Building Accurate Energy-Use Statistics for Data Centers

Yong-Zhen Wang; Te Han; Yi-Ming Wei

doi:10.1016/j.eng.2025.12.014

Engineering ›› :202512014 DOI: 10.1016/j.eng.2025.12.014

Research Engineering Management—Perspective

research-article

Building Accurate Energy-Use Statistics for Data Centers

Yong-Zhen Wang ^a^,^b
, Te Han ^b^,^c^,^d^,^e^,^*
, Yi-Ming Wei ^b^,^c^,^d^,^e^,^*

Author information +

History +

PDF (1547KB)

Abstract

With the rapid expansion of cloud computing and large-scale artificial intelligence models, building accurate and transparent energy-use statistics for data centers has become a critical challenge for global energy systems and climate governance. Existing studies report strikingly divergent estimates of global data center electricity consumption, ranging from 196 to 1200 TW h in 2020, a more than sixfold difference. Such discrepancies reveal profound uncertainties and structural deficiencies in current energy accounting frameworks. Conventional estimation approaches rely heavily on indirect assumptions, proxy indicators, or highly aggregated regional and national statistics, obscuring the true electricity demand of data centers. This lack of statistical transparency distorts energy and carbon accounting, weakens power system planning, and constrains the effective integration of renewable energy with rapidly growing computing demand. This paper highlights that data centers should be treated as a distinct and strategically important end-use energy sector. It emphasizes the need for grid-informed energy registration, enhanced artificial intelligence identification techniques to improve the accuracy and verifiability of energy statistics. Furthermore, the paper emphasizes that policymakers should establish coordinated policy frameworks, enforce standardized energy reporting, and design appropriate incentive mechanisms to encourage data centers to participate in demand response programs and electricity markets, thereby unlocking load flexibility and supporting a secure, low-carbon energy transition.

Cite this article

Download citation ▾

Yong-Zhen Wang, Te Han, Yi-Ming Wei. Building Accurate Energy-Use Statistics for Data Centers. Engineering 202512014 DOI:10.1016/j.eng.2025.12.014

登录浏览全文

4963

注册一个新账户忘记密码

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Masanet E, Shehabi A, Lei N, Smith S, Koomey J. Recalibrating global data center energy-use estimates. Science 2020; 367(6481):984-6.

[2]	Crawford K. Generative AI’s environmental costs are soaring—and mostly secret. Nature 2024; 626(8000):693.

[3]	Debus C, Piraud M, Streit A, Theis F, Götz M. Reporting electricity consumption is essential for sustainable AI. Nat Mach Intell 2023; 5(11):1176-8.

[4]	Mytton D, Ashtine M. Sources of data center energy estimates: a comprehensive review. Joule 2022; 6(9):2032-56.

[5]	Kamiya G, Bertoldi P. Energy consumption in data centres and broadband communication networks in the EU. Report. Luxembourg: Publications Office of the European Union; 2024.

[6]	Masanet E, Lei N, Koomey J. To better understand AI’s growing energy use, analysts need a data revolution. Joule 2024; 8(9):2427-36.

[7]	Tang G, Jiang W, Xu Z, Liu F, Wu K. NIPD: non-intrusive power disaggregation in legacy datacenters. IEEE Trans Comput 2017; 66(2):312-25.

[8]	De Vries A. The growing energy footprint of artificial intelligence. Joule 2023; 7(10):2191-4.

[9]	Lei N, Masanet E. Climate-and technology-specific PUE and WUE estimations for U.S. data centers using a hybrid statistical and thermodynamics-based approach. Resour Conserv Recycl 2022;182:106323.

[10]	Pei Q, Yuan Y, Hu H, Wang L, Zhang D, Yan B, et al. Working smarter not harder: hybrid cooling for deep learning in edge datacenters. IEEE Trans Sustain Comput 2025; 10(4):804-19.

[11]	Standard Performance Evaluation Corporation (SPEC). SPECpower ssj® 2008 Results-Query [Internet]. Gainesville: Standard Performance Evaluation Corporation; undated [cited 2025 Nov 28]. Available from: https://www.spec.org/cgi-bin/osgresults?conf=power_ssj2008.

