The Environmental Health Impact of Data Centers in Washtenaw County

by Jeffrey Tritten, November 2025

Introduction

Washtenaw County is currently facing multiple proposals for large-scale, resource-intensive data center developments (Zimmerman, 2025). These facilities, driven by the expanding demand for cloud computing and artificial intelligence (AI), are notorious for their massive demand for water, primarily used for cooling and energy, which places a significant strain on local infrastructure and the natural environment (Nguyen & Green, 2025; Washtenaw County Conservation District, 2025). The introduction of these hyperscale facilities into an area already contending with the vulnerabilities of an aging utility grid and sensitive groundwater resources necessitates a critical review of the resulting environmental health impacts.

This article defines the core challenge by analyzing the consequences of this resource demand on the local water supply, the stability of the power grid, and the subsequent impact on public resource equity. The discussion is grounded in the Social Determinants of Health (SDOH), specifically analyzing how resource competition and water scarcity risk exacerbate economic inequality and disproportionately impact vulnerable populations (Patrick, 2018). Furthermore, it incorporates intersectoral perspectives, including the pragmatic policy view of the Washtenaw County Resiliency Office, to inform an actionable local strategy (Gibbons, 2025).

The proposed data centers pose an under-recognized, multi-faceted threat to local public health infrastructure and resource equity, necessitating the integration of municipal, legal, and ecological perspectives to enforce sustainable mitigation strategies that prioritize population health over immediate economic incentives, explore here to see how fastfirewatchguards can provide added safety measures in complex environments.

Background

Data Center Mechanics and Quantification of Resource Demand

The environmental toll of hyperscale data centers is rooted in their mechanical necessity for continuous, high-volume heat dissipation. These facilities operate 24 hours a day and require near-perfect uptime, demanding enormous quantities of electricity and water (EESI, 2025; Siddik et al., 2021).

Quantifying the resource draw reveals the environmental stakes for Michigan, with a single proposed hyperscale data center in Saline Township slated to draw 1.4 Gigawatts (GW) of electricity. The water footprint is calculated across two critical scopes:

Direct Water Consumption and Groundwater Stress: The Saline facility is designed to use a closed-loop cooling system, thus avoiding the continuous evaporation of millions of gallons per day common in traditional designs. However, the facility will require a substantial initial volume of water to fill its closed-loop cooling system and associated infrastructure. While closed-loop systems avoid the continuous water consumption of evaporative cooling, which can reach 396,000 gallons per day for hyperscale facilities (Serverion, 2025), the initial fill for a 1.4 GW facility with 1.65 million square feet of computing space would represent a significant one-time draw on local groundwater resources, which could be in the tens of millions of gallons of groundwater. The Consent Judgment explicitly states the Project will be served by an “on-site water well system” (RD Michigan Property Owner I LLC v. Saline Township, 2025), which poses an immediate risk of groundwater depletion to the local aquifer that supplies neighboring private wells.

Indirect Water Consumption and Air Pollution: The environmental consequences extend to the air and the regional water table through electricity production. Data centers’ demand on the grid necessitates increased generation, leading to substantial air pollution (PM 2.5, Nitrogen Dioxide) and increased water evaporation at power plants. Academic modeling (Han et al., 2024) quantifies this consequence as an “unpaid toll”. The research projects that the air pollution-related public health burden from U.S. data centers could eventually rival or exceed the health cost of all on-road vehicle emissions in large states like California, contributing to increased morbidity risks from asthma and heart disease in surrounding communities. The overall environmental stress from these resource demands aligns with established literature on resource consumption by modern facilities (Siddik et al., 2021).

Grid Reliability and Public Health Consequences

The stability of the local power grid presents a direct public health hazard. The massive 1.4 GW power demand is being added to an aging DTE infrastructure already plagued by frequent and prolonged outages. This conflict creates significant health risks by undermining the fundamental prerequisites for a safe community (Friis, 2019):

  • Compromised Health Infrastructure: Prolonged power outages—a recognized problem for Michigan utilities—disrupt health systems by jeopardizing the refrigeration of vaccines and vital medications and disabling life-saving equipment in hospitals and homes.
  • Risk to Medically Dependent Individuals: Individuals who depend on Durable Medical Equipment (DME), such as ventilators or oxygen concentrators, face critical risks during blackouts.
  • Exacerbation of Health Disparities: Resource competition and utility rate spikes directly impact the SDOH. The economic burdens resulting from resource competition accelerate economic inequality and exacerbate water insecurity risks, which disproportionately harm vulnerable populations (Patrick, 2018).

