pith. sign in

arxiv: 2412.06288 · v3 · pith:ZJQE6ACAnew · submitted 2024-12-09 · 💻 cs.CY

The Unpaid Toll: Estimating and Addressing the Public Health Impact of Data Centers

Yuelin Han , Zhifeng Wu , Pengfei Li , Adam Wierman , Shaolei Ren This is my paper

Pith reviewed 2026-05-23 07:52 UTC · model grok-4.3

classification 💻 cs.CY
keywords data centerspublic healthair pollutionAI energy demandhealth costssustainabilityresource managementelectricity consumption
0
0 comments X

The pith

Growing AI demand will drive U.S. data center air pollution health costs above $20 billion per year by 2028.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper develops a method to quantify air pollutant emissions from electricity generation serving data centers and converts those emissions into estimated public health costs. It projects that rising computing and AI needs will push the nationwide annual total above $20 billion by 2028. Although the national figure remains modest relative to other sources, the costs concentrate heavily in certain counties, where per-household burdens can reach seven times the average. The authors then present a management approach that factors health impacts into decisions about where and when data centers draw power. The estimates matter because they surface a concrete external cost of computing expansion that standard environmental reviews have largely omitted.

Core claim

The paper claims that criteria air pollutant emissions tied to U.S. data center electricity use generate public health burdens projected to exceed $20 billion annually by 2028 under continued AI growth, with sharp geographic disparities that place some counties under per-household loads seven times the national average, and shows that a health-informed computing framework can reduce those burdens by guiding resource allocation across space and time.

What carries the argument

A methodology that links data center electricity consumption to power-plant emission factors, then monetizes the resulting criteria air pollutant health damages, paired with a health-informed computing framework that adds those damages to data center scheduling and placement decisions.

If this is right

  • National public health costs from data centers will scale directly with AI and computing demand growth.
  • Health burdens will remain geographically concentrated, producing local impacts several times the national average in the hardest-hit counties.
  • Resource management that explicitly includes health costs can lower total damages while preserving sustainability targets.
  • Energy disclosure rules should expand to report public health impacts alongside carbon metrics.
  • Management and policy attention must extend to every community affected by the associated emissions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The uneven distribution could prompt regulators to require health impact assessments before approving new data center sites.
  • The management framework might be tested first in high-burden regions to measure actual reductions in modeled health costs.
  • Similar emission-to-health-cost modeling could be applied to other large, flexible electricity loads such as cryptocurrency mining or hydrogen production.
  • If the projections hold, they supply a quantitative basis for communities to negotiate mitigation payments or renewable energy requirements from data center operators.

Load-bearing premise

The calculations rest on emission factors, grid mix assumptions, and health damage valuations whose accuracy is not fully validated in the presented work.

What would settle it

Independent county-level air quality or health outcome data collected before and after major data center expansions or shutdowns that either matches or deviates from the model's per-household cost predictions would confirm or refute the estimates.

Figures

Figures reproduced from arXiv: 2412.06288 by Adam Wierman, Pengfei Li, Shaolei Ren, Yuelin Han, Zhifeng Wu.

