Data Centers – Coming to a Town Near You

Across the nation communities are facing the prospect of new data centers to power the artificial intelligence (AI) boom. These are not the smaller scale data centers that we have known since the 1990s to power the internet, but rather large-scale “hyper” facilities that can sprawl over as much as 2,000 acres, run 24 hours a day, and consume enormous amounts of both energy and water.  Hyperscale data centers can reach well over a million square feet, with Google’s first hyperscale data center encompassing 1.3 million square feet.

Currently, 90% of existing data centers use less than 50 megawatts. The new generation of AI data centers include proposals for massive complexes that would demand 1,000 to 4,000 megawatts (or one to four gigawatts) of electricity; and often use from one to five million gallons of water per day for cooling purposes. The buildout of these speculative data centers can leave local communities with noise and air pollution, a strain on electrical grid capacity and water resources, significant climate and health impacts, and increases in utility bills.  Although promising economic benefits for local communities, this rapid expansion is often happening with limited public information about the long-term impacts new data centers could bring to an area.

The surge in artificial intelligence and cloud computing has led to massive industrial data center growth. As of January of this year, there are over 3,900 operational data centers in the U.S. alone, nearly 37% of the world’s total. And, as of April 2026, data centers in current operation and planned, can be found in nearly every state in the lower 48, in both rural and urban areas.

Impacts on Communities from Data Centers

With this surge in the development of data centers, communities need to be aware of the potential impacts in advance of their development. New data centers can deliver on the promise of some possible investments for communities and provide a temporary surge in jobs. However, unless there is an intentional and managed approach to the development of these data centers, the risk and the consequences will rest solely on the shoulders of the local community.

Below you will find a list of impacts that communities are already experiencing. Use this as guidance when engaging in a public forum with your community about data centers. These issues need to be addressed in advance of the approval of any data center project.

1. Impacts on water resources.

Recent estimates project that by 2028, individual AI-related data centers in the U.S. could require up to 32 billion gallons of water annually, enough to support roughly 360,000 households’ indoor water use. This massive demand is a pressing problem for communities already facing drought or depleting water supplies. Two-thirds of all data centers built or in development since 2022 are located in water-stressed areas such as southern Arizona, the Colorado River Basin and Texas. Fifty-seven percent of data centers use potable water for cooling, the same source of water used by residential and municipal systems. Evaporative cooling fully removes that water from the local watershed without returning it even as treated effluent, further depleting an already stressed water resource.

Consider: Some data centers are using cooling approaches that reduce water demand, such as traditional air cooling or liquid immersion. Others are leveraging reclaimed wastewater from utilities instead of drawing from local supplies.

2. Impacts on and from energy resources.

World Resource Institute reports that with the need for a constant and reliable power supply to keep data centers running 24/7, many facilities rely on gas-fired generation for routine operations alongside diesel generators for emergency back-up. Both sources of energy produce air pollution, emit greenhouse gas emissions and pose health risks to local communities.

Consider: Some facilities are beginning to adopt cleaner on-site power alternatives. Battery energy storage systems can replace diesel generators during outages, avoiding their high pollution. Pairing hybrid microgrids with renewables is another alternative that supports community resilience without increasing emissions. Local and state officials can help by integrating clean power and emissions considerations into permitting and comprehensive planning, requiring companies to disclose on-site power strategies; and evaluating cleaner alternatives during permitting review.

3. Short term and long-term noise impacts.

Short term noise impacts are experienced during the construction process, which will vary in duration depending on the size and speed of construction of the data center facility. However, once completed, data centers run on a 24/7 schedule and can be a constant source of noise disturbance. In particular, roof top cooling systems and backup generators (both external units) can be disruptively noisy. Smaller diesel generators reportedly reach 85 decibels, levels that can harm hearing. Larger industrial units can approach 100 decibels — equivalent to a motorcycle or jackhammer. Experiences have been reported from nearby residents on impacts such as headaches, disruptions of sleep and lowered quality of life, which were echoed during a recent congressional roundtable.

Consider: To address this, some data center developers and local governments are beginning to use measures such as natural acoustic barriers, greater setbacks from homes and improved insulation for backup generators. Emerging technologies, including liquid cooling and improved airflow design, can also lower fan and HVAC noise. In places where zoning codes include noise protections, local governments can use the permitting processes to require limits on projected decibel levels, noise mitigation plans and ongoing monitoring commitments from the data center facility.