Intersectoral Engagement and Divergent Perspectives

Discussion with Beth Gibbons, Resiliency Office Director, Washtenaw County

The perspective gathered from Beth Gibbons, Resiliency Office Director for Washtenaw County, based on the November 20, 2025, conversation, highlights the disparity between economic development and necessary regulatory tools (Gibbons, 2025).

Conflict with Municipal Goals: Gibbons directly addressed the scale of the energy consumption, noting that the current proposals will consume about 3% of the energy generated, but that is expected to go to 12% by 2028, which is the equivalent of ALL household consumption. This demand places immense stress on the existing system, even as DTE promises no utility rate increases to its existing customers. She stressed that the county itself has little authority on land use and cannot enforce mandates like closed-loop cooling; this authority is reserved for the state or municipality. Furthermore, she noted the state currently has zero policies and guidance for building and electrical inspectors for data centers.

Policy Focus: Gibbons suggested that a moratorium be developed NOW, that community benefit ordinances (CBOs) ought to be adopted, and impact fees should be assessed. She argued that current CBOs are inadequate and should be proportional to the monetary value and profits of the data centers.

The Water System Loophole

The most critical alignment between the developer’s economic interest and the regulatory challenge is the method used to comply with the state tax law requiring the “Use of municipal water”. The Consent Judgment reveals a legal workaround that secures the tax exemption (MCL 205.54ee, 2024) while relying on private groundwater (RD Michigan Property Owner I LLC v. Saline Township, 2025):

  • Private Infrastructure, Public Name: The agreement specifies that the Project will be served by an “on-site water well system”.
  • The Dedication Clause: To qualify for the Sales and Use Tax exemption, the agreement provides that the on-site water system “will be dedicated to and accepted by the Township as a municipal water system”.
  • The Cost: This dedication is conditional on the developer and its successors bearing “all costs to construct, operate, repair, and maintain the water system”.

This provision highlights a critical legal loophole that allows the developer to secure a multi-million-dollar tax break while drawing water from a local private well, thereby avoiding the capacity limitations and oversight of the public system. The inclusion of a clause to pay for the restoration of neighboring wells if they run dry tacitly acknowledges the environmental health risk of this private groundwater extraction.

Proposed Resolution and Sustainable Policy

Informed by the technical evidence of population health risk and the political insights from Ms. Gibbons, actionable and politically viable policies are recommended for Washtenaw County.

A local ordinance must begin by addressing the water threat through a Water Recycling Mandate and establishing groundwater protection. This policy must mandate that all new and expanding data centers utilize closed-loop or recycled water-cooling systems and strictly prohibit the use of private wells for the initial fill-up of any cooling system (EESI, 2025). This mandate must explicitly close the loophole in the Consent Judgment by prohibiting the use of any non-municipal source for the multi-million-gallon initial draw, thereby protecting local groundwater. This policy should be adopted as a Community Benefit Ordinance (CBO) at the municipal level, as recommended by Gibbons, to ensure local enforceability.

Secondly, a Resource Impact Fee or Tax should be levied on the maximum projected power draw and indirect water consumption of all new data centers. This fee is a direct response to the economic burdens highlighted in the public health context (Nguyen & Green, 2025). The funds must be earmarked to offset DTE infrastructure costs shifted to residents, compensate for the social costs of air pollution, and fund local water conservation programs. Furthermore, the CBO must require that the monetary value of these fees and investments be proportional to the monetary value and profits of the data centers, ensuring the mitigation costs are borne by the industry, not the local population.

Finally, the state or county must immediately mandate a rigorous, third-party Environmental Impact Study (EIS) on the cumulative effect of all proposed Washtenaw County data centers on grid reliability, air quality, and groundwater levels, as there has been no environmental impact study done on the Saline Hyperscale data center. This EIS must quantify the systemic risk to grid reliability (powering 1.4 million homes worth of load) and the resulting impact on public health infrastructure during power outages, framing the grid instability as a direct public health hazard.