Figure 1
Figure 1. Figure 1: The overview of data centers’ contribution to air pollutants and public health impacts. Scope-1 and scope-2 impacts occur during the operation of data centers (“operational”), whereas scope-3 impacts arise from activities across the supply chain (“embodied”). Under the Clean Air Act, the U.S. EPA is authorized to regulate the emission levels of criteria air pollu￾tants, reducing concentrations to comply wi… view at source ↗
Figure 2
Figure 2. Figure 2: The county-level total scope-1 health cost of data center backup generators operated in Virginia (mostly in Loudoun County, Fairfax County, and Prince William County) [62]. The backup generators are assumed to emit air pollutants at 10% of the permitted levels per year. The total annual public health cost is $220-300 million, including $190-260 million incurred in Virginia, West Virginia, Maryland, Pennsyl… view at source ↗
Figure 3
Figure 3. Figure 3: Public health costs of electricity generation and on-road emissions in the contiguous U.S. in 2023 and 2028 [39]. The error bars represent high and low esti￾mates returned by COBRA using two different exposure￾response functions. Based on the emission data projected by the U.S. EPA’s COBRA modeling tool [39], we show in [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The county-level total health cost of U.S. data centers from 2019 to 2023. (a) Health cost map; (b) CDF of county-level health cost; (c) Top-10 counties by total health cost. 5.1.2 Uneven distribution of data centers’ public health impacts. Next, [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: The county-level per-household health cost of U.S. data centers from 2019 to 2023. (a) Per-household health cost map; (b) CDF of county-level per-household health cost; (c) Top-10 counties by per-household health cost. IR represents “County-to-Nation Per-Household Median Income Ratio.” different communities in terms of the public health cost suggests that we need to carefully examine the local and regional… view at source ↗
Figure 7
Figure 7. Figure 7: The county-level per-household health cost of two U.S. technology companies in 2023. 5.2 Public Health Impact of Generative AI Training We now study the health impact of a specific computing task. Specifically, we consider the training of an LLM and assume that the electricity consumption is the same as training Llama-3.1 recently released by Meta [92]. As the scope-2 impact is dominant and the power alloc… view at source ↗
Figure 8
Figure 8. Figure 8: Analysis of marginal scope-2 carbon emission rates and public health costs over 114 U.S. regions between October 1, 2023 and September 30, 2024 [77]. (a) In 110 out of the 114 U.S. regions (96%), the normalized IQR of marginal health cost is higher than that of marginal carbon intensity. (b) In 90 out of the 114 U.S. regions (79%), the normalized standard deviation of marginal health cost is higher than th… view at source ↗
Figure 9
Figure 9. Figure 9: Correlation analysis. (a) CDF of correlation coefficients between hourly health prices and marginal carbon emission rates for all the U.S. regions; (b) Scatter plot of health price and marginal carbon emission rate (annual average in 2023) across Meta’s U.S. data center locations. joint consideration of carbon-aware and health-informed GLB is interesting, we exclude such an analysis to better contrast the … view at source ↗
Figure 10
Figure 10. Figure 10: State-level electricity consumption of U.S. data centers in 2023 [5]. B Additional Results for Health-Informed GLB B.1 Details of the experiment setup We use Meta’s electricity consumption for each U.S. data center location in 2023 [37] for our experiments [PITH_FULL_IMAGE:figures/full_fig_p028_10.png] view at source ↗
read the original abstract

The surging demand for artificial intelligence (AI) has led to a rapid expansion of energy-intensive data centers, contributing to criteria air pollutant emissions and raising public health concerns that have received comparatively limited attention in sustainability assessments. This paper introduces a principled methodology to model air pollutant emissions for data centers and estimate the public health impacts. Our findings reveal that the growing demand for AI and computing technologies is projected to push the total annual public health burden of U.S. data centers up to more than $20 billion in 2028. Although national-level impacts remain modest, data center health costs are unevenly distributed: in the most affected counties, the estimated per-household health burden can reach about seven times the national average. Next, we propose a health-informed computing framework that explicitly incorporates public health impacts into data center resource management across space and time, mitigating public health costs while supporting environmental sustainability. More broadly, we recommend extended energy reporting to include public health impact of data centers and paying attention to all impacted communities.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The manuscript develops a methodology to estimate criteria air pollutant emissions from U.S. data centers driven by AI demand, translates these into monetized public health damages using location-specific grid mixes and exposure models, and projects an annual national burden exceeding $20 billion by 2028. It documents spatial heterogeneity (with some counties experiencing per-household burdens up to seven times the national average), proposes a health-informed computing framework for spatiotemporal resource allocation that internalizes these externalities, and recommends expanded energy reporting that includes public health metrics.

Significance. If the electricity-demand projections, emission factors, grid-mix assumptions, and health-damage valuations hold under scrutiny, the work supplies a concrete, spatially resolved quantification of an externality that has been largely absent from data-center sustainability assessments. The health-informed framework offers a practical mechanism for operators to trade off compute performance against public-health costs, and the call for extended reporting could influence both policy and industry standards.

minor comments (3)
  1. The abstract omits any mention of the underlying models, data sources, or validation steps; expanding it by one or two sentences to summarize the methodology would improve accessibility without altering length constraints.
  2. Figure captions and axis labels in the spatial-disaggregation maps (e.g., county-level burden choropleths) would benefit from explicit units and a note on the base year for the $20B projection.
  3. A short table listing the primary data sources (EIA, EPA emission factors, health-cost valuations) and their vintages would aid reproducibility.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive assessment of the manuscript, accurate summary of its contributions, and recommendation for minor revision. The work quantifies the public health externality of data center emissions under AI-driven demand growth and introduces a framework to internalize those costs in resource allocation.