4. Reshaping and reprioritizing land use.

Dedicating land to industrial uses such as data centers may conflict with agricultural and rural land use priorities. This can lead to the loss of productive farmland, which is already declining due to development pressures, land consolidation and competition. Massive data centers are reshaping how land is used in many communities. In 2024, the average data center sitecovered about 224 acres or 0.35 square miles — roughly the size of 450 football fields. This represents a 144% increase in the typical size of a data center since 2022. For urban and suburban communities, the question is whether land should be prioritized to meet the basic needs of a community such as providing additional workforce housing opportunities; and/ or prioritized to meet quality of life factors for the community such as green space, community gardens, recreational space or even habitat restoration for woodlands, prairies or other representative habitats of the region. These are all important and on-going land use and zoning questions for any community or county, which become even more highlighted with the looming pressure of large-scale data centers.

Consider: For the short term, local governments are taking the temporary measure of putting moratoria in place while they better understand both the impacts from data centers and the path forward for possible zoning or permitting restrictions on data centers. Some local governments are considering policies to treat data centers as a distinct land use, applying additional review processes, and in rural areas, steering them away from prime farmland. This might include strengthened agricultural zoning and implementing performance standards and monitoring to address buffering, noise, traffic and long-term management of a site.

5. Pressure on marginalized communities.

Emerging analyses suggest that some data center siting patterns could compound long-standing environmental and economic challenges in already marginalized communities. A national review of roughly 700 data centers across the country found that nearly half of these data centers are located in census tracts with above-median environmental burdens, such as air pollution, park access issues and water pollution, as measured by the Center for Disease Control’s Environmental Justice Index. Many data centers are located in areas with social vulnerability indicators, such as poverty and lower education levels.

Consider: More research is needed to fully understand these trends. Yet preliminary findings underscore the importance of ensuring that new data center development does not reinforce existing inequities. Local governments and communities need to navigate the trade-offs of data center siting to ensure transparency and accountability in order to protect both human health and equity.

6. Economic trade-offs that favor local communities.

Data center development can create real economic opportunities for local communities, although most data center jobs are temporary and generated during the construction phase. Jobs pertaining to actually running data centers are limited in terms of actual number of employees. The other side of that narrative is that data centers generate revenue for local and state government through property, sales and use taxes. However, many states and municipalities are also offering sizeable tax incentives to attract data centers, offsetting the gains in revenues. In addition, with the building of large data centers, there may be investments borne by a community in water systems, electrical substations, transmission lines, roads and other infrastructure requirements. The recovery on these investments may outlive the life of a data center, and if so, will the municipality bear the burden over the long term? Who may be left paying for and maintaining an infrastructure that is no longer needed? The question then becomes how to balance these economic tradeoffs and potential liabilities both over the short and long term.

Consider: The burden rests with local governments to develop more safeguards and tools to be used in their negotiations with developers. To begin with, more local governments are imposing temporary moratoria while they assess the potential tradeoffs and develop siting criteria, procedural protections, and community safeguards. Community priorities, along with those of the government and developers have to be carefully weighed if a community is going to benefit rather than bear all the risks of a data center.

7. Unaccounted for concentrated and/or polluted discharge from data centers.

Data center cooling generates two distinctive discharge streams: concentrated blowdown from evaporative cooling (containing biocides, corrosion inhibitors, heavy metals, and potentially PFAs); and thermal effluent from free-cooling configurations – each with different regulatory treatment requirements. In particular, PFAs is data center coolants is an emerging and unregulated issue. Some nonwatery liquid coolants and F-gas refrigerants used in air cooling are PFAS compounds. Current pretreatment program monitoring requirements do not broadly address this chemistry.

Consider:  Local governments and communities must demand transparency. Nondisclosure agreements and voluntary-only reporting must be eliminated. Municipal utilities need to require mandatory data center facility-level water discharge reporting so municipalities can manage cumulative impacts.