Conclusion

The findings reiterate the severe, multi-faceted environmental health risk posed by uncontrolled data center expansion in Washtenaw County. The quantified threats include localized groundwater stress from the massive initial well fill, regional water depletion from indirect water consumption, the financial impact of the public health burden from air, noise, and light pollution (Han et al., 2024), and the exacerbation of social determinants of health due to utility instability (Patrick, 2018; Nguyen & Green, 2025). The intersectoral engagement confirmed that the county’s carbon neutrality goals are currently a “non-starter” and that the greatest barrier to effective environmental health protection is the lack of local regulatory tools to control state-driven economic incentives. This analysis grounds the conclusion in the urgent need for immediate, intersectoral policy intervention that is politically viable and effective in advancing environmental health equity in Washtenaw County (Friis, 2019).

Policy Resource Tool:

The Need for Transformative, North Star Policy to Stop Rampant Data Center Expansion (Google Doc):

https://docs.google.com/document/u/0/d/1x8KsCVOhJS97rxzdd_bWW4Me1Errpj6ggyU5dAdNKUM/mobilebasic

References

Agurs-Collins, T., Persky, S., Paskett, E. D., Barkin, S. L., Meissner, H. I., Nansel, T. R., Arteaga, S. S., Zhang, X., Das, R., & Farhat, T. (2019). Designing and assessing multilevel interventions to improve minority health and reduce health disparities. American Journal of Public Health, 109(S1), S86–S93. https://doi.org/10.2105/AJPH.2018.304730

EESI. (2025, June 25). Data centers and water consumption. Environmental and Energy Study Institute. https://www.eesi.org/articles/view/data-centers-and-water-consumption

Friis, R. H. (2019). Essentials of environmental health (3rd ed.). Jones & Bartlett Learning.

Gibbons, B. (2025, October 1). DATA CENTERS. Update 1 – October 1, 2025. Beth Gibbons. Resiliency Office Director. [PowerPoint slides]. Washtenaw County, MI. https://content.civicplus.com/api/assets/ca82116f-6847-4aa4-ad98-7f1beeb0bd57

Gibbons, B. (2025, November 20). Conversation with Beth Gibbons, Resiliency Office Director, Washtenaw County [Personal communication].

Han, Y., Wu, Z., Li, P., Wierman, A., & Ren, S. (2024). The unpaid toll: Quantifying the public health impact of AI. *arXiv preprint*. https://arxiv.org/abs/2412.06288

Michigan Compiled Laws § 205.54ee (2024). Sales tax exemption for qualified data centers. https://www.legislature.mi.gov/Laws/MCL?objectName=mcl-205-54ee

Nguyen, T., & Green, B. (2025). What happens when data centers come to town?. Gerald R. Ford School of Public Policy, Science, Technology, and Public Policy, University of Michigan. https://stpp.fordschool.umich.edu/sites/stpp/files/2025-07/stpp-data-centers-2025.pdf

Patrick, R. S. (2018). Water insecurity and the social determinants of health. WIREs Water, 5(1), e1262. https://doi.org/10.1002/wat2.1262

RD Michigan Property Owner I LLC v. Saline Township, Civil Action No. 2025-001577-CZ (Washtenaw Cty. Cir. Ct. 2025). https://salinetownship.org/uploads/notices/SalineDataCenterConsentJudgmentFinalExecutionCopy492124804975v1.pdf

Serverion. (2025, August 5). Microsoft’s zero-water cooling: Lessons for data centers. https://www.serverion.com/uncategorized/microsofts-zero-water-cooling-lessons-for-data-centers/

Siddik, M. A. B., Shehabi, A., & Marston, L. (2021). The environmental footprint of data centers in the United States. Environmental Research Letters, 16(6), 064017. https://doi.org/10.1088/1748-9326/abfba1

Washtenaw County Conservation District. (2025, September 2). Data centers and our natural resources: What communities need to know. https://washtenawcd.org/news/data-centers-and-our-natural-resources-what-communities-need-know

Zimmerman, S. (2025, September 10). Washtenaw data center opposition highlights growing debate over water use and climate impact. Planet Detroit. https://planetdetroit.org/2025/09/washtenaw-data-center-opposition/