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper's central projection ($20B health burden in 2028) is constructed from external inputs: electricity-demand forecasts, location-specific grid mixes, pollutant emission factors, exposure models, and monetized health damage valuations. These are described as drawn from independent data sources and standard modeling practices rather than fitted to the target output or defined in terms of the result itself. No equations reduce the prediction to its own inputs by construction, no load-bearing self-citations are invoked to justify uniqueness or ansatzes, and the health-informed framework is presented as an application of the independently derived impacts. The derivation chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available, so specific free parameters, axioms, and invented entities cannot be extracted. The work presumably relies on standard emission factors and health impact functions drawn from prior environmental science literature.

pith-pipeline@v0.9.0 · 5715 in / 1158 out tokens · 33158 ms · 2026-05-23T07:52:37.238599+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Representational Harms in LLM-Generated Narratives Against Global Majority Nationalities

    cs.CL 2026-04 unverdicted novelty 5.0

    LLMs generate narratives containing persistent stereotypes, erasure, and one-dimensional portrayals of Global Majority national identities, with minoritized groups overrepresented in subordinated roles by more than fi...

  2. What if AI systems weren't chatbots?

    cs.CY 2026-05 unverdicted novelty 3.0

    Chatbot AI systems often fail complex needs while projecting authority, contributing to deskilling, labor displacement, economic concentration, and high environmental costs, so alternative pluralistic and task-specifi...

Reference graph

Works this paper leans on

109 extracted references · 109 canonical work pages · cited by 2 Pith papers · 1 internal anchor

  1. [1]

    Centers for Disease Control and Prevention

    U.S. Centers for Disease Control and Prevention. Artificial intelligence and machine learning: Apply- ing advanced tools for public health.https://www.cdc.gov/surveillance/data-modernization/ technologies/ai-ml.html

    work page

  2. [2]

    Mary Tran. U.S. Department of State DipNote: New air quality dash- board uses AI to forecast pollution levels.https://www.state.gov/ new-air-quality-dashboard-uses-ai-to-forecast-pollution-levels/, May 2024

    work page 2024

  3. [3]

    How artificial intelligence and ma- chine learning can help healthcare systems respond to COVID-19.Machine Learning, 110:1–14, 2021

    Mihaela Van der Schaar, Ahmed M Alaa, Andres Floto, Alexander Gimson, Stefan Scholtes, Angela Wood, Eoin McKinney, Daniel Jarrett, Pietro Lio, and Ari Ercole. How artificial intelligence and ma- chine learning can help healthcare systems respond to COVID-19.Machine Learning, 110:1–14, 2021

    work page 2021

  4. [4]

    Smith, Alex Hubbard, Alex Newkirk, Nuoa Lei, Md Abu Bakar Siddik, Billie Holecek, Jonathan Koomey, Eric Masanet, and Dale Sartor

    Arman Shehabi, Sarah J. Smith, Alex Hubbard, Alex Newkirk, Nuoa Lei, Md Abu Bakar Siddik, Billie Holecek, Jonathan Koomey, Eric Masanet, and Dale Sartor. 2024 United States data center energy usage report.Lawrence Berkeley National Laboratory LBNL-2001637, December 2024

    work page 2024

  5. [5]

    EPRI. Powering intelligence: Analyzing artificial intelligence and data center energy consump- tion.White Paper on Technology Innovation Report, 2024.https://www.epri.com/research/products/ 3002028905

    work page 2024

  6. [6]

    Department of Energy

    U.S. Department of Energy. Recommendations on powering artificial intelligence and data center infrastructure, Jul. 2024

    work page 2024

  7. [7]