 

World Resources Institute’s Summary of Governance Tools to Mitigate Data Center Impacts on Communities

Impact Area	Governance Tool
Energy Use	Dedicated rate classes for large energy users Cost recovery safeguards
Water Protection	Water use monitoring and public disclosure< Drought contingency planning Reclaimed water use
Air Quality	Replacing diesel with battery storage Emissions review during permitting On-site clean power requirements
Land Use Planning	Agricultural zoning protections Performance standards noise, traffic and buffering
Equity and Community Engagement	Community benefits agreements Transparent siting requirements Limiting use of NDAs Public disclosure of resource demands

 

The Process for Developing a Data Center (and possible points of intervention)

Before a data center is built, a developer needs upfront capital – data centers cost around $10 million dollars per megawatt, meaning a 40 megawatt center will cost at least $400 million to build. There are five stages to the development of a data center:

1. Site selection. This stage entails finding a site that fits the basic criteria: cloud connectivity and network availability; access to stable electric grid and abundant power supply; proximity to the industry it will serve and/or population centers; and relatively predictable weather. This phase can take up to a year, particularly in dense urban areas.
2. Permitting. This next stage is getting permits and local approvals, as well as apply for any incentives a state may have. Typically, permits include zoning (local), building (local), energy (state), and environmental (state), and depending on municipal and state requirements, this stage can last up to a year and a half, or longer. If an environmental impact report or environmental assessment is required, public review and comment opportunities provide a n appropriate possible point of intervention.
3. Design and pre-development. The next stage includes the architectural design of the data center, as well as pre-development work such as bringing in high voltage power lines, building new substations, and connecting to water lines. Depending on any challenges that occur, this stage could take up to a year and a half.
4. Construction. This Stage – which can take up to three years – includes building the actual core and shell of the building and installing the required components (uninterruptible power supply, Graphics Processing Units (GPUs), generators, air conditioners, servers, etc.). These components represent the most costly piece, with top of the line GPUs costing over $25,000 for a single GPU – a data center may have hundreds or even thousands of GPUs.  Once construction is complete, energy must be procured.
5. Procurement. The final phase – energy procurement – involves submitting a request for power within a given location’s electricity provider through power purchase agreements or grid interconnections and could take up to a year, depending on the length of the application process. Another option is for supplying their own power behind the meter through virtual power plants, In cases where there is existing high demand on the grid, a grid operator may place a moratorium on large load interconnection. This stage is often done concurrently or towards the end of the construction process.

 

Taking Action: A Few of the Early-stage Promising Stories

Communities, local governments and states are taking the issue of the uncertainties of data centers into their own hands by putting pressure on developers, pushing the pause button on moving forward through the use of tools such as moratoria, introducing bills and land use guidance or permitting requirements to address specific impacts on local communities. Here are some examples of the early success stories to date on taking a more precautionary approach to the development of data centers:

• A Public First poll released in February found that 26 percent of Americans support building more data centers, the lowest approval of any nation surveyed.
• According to The Financial Times, dozens of projects worth at least $156 billion have been blocked or stalled since 2025 due to local opposition.
• Monterey Park, CA voted 86 percent to 14 percent to permanently ban data centers from the city. It is the first US city to do it through a ballot initiative. Campaign organizer Steven Kung called it “a landslide victory.” On the reasons: “The noise pollution, the air pollution, the rise in the electricity rates. The deal just didn’t make sense and it doesn’t make sense for most, if not all, cities data centers go to.”
• Last year, utility AEP Ohiointroduced a new rate schedule for data centers requiring them to pay for at least 85% of the energy they are subscribed for, regardless of how much they actually use. Other states, like Oregon, Minnesota and Missouri, have required utilities to create new billing classes and rate structures for large energy users so that their costs aren’t shared by households and businesses.
• In Ohio, a proposal to rezone 300 acres of farmland to support data center development sparked such widespread opposition that the landowners ultimately withdrew their application.
• The San Diego Chapter of the Sierra Club filed a lawsuit in May 2026 challenging the Imperial County, CA Board of Supervisors’ April decision to allow several parcels of land to be joined together to build a data center complex.
• The Tonawanda Seneca Nation and Sierra Club Niagara Group successfully filed a lawsuit to stop the development of a huge data center in Genesee County. Today, the Niagara Group is fighting plans for a much larger data center on 30,000 pristine acres of natural land supporting the Tonawanda Seneca Nation, the Iroquois National Wildlife Refuge, and several state wildlife areas which contain clean air, abundant water, and uniquely connected wildlife habitat.
• The Southern Environmental Law Center and Sierra Club launched a lawsuit In February 2026 to challenge Georgia regulators over an air permit for new fossil fuel infrastructure tied to data center development.
• Sierra Club, MPSC, Michigan Environmental Council, NRDC (Natural Resources Defense Council), and Citizens Utility Board of Michigan filed a petition objecting to a fast-track approval for DTE Electric’s contracts to serve a massive data center in Saline Township.
• The NAACP and Earthjustice are suing Elon Musk’s company xAI over air pollution from an unpermitted power plant built for an AI data center.
• Earthjustice stopped a congressional bill that would have exempted data centers for key environmental protections.
• Maryland recently passed a law that requires the Public Service Commission to create a new rate specifically for data centers to ensure that everyday people don’t foot the bill.
• Governor Murphy of New Jersey signed into lawA5466, a bill which requires the Board of Public Utilities to study the effects of data centers on electricity costs. This bill had strong bipartisan support and is an important step towards understanding the effects data centers have on ratepayer’s bills.
• Dozens of other states, counties and cities nationwide have implemented temporarymoratoriums—pauses on new construction or permitting—as they evaluate the infrastructure’s resource demands, including:

 

STATES THAT HAVE TAKEN ACTION

(reported by the National Conference of State Legislatures)

Fourteen states are now considering bans on data centers, most often driven by energy and water considerations, as well as opposition from local communities. This is a dynamic situation, with more states joining in. Below you will see an indication of the approach different states are taking.

• New York:The state legislature passed the Responsible Data Center Development Act, imposing a one-year moratorium on the permitting of large data centers (over 20 megawatts) while the state completes environmental impact studies
• Oklahoma:Enacted an immediate moratorium on all new data center construction until November 2029 (Senate Bill 1488)
• Vermont:Paused the permitting of AI data facilities (above 10 megawatts) until July 2030 (Senate Bill 205)
• Maryland, Georgia, & Michigan:All introduced statewide moratorium legislation imposing varying restrictions or temporary pauses on data center permits while oversight frameworks are developed
• Maine:Lawmakers passed a statewide moratorium, though it was ultimately vetoed by the Governor
• Delaware: Moratorium on permits for data centersover 100 MW until January 2027
• Minnesota: Bans new data center permits until the Public Utility Commission submits a comprehensive report, until at least July 2027
• New Hampshire: One year moratorium on data center construction
• Pennsylvania: Three-year moratorium on new data centers paired with impact studies; or empower municipalities to prohibit data centers
• South Carolina: Forbids local approval of data centers until the Legislature adopts a “comprehensive oversight” framework
• South Dakota: One year moratorium on construction or expansion of data centers
• Vermont: Moratorium on data centers until 2030, requires impact study
• Virginia: Prohibits final approval of data centers until all interconnection requests are fulfilled, or until July 2028
• Wisconsin: Prohibits data center operation until 14 conditions are satisfied

 

MORE THAN 500 COUNTIES AND MUNICIPALITIES HAVE TAKEN ACTION

(reported by The Guardian)

Today, three-quarters of Americans oppose local datacenter development, a swing of more than 30 points in just a year. More than 500 counties and municipalities have passed datacenter bans or moratoria, while dozens of proposed projects have been killed outright. And the issue has come to define some of the highest-profile political races in the country.

 

The Positioning on Data Centers Coming Out of the U.S. Federal Government

Federal data center legislation primarily focuses on accelerating infrastructure permitting for data centers, mitigating massive energy and water impacts, while addressing localized pushback. Congress is currently evaluating several major proposals including:

Key Federal Legislative Proposals:

The DATA Act of 2026 (S. 3673): Introduced by Sen. Tom Cotton, this legislation seeks to exempt fully off-grid, privately financed data center power suppliers from Federal Power Act and Department of Energy regulations, allowing facilities to bypass long electrical grid interconnection queues.

Data Center Transparency Act (H.R. 6984): This bill directs the EPA to closely track and update data every three months on the cumulative environmental impacts of data centers, specifically targeting water consumption, utility costs, and greenhouse gas or air pollution.

Protecting Families from AI Data Center Energy Costs Act (H.R. 6529): Targets grid reliability and ratepayer protection to ensure that massive corporate data center loads do not drive up localized residential electricity bills.