[12]	Li Y, Hu H, Wen Y, Zhang J.Power data classification: a hybrid of a novel local time warping and LSTM. arXiv:1608.04171.

[13]	Karimi M, Arab F.Machine learning based framework for estimation of data center power using acoustic side channel. arXiv:2008.02481.

[14]	Tanoni G, Principi E, Squartini S. Non-intrusive load monitoring in industrial settings: a systematic review. Renew Sustain Energy Rev 2024;202:114703.

[15]	Li Y, Yao R, Qin D, Wang Y. Lightweight federated learning for on-device non-intrusive load monitoring. IEEE Trans Smart Grid 2025; 16(2):1950-61.

[16]	Frijling R. EED audit and energy saving obligations for Dutch data centers [Internet]. Lincoln: EkkoSense; 2024 May 16 [cited 2025 Apr 3]. Available from: https://www.ekkosense.com/resources/industry-insight/eed-audit-and-energy-saving-obligations-for-dutch-data-centers.

[17]	Avgerinou M, Bertoldi P, Castellazzi L. Trends in data centre energy consumption under the European code of conduct for data centre energy efficiency. Energies 2017; 10(10):1470.

[18]

European Commission. Commission adopts EU-wide scheme for rating sustainability of data centres [Internet]. Brussel: European Commission; 2024 Mar 15 [cited 2025 Apr 3]. Available from: https://energy.ec.europa.eu/news/commission-adopts-eu-wide-scheme-rating-sustainability-data-centres-2024-03-15_en.

[19]	Energy Authority. Reporting from data centres to the European database has started. Jacksonville: Energy Authority; 2024 Oct 14 [cited 2025 Apr 3]. Available from: https://energiavirasto.fi/en/-/reporting-from-data-centres-to-the-european-database-has-started.

[20]

European Commission. Commission Delegated Regulation (EU) 2024/1364 of 14 March 2024 on the first phase of the establishment of a common Union rating scheme for data centres [Internet]. Brussel: European Commission; 2024 Mar 15 [cited 2025 Apr 3]. Available from: https://eur-lex.europa.eu/eli/reg_del/2024/1364/oj.

[21]

Zhu X, Jiang W, Liu F, Zhang Q, Pan L, Chen Q, et al. Heat to power: thermal energy harvesting and recycling for warm water-cooled datacenters. In: Proceedings of the 2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture (ISCA); 2020 May 30-Jun 3; held virtually. New York City: IEEE Press; 2020. p. 405-18.

[22]	Wang Y, Han Y, Shen J, Han K, Li J, Ji W, et al. Data center integrated energy system for sustainability: generalization, approaches, methods, techniques, and future perspectives. Innov Energy 2024; 1(1):100014.

[23]

Pei Q, Wang L, Zhang D, Yan B, Yu C, Liu F. InferCool: enhancing AI inference cooling through transparent, non-intrusive task reassignment. In: Proceedings of the 2024 ACM Symposium on Cloud Computing, SoCC’24; 2024 Nov 20-22; Redmond, WA, USA. New York City: Association for Computing Machinery (ACM); 2024. p. 487-504.

[24]	Liu F, Zhou Z, Jin H, Li B, Li B, Jiang H. On arbitrating the power-performance tradeoff in SaaS Clouds. IEEE Trans Parallel Distrib Syst 2024; 25(10):2648-58.

[25]	Cao Y, Cheng M, Zhang S, Mao H, Wang P, Li C, et al. Data-driven flexibility assessment for internet data center towards periodic batch workloads. Appl Energy 2022;324:119665.

[26]	Zeng B, Zhou Y, Xu X, Cai D. Bi-level planning approach for incorporating the demand-side flexibility of cloud data centers under electricity-carbon markets. Appl Energy 2024;357:122406.

[27]	Han O, Ding T, Mu C, Jia W, Ma Z. Waste heat reutilization and integrated demand response for decentralized optimization of data centers. Energy 2023;264:126101.