    Sustainable AI: Environmental implica- tions, challenges and opportunities

    Carole-Jean Wu, Ramya Raghavendra, Udit Gupta, Bilge Acun, Newsha Ardalani, Kiwan Maeng, Gloria Chang, Fiona Aga, Jinshi Huang, Charles Bai, et al. Sustainable AI: Environmental implica- tions, challenges and opportunities. InProceedings of Machine Learning and Systems, volume 4, pages 795–813, 2022. 19

    work page 2022

  8. [8]

    Lee, Gu-Yeon Wei, David Brooks, and Carole-Jean Wu

    Udit Gupta, Young Geun Kim, Sylvia Lee, Jordan Tse, Hsien-Hsin S. Lee, Gu-Yeon Wei, David Brooks, and Carole-Jean Wu. Chasing carbon: The elusive environmental footprint of computing.IEEE Micro, 42(4):37–47, jul 2022

    work page 2022

  9. [9]

    Islam, and Shaolei Ren

    Pengfei Li, Jianyi Yang, Mohammad A. Islam, and Shaolei Ren. Making AI less ’thirsty’.Commun. ACM, 68(7):54–61, June 2025

    work page 2025

  10. [10]

    Environmental report.https://sustainability.google/reports/ google-2024-environmental-report/, 2024

    Google. Environmental report.https://sustainability.google/reports/ google-2024-environmental-report/, 2024

    work page 2024

  11. [11]

    DCFlex initiative.https://msites.epri.com/dcflex, 2024

    EPRI. DCFlex initiative.https://msites.epri.com/dcflex, 2024

    work page 2024

  12. [12]

    Deepspeed-MOE: Advancing mixture-of-experts inference and training to power next-generation AI scale

    Samyam Rajbhandari, Conglong Li, Zhewei Yao, Minjia Zhang, Reza Yazdani Aminabadi, Am- mar Ahmad Awan, Jeff Rasley, and Yuxiong He. Deepspeed-MOE: Advancing mixture-of-experts inference and training to power next-generation AI scale. InICML, 2022

    work page 2022

  13. [13]

    Junkyard computing: Repurpos- ing discarded smartphones to minimize carbon

    Jennifer Switzer, Gabriel Marcano, Ryan Kastner, and Pat Pannuto. Junkyard computing: Repurpos- ing discarded smartphones to minimize carbon. InProceedings of the 28th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2, ASPLOS 2023, page 400–412, New York, NY, USA, 2023. Association for Computin...

    work page 2023

  14. [14]

    Jaylen Wang, Daniel S. Berger, Fiodar Kazhamiaka, Celine Irvene, Chaojie Zhang, Esha Choukse, Kali Frost, Rodrigo Fonseca, Brijesh Warrier, Chetan Bansal, Jonathan Stern, Ricardo Bianchini, and Akshitha Sriraman. Designing cloud servers for lower carbon. In2024 ACM/IEEE 51st Annual Inter- national Symposium on Computer Architecture (ISCA), pages 452–470, 2024

    work page 2024

  15. [15]

    Carbon-aware computing for datacenters.IEEE Transac- tions on Power Systems, 38(2):1270–1280, 2023

    Ana Radovanovi´c, Ross Koningstein, Ian Schneider, Bokan Chen, Alexandre Duarte, Binz Roy, Diyue Xiao, Maya Haridasan, Patrick Hung, Nick Care, Saurav Talukdar, Eric Mullen, Kendal Smith, MariEllen Cottman, and Walfredo Cirne. Carbon-aware computing for datacenters.IEEE Transac- tions on Power Systems, 38(2):1270–1280, 2023

    work page 2023

  16. [16]

    Reducing the carbon impact of generative AI inference (today and in 2035)

    Andrew A Chien, Liuzixuan Lin, Hai Nguyen, Varsha Rao, Tristan Sharma, and Rajini Wijayawar- dana. Reducing the carbon impact of generative AI inference (today and in 2035). InProceedings of the 2nd Workshop on Sustainable Computer Systems, HotCarbon ’23, New York, NY, USA, 2023. Association for Computing Machinery

    work page 2035

  17. [17]

    Hanafy, Qianlin Liang, Noman Bashir, Abel Souza, David Irwin, and Prashant Shenoy

    Walid A. Hanafy, Qianlin Liang, Noman Bashir, Abel Souza, David Irwin, and Prashant Shenoy. Going green for less green: Optimizing the cost of reducing cloud carbon emissions. InProceedings of the 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 3, ASPLOS ’24, page 479–496, New York, NY, US...