Local Control Protection Act (Introduced June 2026): Aimed at halting out-of-state tech corporations from overriding municipal zoning denials through federal litigation, and requires enforced community benefit agreements for federally incentivized builds.

Federal Executive Action and Guidelines:

Executive Order 14318: Signed by the Trump Administration, this EO focuses on streamlining and accelerating the permitting of data center infrastructure, transmission lines, and energy supply chains on appropriate federal lands to facilitate AI infrastructure.

Agency Guidance & Standards: The Department of Energy and the EPA continue to issue guidance on optimizing energy use and streamlining Clean Air Act resources for developers. Additionally, many developers have signed a voluntary Ratepayer Protection Pledge with the White House to cover the costs of new electric generation resources needed for their infrastructure.

 

TOOLBOX RESOURCES: Getting Started in Your Own Community

Expansion of data centers is a given, however, the way it happens is not. Today’s expansion is unprecedented, unfolding at a much larger scale and pace than ever before. This sense of urgency is fueled by billions of dollars in private investment, state tax incentives,  and federal directives aimed at fast-tracking permitting.

Communities must demand of decision makers clear (regulatory) guidelines on land use and siting of data centers (including consideration of noise factors). Developers must declare their energy procurement and energy generation sources (including integration of renewable energy sources), water sources and water management plan. Before any permitting is secured, a cost recovery plan for local communities must be communicated and committed to by the developer and owner of the operation. How these issues are approached are all determining factors in whether a data center contributes to local communities, strengthens their economies and infrastructure – or becomes a burden. Here are some links to resources that might be helpful to communities facing the uncertainty of data center development in their area:

• The Sierra Club’s Local Guidelines for Data Centers provides a range of local policies needed to address land use, energy, carbon emissions and water impacts.
• Sierra Club’s Unconstrained Demand provides an analysis of the unprecedented growth of AI centers in Virginia, the epicenter of unchecked rapid expansion in a state that lacks any policy protections. Excellent situation analysis and policy recommendations and guidelines.
• Sierra Club’s Recommendations for Data Centers identifies policies that are not meant to stem the growth of data centers or any other specific end-use, but rather to ensure that data centers do no harm to communities, the environment or social equity outcomes.
• A new Sierra Club report,Data Center State Policies, 2026, contains recommendations for state utility regulators and policymakers dealing with data centers. These guidelines are intended to protect customer electricity rates and public health, mandate transparency, and prioritize on-site low-carbon energy use. Check the webpage Demanding Better for more information on this issue.
• TheNAACP’s Frontline Framework Community Guiding Principles offers practical strategies to ensure data centers are planned and built in ways that bring together community voices, environmental and climate justice framing.
• NAACP’s Community Benefits Agreement template helps communities and decision-makers advance transparency, accountability and shared governance in data center development.
• World Resources Institute’s Data Centers and Rising Energy Demand in the US looks to meet the challenge of the rapid expansion of data centers in the US by addressing three interconnected issues: meeting demand sustainably, ensuring state and federal policies enable smart and affordable growth, and delivering benefits while avoiding harm in local communities.
• North Star Data Center Policy Toolkit maps out the possible local policy interventions to stop, slow, and restrict rampant data center development at the town, city, and county level.
• Regulating Data Center Water Use in California evaluates how to better manage water impacts on local communities and the environment; and includes current knowledge on data center water use, a mapped approach to the policy and regulatory framework for direct data center water use in California, and provides recommendations.
• Next 10’s The Intersection of Data Center Development, Water Availability, and Environmental Justice in California asks a question that doesn’t get posed enough:  What could be done right now to make the planning and development of data centers in California both more sustainable and more equitable?
• Tennessee Lookout’s report on Here’s what communities can do when data centers arrive is a community toolkit developed through interviews with 50 experts on how ordinary people can respond and prepare to handle data center projects beforehand.
• Kentucky Resource Council’s Data Center Toolkit and Frequently Asked Questions answer common questions about data centers and their impacts. This resource, drafted in partnership with the Kentucky Sierra Club, is meant to help Kentuckians understand what hyperscale data centers are, what impacts they have, and what local communities and concerned citizens can do to mitigate the worst impacts.
• Hyperscale Data Centers in Wisconsin: BIG TECH UNCHECKED, a toolkit for community action.
• Alliance for the Great Lakes’ A Community Checklist for Evaluating Data Center Impacts. Asking informed questions early in the process can help protect local water resources, public health, and overall community well-being. This checklist is intended for early engagement with local officials, planning boards, and developers.
• Hubbell Environmental Law Initiative’s Iowa Data Center Decision Tool is an objective risk management resource to help communities prepare for data center discussions by identifying key questions before decisions are made.

