & Exhibition (DATE).

[3]  Severin Borenstein Electricity pricing that reflects its real-time cost , 2009 .

[4]  G. Goldman,et al.  A Survey of Utility Experience with Real Time Pricing Author , 2004 .

[5]  Inês L. Azevedo,et al.  Power usage effectiveness in data centers: overloaded and underachieving , 2016 .

[6]  Richard E. Brown,et al.  United States Data Center Energy Usage Report , 2016 .

[7]  Thu D. Nguyen,et al.  Cost-and Energy-Aware Load Distribution Across Data Centers , 2009 .

[8]  M Granger Morgan,et al.  Marginal emissions factors for the U.S. electricity system. , 2012, Environmental science & technology.

[9]  Girish Ghatikar,et al.  Demand Response Opportunities and Enabling Technologies for Data Centers: Findings From Field Studies , 2012 .

[10]  Adam Wierman,et al.  Data center demand response: avoiding the coincident peak via workload shifting and local generation , 2013, SIGMETRICS '13.

[11]  Srinivasan Keshav,et al.  It's not easy being green , 2012, CCRV.

[12]  David Costenaro,et al.  The Megawatts behind Your Megabytes: Going from Data-Center to Desktop , 2012 .

[13]  Bruce M. Maggs,et al.  Cutting the electric bill for internet-scale systems , 2009, SIGCOMM '09.

[14]  William D. Nordhaus,et al.  Environmental Accounting for Pollution in the United States Economy , 2011 .

[15]  Lachlan L. H. Andrew,et al.  Greening geographical load balancing , 2011, PERV.

[16]  Peter J. Adams,et al.  Reduced-form modeling of public health impacts of inorganic PM2.5 and precursor emissions , 2016 .

[17]  Dag Lunden,et al.  Life Cycle Assessment of ICT , 2014 .