    work page 2024

  18. [18]

    2024 H1 semi-annual monitoring report (Intel Ocotillo facility).https://www.exploreintel

    Intel. 2024 H1 semi-annual monitoring report (Intel Ocotillo facility).https://www.exploreintel. com/ocotillo, 2024

    work page 2024

  19. [19]

    Interagency monitoring of protected visual environments

    UC Davis Air Quality Research Center. Interagency monitoring of protected visual environments. https://airquality.ucdavis.edu/improve

    work page

  20. [20]

    U.S. EPA. Research on health effects from air pollution.https://www.epa.gov/air-research/ research-health-effects-air-pollution

    work page

  21. [21]

    Long-term exposure to PM 2.5 and cognitive decline: A longitudinal population-based study.Journal of Alzheimer’s Disease, 80(2):591–599, 2021

    Giulia Grande, Jing Wu, Petter LS Ljungman, Massimo Stafoggia, Tom Bellander, and Debora Rizzuto. Long-term exposure to PM 2.5 and cognitive decline: A longitudinal population-based study.Journal of Alzheimer’s Disease, 80(2):591–599, 2021

    work page 2021

  22. [22]

    U.S. EPA. User’s manual for the co-benefits risk assessment (COBRA) screening model.https: //www.epa.gov/cobra/users-manual-co-benefits-risk-assessment-cobra-screening-model. 20

    work page

  23. [23]

    Estimates of global mortality burden associated with short-term exposure to fine particulate matter (PM2.5).The Lancet Planetary Health, 8(3):e146–e155, 2024

    Wenhua Yu, Rongbin Xu, Tingting Ye, Michael J Abramson, Lidia Morawska, Bin Jalaludin, Fay H Johnston, Sarah B Henderson, Luke D Knibbs, Geoffrey G Morgan, et al. Estimates of global mortality burden associated with short-term exposure to fine particulate matter (PM2.5).The Lancet Planetary Health, 8(3):e146–e155, 2024

    work page 2024

  24. [24]

    U.S. EPA. What is cross-state air pollution?https://www.epa.gov/Cross-State-Air-Pollution/ what-cross-state-air-pollution

    work page

  25. [25]

    Trivikrama Rao

    Jian Zhang and S. Trivikrama Rao. The role of vertical mixing in the temporal evolution of ground- level ozone concentrations.Journal of Applied Meteorology, 38(12):1674–1691, 1999

    work page 1999

  26. [26]

    Guidance for evaluating human health impacts in environmental assessment: Air quality.https://iaac-aeic.gc.ca/050/documents/p80054/119376E.pdf, 2016

    Health Canada. Guidance for evaluating human health impacts in environmental assessment: Air quality.https://iaac-aeic.gc.ca/050/documents/p80054/119376E.pdf, 2016

    work page 2016

  27. [27]

    World Health Organization. Air pollution is responsible for 6.7 million premature deaths every year.https://www.who.int/teams/environment-climate-change-and-health/ air-quality-and-health/health-impacts/types-of-pollutants

    work page

  28. [28]

    Ambient (outdoor) air pollution.https://www.who.int/news-room/ fact-sheets/detail/ambient-(outdoor)-air-quality-and-health, 2024

    World Health Organization. Ambient (outdoor) air pollution.https://www.who.int/news-room/ fact-sheets/detail/ambient-(outdoor)-air-quality-and-health, 2024

    work page 2024

  29. [29]

    Global burden of disease 2021: Find- ings from the GBD 2021 study.https://www.healthdata.org/research-analysis/library/ global-burden-disease-2021-findings-gbd-2021-study, May 2024

    Institute for Health Metrics and Evaluation (IHME). Global burden of disease 2021: Find- ings from the GBD 2021 study.https://www.healthdata.org/research-analysis/library/ global-burden-disease-2021-findings-gbd-2021-study, May 2024

    work page 2021

  30. [30]

    U.S. EPA. Human health and environmental impacts of the electric power sector.https://www.epa. gov/power-sector/human-health-environmental-impacts-electric-power-sector

    work page

  31. [31]

