TR-2022-05
Synergistic Datacenter Load Adaptation: Sharing Information with the Grid is Key
Liuzixuan (Peter) Lin; Andrew A Chien. 27 June, 2022.
Communicated by Andrew Chien.
Abstract
The rapid growth of cloud computing has fueled concerns about its growing datacenter (DC) carbon emissions. Dynamic load shifting is one promising solution. Recent work shows that shifting by cloud datacenters (200MW to 1GW) can have adverse impacts on grid carbon emissions and datacenter capacity.To reduce such negative consequences of adaptive DC power management, we study several approaches that attempt to constructively resolve the tension between DC and grid needs. First, we consider DC load control with state-of-the-art online control techniques. Second, we consider adding external coordinators that examine DC load change requests and smooth their aggregate change to reduce negative grid impact. Third, we consider a new approach, in-advance information sharing by DCs with the power grid.
Results show that independent DC control alone or even augmented with coordinators cannot eliminate the adverse impacts of dynamic DC loads. However, an approach with in-advance load adaptation and sharing load-plan information can enable effective grid-wide optimization, delivering 80% of achievable grid benefits.
Original Document
The original document is available in PDF (uploaded 27 June, 2022 by Andrew Chien).
Additional Document Formats
The document is also available in PDF (uploaded 3 December, 2022 by Andrew Chien).
NOTE: The author warrants that these additional documents are identical with the originial to the extent permitted by the translation between the various formats. However, the webmaster has made no effort to verify this claim. If the authenticity of the document is an issue, please always refer to the "Original document." If you find significant alterations, please report to webmaster@cs.uchicago.edu.