 

Marine-Based Data Centers – Looking Forward to a Big Unknown, and Yet to be Proven Option

The next frontier for siting and experimenting with locations for data centers has arrived. Experimental data centers in the marine environment range from submerged capsules to floating platforms with the hope that sustainable solutions may be found for the massive energy, water and cooling demands of modern data centers.

In particular, interest in the marine environment is based on the notion that utilizing ocean water, considered a renewable resource, for cooling (cold seawater absorbs heat from servers) will reduce energy consumption and eliminate the need for freshwater usage. The second part of the equation is that floating platforms (on the surface of the ocean) could be wave or current powered, harnessing marine renewable wave-powered energy directly from the motion of the surface of the sea. Both submerged capsule and floating platform data centers are very much in the preliminary stages of exploration and actual implementation.

The Challenges of Marine-based Data Centers

Although marine-based data centers are emerging as a sustainable solution to the enormous energy, water and cooling demands of modern AI support systems, they will be facing some unique challenges by moving from land-based to marine systems.

The Potential Environmental Impacts of Marine-based Data Centers

Although some experimentation has taken place, a complete understanding of the potential environmental impacts to the marine environment can only be extrapolated from other types of human use impacts. Areas of concern include:

• Heat emissions. There is uncertainty about concentrated thermal emissions from massive data centers altering localized water temperatures and disrupting thermoclines, in combination with the uncertainty of marine heat waves.
• De-oxygenation. High temperature discharge water from a data center could lead to marine water de-oxygenation, inhibiting continued and sustained oxygenation and impacting different marine form.
• Acoustic disturbance. Persistent low frequency noise from cooling pumps and other mechanical equipment can interfere with navigation, communication and life history patterns of marine mammals and sea turtles.
• Electromagnetic fields. Transmission and power cables extending from shoreline to offshore data centers generate electromagnetic fields that interfere with the echolocation navigation and communication of marine mammals and other marine species.
• Seabed and habitat destruction. Both submerged capsule and floating platform data centers result in habitat and species disruption and destruction (particularly to the benthic habitat) during the installation and maintenance stages.
• Anti-fouling chemical contamination. Toxic anti-fouling coatings are frequently used in the marine environment to combat marine organisms such as barnacles and algae attaching to submerged infrastructure, which in turn can leach heavy metals into the water column.

Although there remain concerns and viability questions about underwater data centers, it does force the conversation about evolving the technology in a manner that reduces energy use and improves cooling efficiency. It also raises the question about long-term ecological effects and the impacts of data centers on both land-based and marine ecosystems. So far, there seems to be little discourse on the nexus between large-scale data centers and the 30 X 30 initiative – the global UN initiative to conserve 30% of the planet’s representative habitats on land, the ocean and inland waters by 2030 in order to halt biodiversity loss, mitigate the impacts of climate change and build ecological resilience.

 

Where do we go from here?

At this juncture, we must ask the question – do we really need to build massive data centers disrupting life in rural and urban American communities alike? Does it seem outlandish to construct floating data centers in the ocean or sending them into space? In an interview with UC Berkeley News, Professors Carl Boettiger and Fernando Pérez, both leaders at the Eric and Wendy Schmidt Center for Data Science & Environment, along with Cassie Buhler, a postdoctoral fellow at the Cooperative Institute for Research in Environmental Sciences (CIRES) at CU Boulder, suggest that perhaps taking the time to build AI smarter, rather than just bigger, could provide us with a path forward with a smaller environmental footprint.

Embracing more efficient open-source language models that are catching up with the capabilities of large enterprise models like ChatGPT and Claude, we may be able to find a scale of AI technology appropriate to our actual needs and uses, while taking a kinder, gentler approach to the scale and scope of data centers and their impacts on the environment, the climate and local communities.