    U.S. EPA. Power plants and neighboring communities.https://www.epa.gov/power-sector/ power-plants-and-neighboring-communities

    work page

  32. [32]

    Mortality risk from united states coal electricity generation.Science, 382(6673):941–946, 2023

    Lucas Henneman, Christine Choirat, Irene Dedoussi, Francesca Dominici, Jessica Roberts, and Cor- win Zigler. Mortality risk from united states coal electricity generation.Science, 382(6673):941–946, 2023

    work page 2023

  33. [33]

    The costs to health and the environment from indus- trial air pollution in Europe — 2024 update.https://www.eea.europa.eu/publications/ the-cost-to-health-and-the, 2025

    European Environment Agency. The costs to health and the environment from indus- trial air pollution in Europe — 2024 update.https://www.eea.europa.eu/publications/ the-cost-to-health-and-the, 2025

    work page 2024

  34. [34]

    Diesel pollution from data centers.https://ecology.wa.gov/ air-climate/air-quality/data-centers

    Washington Department of Ecology. Diesel pollution from data centers.https://ecology.wa.gov/ air-climate/air-quality/data-centers

    work page

  35. [35]

    White House

    The U.S. White House. Executive order on advancing United States leadership in artificial intelli- gence infrastructure.https://www.federalregister.gov/documents/2025/01/17/2025-01395/ advancing-united-states-leadership-in-artificial-intelligence-infrastructure, Jan- uary 2025

    work page 2025

  36. [36]

    Model information.https://github.com/meta-llama/llama-models/blob/main/ models/llama3_1/MODEL_CARD.md

    Meta Llama. Model information.https://github.com/meta-llama/llama-models/blob/main/ models/llama3_1/MODEL_CARD.md

    work page

  37. [37]

    Sustainability report.https://sustainability.atmeta.com/ 2024-sustainability-report/, 2024

    Meta. Sustainability report.https://sustainability.atmeta.com/ 2024-sustainability-report/, 2024

    work page 2024

  38. [38]

    Investing in the rising data center economy.White Paper, January 2023

    McKinsey. Investing in the rising data center economy.White Paper, January 2023

    work page 2023

  39. [39]

    U.S. EPA. Co-benefits risk assessment health impacts screening and mapping tool (COBRA).https: //cobra.epa.gov/

    work page

  40. [40]

    National Center for Health Statistics

    U.S. National Center for Health Statistics. Factstats: Asthma.https://www.cdc.gov/nchs/fastats/ asthma.htm. 21

    work page

  41. [41]

    Vehicles registered by county.https://www.dmv.ca

    California DMV. Vehicles registered by county.https://www.dmv.ca. gov/portal/dmv-research-reports/research-development-data-dashboards/ vehicles-registered-by-county/

    work page

  42. [42]

    Report to the Governor and the General Assembly of Virginia: Data centers in Virginia (JLARC report 158), December 2024

    Virginia Joint Legislative Audit and Review Commission. Report to the Governor and the General Assembly of Virginia: Data centers in Virginia (JLARC report 158), December 2024

    work page 2024

  43. [43]

    Health risks from diesel emissions in the Quincy area.Air Quality Program Report (Publication 20-02-019), August 2020

    Washington Department of Ecology. Health risks from diesel emissions in the Quincy area.Air Quality Program Report (Publication 20-02-019), August 2020

    work page 2020

  44. [44]

    U.S. EPA. Use of backup generators to maintain the reliability of the electric grid.https://www.epa.gov/system/files/documents/2025-05/ rice-memo-on-duke-energy-regulatory-interpretation-04_17_25.pdf, May 2025

    work page 2025

  45. [45]

    Donahue, Allen L

    Neil M. Donahue, Allen L. Robinson, and Spyros N. Pandis. Atmospheric organic particulate matter: From smoke to secondary organic aerosol.Atmospheric Environment, 43(1):94–106, 2009

    work page 2009

  46. [46]

    U.S. EPA. Benefits and costs of the Clean Air Act 1990-2020, the sec- ond prospective study.https://www.epa.gov/clean-air-act-overview/ benefits-and-costs-clean-air-act-1990-2020-second-prospective-study, March 2011

    work page 1990

  47. [47]

