Data Center Power Play

The document details the use of a modified Regional Energy Deployment System (ReEDS) model to project US electricity demand from data centers and the resulting effects on the energy transition. It outlines five demand scenarios, including 'high data center demand growth', and evaluates them against various policy frameworks such as the Inflation Reduction Act (IRA) and the One Big Beautiful Bill Act (OBBBA). The analysis incorporates specific state-level emissions caps for Illinois, Michigan, and Wisconsin, and adjusts technology costs for nuclear and carbon capture and storage (CCS) to reflect conservative estimates. It also details the calculation of health damages from criteria pollutants and the social cost of greenhouse gas emissions.

• Under the reference demand growth case, data centers are projected to increase their share of total US electricity demand from 4.4 percent in 2023 to 11.6 percent by 2030 and 15 percent by 2050.
• Data centers are identified as the primary driver of electricity demand growth in the near term, accounting for 46 percent of the increase between 2025 and 2030, though this share drops to 27 percent by 2050 as other sectors, particularly transportation, electrify.
• The analysis assumes improvements in power usage efficiency (PUE) will reduce data center energy use by nearly 11 percent from 2024 to 2028 and 18 percent by 2035, driven by liquid-cooled servers and a shift toward co-located and hyperscale facilities.
• The model incorporates conservative cost assumptions for new nuclear plants and carbon capture and storage (CCS) projects, noting that actual CCS projects have historically achieved lower capture rates than the 95 percent rate assumed in the analysis.
• The One Big Beautiful Bill Act (OBBBA) is modeled as having more stringent restrictions on the source of manufactured products for clean energy projects starting January 1, 2026, which the authors assume will make lithium-ion battery storage projects unable to meet these requirements after 2026 due to reliance on Chinese imports.
• The study models specific decarbonization targets for three states: Illinois (100 percent carbon-free electricity by 2050), Michigan (100 percent by 2045), and Wisconsin (100 percent clean energy standard by 2050).

APA

Clemmer, S., Chavez, M., Dotson, S., Gignac, J., Sattler, S., & Shaver, L. (2026). Data Center Power Play. Union of Concerned Scientists. https://www.ucs.org/sites/default/files/2026-01/Data-Center-Power-Play-appendix.pdf

Chicago

Clemmer, Steve, Maria Chavez, Samuel Dotson, James Gignac, Sandra Sattler, and Lee Shaver. Data Center Power Play. Union of Concerned Scientists, 2026. https://www.ucs.org/sites/default/files/2026-01/Data-Center-Power-Play-appendix.pdf.

Wikipedia

{{cite report |last1=Clemmer |first1=Steve |last2=Chavez |first2=Maria |last3=Dotson |first3=Samuel |last4=Gignac |first4=James |last5=Sattler |first5=Sandra |last6=Shaver |first6=Lee |title=Data Center Power Play |publisher=Union of Concerned Scientists |date=January 2026 |url=https://www.ucs.org/sites/default/files/2026-01/Data-Center-Power-Play-appendix.pdf |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{clemmer2026data, author = {Clemmer, Steve and Chavez, Maria and Dotson, Samuel and Gignac, James and Sattler, Sandra and Shaver, Lee}, title = {{Data Center Power Play}}, institution = {Union of Concerned Scientists}, year = {2026}, month = jan, url = {https://www.ucs.org/sites/default/files/2026-01/Data-Center-Power-Play-appendix.pdf}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }