Data center power control for frequency regulation

The power consumption of a data center with a large number of computer servers can be controlled precisely on fast timescales, making data centers rather valuable assets to the electric grid. This paper explores the potential of using large-scale data centers to provide frequency regulation. In particular, Dynamic Voltage Frequency Scaling (DVFS) is employed to adjust the power consumptions of individual computer servers. An aggregated power response model is developed for the entire data center. Service response time model is incorporated to ensure performance of the data center. A novel power management strategy is then proposed to control the CPU frequencies of individual servers to follow a given regulation signal while respecting desired response time requirement. Simulations based on real workload traces and real regulation signals are performed. The results indicate that the regulation signal can be tracked very accurately under the proposed control strategy, and a single data center can achieve about a half-million-dollar revenue by participating in the regulation market without affecting its service quality.

[1]  Anand Sivasubramaniam,et al.  Benefits and limitations of tapping into stored energy for datacenters , 2011, 2011 38th Annual International Symposium on Computer Architecture (ISCA).

[2]  Yefu Wang,et al.  Capping the electricity cost of cloud-scale data centers with impacts on power markets , 2011, HPDC '11.

[3]  Myron Hlynka,et al.  Queueing Networks and Markov Chains (Modeling and Performance Evaluation With Computer Science Applications) , 2007, Technometrics.

[4]  Luiz André Barroso,et al.  The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines , 2009, The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines.

[5]  Arnold O. Allen,et al.  Probability, statistics and queueing theory - with computer science applications (2. ed.) , 1981, Int. CMG Conference.

[6]  Margaret Martonosi,et al.  Computer Architecture Techniques for Power-Efficiency , 2008, Computer Architecture Techniques for Power-Efficiency.

[7]  Xue Liu,et al.  Minimizing Electricity Cost: Optimization of Distributed Internet Data Centers in a Multi-Electricity-Market Environment , 2010, 2010 Proceedings IEEE INFOCOM.

[8]  J. Koomey.  Worldwide electricity used in data centers , 2008 .

[9]  T. N. Vijaykumar,et al.  Joint optimization of idle and cooling power in data centers while maintaining response time , 2010, ASPLOS XV.

[10]  S. Borkar,et al.  Total power optimization by simultaneous dual-Vt allocation and device sizing in high performance microprocessors , 2002, Proceedings 2002 Design Automation Conference (IEEE Cat. No.02CH37324).

[11]  Duncan S. Callaway.  Tapping the energy storage potential in electric loads to deliver load following and regulation, with application to wind energy , 2009 .

[12]  François Bouffard,et al.  Decentralized Demand-Side Contribution to Primary Frequency Control , 2011, IEEE Transactions on Power Systems.

[13]  Richard E. Brown,et al.  Report to Congress on Server and Data Center Energy Efficiency: Public Law 109-431 , 2008 .

[14]  Sandia Report,et al.  Energy Storage for the Electricity Grid: Benefits and Market Potential Assessment Guide A Study for the DOE Energy Storage Systems Program , 2010 .

[15]  David M. Auslander,et al.  Using load switches to control aggregated electricity demand for load following and regulation , 2011, 2011 IEEE Power and Energy Society General Meeting.

[16]  Xiaorui Wang,et al.  SHIP: Scalable Hierarchical Power Control for Large-Scale Data Centers , 2009, 2009 18th International Conference on Parallel Architectures and Compilation Techniques.

[17]  Christoforos E. Kozyrakis,et al.  A Comparison of High-Level Full-System Power Models , 2008, HotPower.

[18]  Sean P. Meyn,et al.  How demand response from commercial buildings will provide the regulation needs of the grid , 2012, 2012 50th Annual Allerton Conference on Communication, Control, and Computing (Allerton).