    U.S. EPA. Summary of the Clean Air Act.https://www.epa.gov/laws-regulations/ summary-clean-air-act

    work page

  48. [48]

    U.S. EPA. National ambient air quality standards (NAAQS) table.https://www.epa.gov/ criteria-air-pollutants/naaqs-table

    work page

  49. [49]

    Laws and regulations.https://ww2.arb.ca.gov/resources/ documents/laws-and-regulations

    California Air Resources Board. Laws and regulations.https://ww2.arb.ca.gov/resources/ documents/laws-and-regulations

    work page

  50. [50]

    State of the air.Report, 2024

    American Lung Association. State of the air.Report, 2024

    work page 2024

  51. [51]

    Ambient air pollution attributable deaths.https://www.who.int/data/ gho/data/indicators/indicator-details/GHO/ambient-air-pollution-attributable-deaths, 2024

    World Health Organization. Ambient air pollution attributable deaths.https://www.who.int/data/ gho/data/indicators/indicator-details/GHO/ambient-air-pollution-attributable-deaths, 2024

    work page 2024

  52. [52]

    2024.Global status report on alcohol and health and treatment of substance use disorders

    World Health Organization. WHO global air quality guidelines.https://www.who.int/ publications/i/item/9789240034228

    work page arXiv

  53. [53]

    U.S. EPA. Projection of counties with monitors that would not meet in 2032.https://www.epa. gov/system/files/documents/2024-02/2024-pm-naaqs-final-2032-projections-map.pdf, February 2024

    work page 2032

  54. [54]

    Stevens and Supreme Court of The United States

    John P. Stevens and Supreme Court of The United States. U.S. reports: Massachusetts v. EPA, 549 U.S. 497.The Library of Congress, 2007

    work page 2007

  55. [55]

    U.S. EPA. Climate change and human health.https://www.epa.gov/climateimpacts/ climate-change-and-human-health

    work page

  56. [56]

    Standards and guidance.https://ghgprotocol.org/

    Greenhouse Gas Protocol. Standards and guidance.https://ghgprotocol.org/

    work page

  57. [57]

    U.S. EPA. Controlling air pollution from stationary engines.https://www.epa.gov/ stationary-engines

    work page

  58. [58]

    Fact sheet on emergency backup generators.https://www.aqmd

    California Air Resources Board. Fact sheet on emergency backup generators.https://www.aqmd. gov/home/permits/emergency-generators

    work page

  59. [59]

    Explaining the Uptime Institute’s tier classification system (April 2021 update).https://journal.uptimeinstitute.com/ explaining-uptime-institutes-tier-classification-system/

    Uptime Institute. Explaining the Uptime Institute’s tier classification system (April 2021 update).https://journal.uptimeinstitute.com/ explaining-uptime-institutes-tier-classification-system/. 22

    work page 2021

  60. [60]

    Issued air permits for data centers.https://www

    Virginia Department of Environmental Quality. Issued air permits for data centers.https://www. deq.virginia.gov/permits/air/issued-air-permits-for-data-centers

    work page

  61. [61]

    Hanna Pampaloni. Heat wave prompts increased data center generator use; turner pushes for tier 4 upgrades.https://www.loudounnow.com/news/ heat-wave-prompts-increased-data-center-generator-use-turner-pushes-for-tier-4-upgrades/ article_60a48bda-dc50-4d1a-8b16-399cd4340350.html, July 2025

    work page 2025

  62. [62]

    Data centers, diesel generators and air quality – PEC web map.https://www.pecva.org/uncategorized/ data-centers-diesel-generators-and-air-quality-pec-web-map/

    Piedmont Environmental Council. Data centers, diesel generators and air quality – PEC web map.https://www.pecva.org/uncategorized/ data-centers-diesel-generators-and-air-quality-pec-web-map/

    work page

  63. [63]

    Government of Canada. Wet cooling towers: Guide to reporting.https://www.canada.ca/en/ environment-climate-change/services/national-pollutant-release-inventory/report/ sector-specific-tools-calculate-emissions/wet-cooling-tower-particulate-guide.html

    work page

  64. [64]

    Wexler, Chris D

    Anthony S. Wexler, Chris D. Wallis, Patrick Chuang, and Mason Leandro. Assessing particulate emis- sions from power plant cooling towers.California Energy Commission Final Project Report (CEC-500- 2023-048), July 2013

    work page 2023

  65. [65]

    Reuters. Data center reliance on fossil fuels may delay clean-energy tran- sition.https://www.reuters.com/technology/artificial-intelligence/ how-ai-cloud-computing-may-delay-transition-clean-energy-2024-11-21/, November 2024

    work page 2024

  66. [66]

    Integrated resource plan.https://www.dominionenergy

    Virginia Electric and Power Company. Integrated resource plan.https://www.dominionenergy. com/-/media/pdfs/global/company/IRP/2024-IRP-w_o-Appendices.pdf, October 2024

    work page 2024

  67. [67]

    Energy Information Administration

    U.S. Energy Information Administration. Annual energy outlook 2025.https://www.eia.gov/ outlooks/aeo

    work page 2025

  68. [68]

    Electricity mix.Our World in Data, 2024

    Hannah Ritchie and Pablo Rosado. Electricity mix.Our World in Data, 2024

    work page 2024

  69. [69]

    New nuclear clean energy agreement with Kairos Power.https://blog.google/ outreach-initiatives/sustainability/google-kairos-power-nuclear-energy-agreement/, October 2024

    Google. New nuclear clean energy agreement with Kairos Power.https://blog.google/ outreach-initiatives/sustainability/google-kairos-power-nuclear-energy-agreement/, October 2024

    work page 2024

  70. [70]

    To land Meta’s massive$10 billion data center, Louisiana pulled out all the stops

    CNBC. To land Meta’s massive$10 billion data center, Louisiana pulled out all the stops. will it be worth it?https://www.cnbc.com/2025/06/25/ meta-massive-data-center-louisiana-cost-jobs-energy-use.html, June 2025

    work page 2025

  71. [71]

    The Associated Press. OpenAI shows off Stargate AI data center in Texas and plans 5 more elsewhere with Oracle, Softbank.https://apnews.com/article/ openai-stargate-oracle-data-center-0b3f4fa6e8d8141b4c143e3e7f41aba1, September 2025

    work page 2025

  72. [72]

    DynamoLLM: De- signing LLM inference clusters for performance and energy efficiency

    Jovan Stojkovic, Chaojie Zhang, Inigo Goiri, Josep Torrellas, and Esha Choukse. DynamoLLM: De- signing LLM inference clusters for performance and energy efficiency. InIEEE International Symposium on High-Performance Computer Architecture (HPCA), 2025

    work page 2025

  73. [73]

    E-waste challenges of generative artificial intelligence.Nature Computational Science, 4:818–823, October 2024

    Peng Wang, Ling-Yu Zhang, Asaf Tzachor, and Wei-Qiang Chen. E-waste challenges of generative artificial intelligence.Nature Computational Science, 4:818–823, October 2024

    work page 2024

  74. [74]

    U.S. EPA. Semiconductor industry.https://www.epa.gov/eps-partnership/ semiconductor-industry

    work page

  75. [75]

    Generative AI: The next S-curve for the semiconductor industry?White Paper, March 2024

    McKinsey. Generative AI: The next S-curve for the semiconductor industry?White Paper, March 2024

    work page 2024

  76. [76]

    U.S. EPA. Avoided emissions and generation tool (A VERT).https://www.epa.gov/avert

    work page

  77. [77]

    WattTime.https://watttime.org/. 23

    work page

  78. [78]

    Environmental sustainability report.https://www.microsoft.com/en-us/ corporate-responsibility/sustainability/report, 2024

    Microsoft. Environmental sustainability report.https://www.microsoft.com/en-us/ corporate-responsibility/sustainability/report, 2024

    work page 2024

  79. [79]

    Life-cycle emissions of AI hardware: A cradle-to-grave approach and generational trends, 2025

    Ian Schneider, Hui Xu, Stephan Benecke, David Patterson, Keguo Huang, Parthasarathy Ran- ganathan, and Cooper Elsworth. Life-cycle emissions of AI hardware: A cradle-to-grave approach and generational trends, 2025

    work page 2025

  80. [80]

    U.S. EPA. Air quality dispersion modeling.https://www.epa.gov/scram/ air-quality-dispersion-modeling

    work page

Showing first 80 references.