As AI-driven data center development accelerates, developers, communities and regulators are increasingly focused on water demand—both the volume required and the sources from which that water will be drawn. While industry attention has largely centered on electricity procurement and grid impacts, the availability and legal entitlement to a firm water supply has become equally material to siting, permitting and community acceptance. Particularly as surface and groundwater supplies become increasingly constrained and new projects are sited in regions experiencing tighter hydrologic conditions or growth-related supply constraints, project teams are increasingly integrating water supply analysis into early-stage development to address issues that can materially affect schedule, financing and long-term operations.
The Escalating Demand for Water
Data centers require substantial water resources, primarily for cooling systems that dissipate heat generated by servers operating around the clock, but indirectly through electricity consumption, much of which is supplied by power plants that themselves rely on significant volumes of water for cooling. Evaporative cooling, for example, is a process in which water absorbs heat but eventually becomes too mineralized for reuse as water vapor escapes, eventually requiring additional water supplies. While water use varies based on facility size, cooling system design, geography and power configuration, hyperscale data centers can require as much as 3-5 million gallons per day. For context, daily demand at that scale is roughly equivalent to the municipal water use of a town of 10,000–50,000 people. Projections—from both corporate forecasts and federal analyses—indicate that this demand is accelerating rapidly, with annual onsite water consumption by U.S. data centers potentially doubling or even quadrupling by 2028. (In Texas, for example, data center water demand was estimated to be approximately 25 billion gallons in 2025. By 2030, that volume could range to between 29 billion and 161 billion gallons annually.)
When indirect use from power generation is included, a data center’s total water footprint can triple. A typical AI-focused data center consumes as much electricity as approximately 100,000 households, while current hyperscale facilities may require 20 times that amount, leading to projections that U.S. electrical demand will more than double by 2030. The Electric Reliability Council of Texas’s estimates peak electricity demand in the state will likely rise by at least 70% by 2031—an increase driven largely by data centers, which represent 73% of the current five-year queue of large-load interconnection requests.
Compounding these pressures is the growing concentration of data center development in regions experiencing high water stress, where total demand approaches or exceeds available supply. An estimated 38% of existing U.S. data centers operate in high water stress areas, and roughly two-thirds of facilities built or under development since 2022 are located in such regions. These siting trends intersect with broader hydrologic realities: Approximately 70% of major aquifers are in long-term decline, intensifying the water constraints confronting municipal and industrial users.
Water Reuse, Rights and Regulatory Fit
Water rights present a separate and often determinative constraint on data center development, particularly in the Southwest and West, where states operate under prior appropriation regimes. Unlike riparian systems in which water use is tied to land ownership, appropriative rights are governed by a strict priority framework—“first in time, first in right”—under which junior users may be curtailed during shortage conditions pursuant to state-administered priority administration or call systems. Data centers relying on junior water rights face heightened exposure to curtailment risk and increased scrutiny by lenders and investors assessing long-term operational resilience.
At the same time, local governments are evaluating access, whether proposed supplies will remain available, and whether cooling systems can continue to operate under reduced allocations or basin-wide cutbacks during drought. In practice, these considerations are now frequently reflected in entitlement conditions, utility service agreements and drought-response planning requirements imposed as part of project approval.
Cooling system designs continue to evolve as well, with some facilities incorporating hybrid configurations that can shift between evaporative cooling—which uses water evaporation to dissipate heat and is more water-intensive but energy-efficient—and dry cooling—which relies primarily on air or refrigerants and significantly reduces water use but increases energy demand and capital cost—depending on season, load or real-time water availability.
While these approaches can materially reduce potable or raw water withdrawals, project sponsors evaluating reuse strategies should account for operational and regulatory complexities. In several jurisdictions, developers and utilities are negotiating first-of-their-kind reclaimed water supply agreements that allocate treatment risk, drought interruptibility, and long-term volume ramp-up in ways not contemplated by traditional municipal service contracts. Structuring these arrangements requires alignment between state reuse permitting regimes, water right transferability, and utility rate authority.
EPA has recently indicated that forthcoming updates to its Water Reuse Action Plan (WRAP) will more directly address industrial-scale reuse, including reuse associated with data center cooling demand. While these updates may provide additional federal guidance and technical resources, primary permitting and allocation authority will remain with the states. In the near term, many state reuse frameworks—historically designed for irrigation, agricultural or limited industrial applications—do not always provide clear pathways for continuous, high-volume cooling use. As a result, enabling large-scale reuse often requires project-specific permit conditions, coordination among multiple agencies, or tailored supply agreements to achieve firm-supply treatment under state law.
At the same time, federal policymakers are considering financial incentives to accelerate reuse infrastructure deployment. A recently introduced bipartisan House bill, H.R. 2940, Advancing Water Reuse Act, would establish a 30 percent tax credit for qualifying water reuse projects that install or modify onsite recycling systems within industrial, manufacturing, data center, or food processing facilities; replace freshwater supplies with municipally provided recycled water; or build or expand municipal recycling systems to secure recycled supply for industrial use. If enacted, such incentives could materially affect the economics of reclaimed water infrastructure and further encourage utility–developer collaboration in high-growth markets.
Even then, reclaimed water quality can be variable and often requires additional treatment, monitoring and contingency backup supplies. Supporting infrastructure—treatment, storage and dual distribution—can entail significant capital investment and long-term operational commitments aligned with utility planning and regulatory approvals.
A State- and Public-Driven Push for Transparency …
These supply and water rights constraints are increasingly reflected in local permitting practice. Decision-makers are paying closer attention to the cumulative, long-term effects of multiple large facilities drawing from the same basin, aquifer or distribution system. Even in the absence of formal disclosure mandates or environmental impact requirements, local governments and water districts are increasingly relying on discretionary land-use authority, utility service conditions, and informal consultation processes to delay approvals, require supplemental water supply analyses or revisit growth assumptions. In practice, these mechanisms have increasingly made demonstrable, drought-resilient water a gating issue for schedule certainty and financing.
Many industrial water users consider their consumption data proprietary or confidential business information, and contractual confidentiality provisions generally prevent disclosure of facilities’ volumetric usage or commitments by utilities. Public records requests directed at utilities for customer-specific usage data are thus frequently challenged, with outcomes often turning on the particulars of state public records law and the scope of applicable exemptions.
This dynamic has also heightened public interest in how large industrial users account for long-term water demand within constrained basins. In several high-growth jurisdictions, sustained concern regarding cumulative withdrawals has contributed to extended entitlement timelines and, in some instances, project reconfiguration as sponsors engaged with local officials to address disclosure and basin-level supply questions.
In response, a growing number of states have pursued legislative measures aimed at increasing transparency around data center water use. In 2025, lawmakers in at least nine states—including Connecticut, Georgia, Illinois, Indiana, Maryland and Oregon—introduced bills requiring disclosure of projected or actual water consumption to local suppliers or state agencies.[1]
Taken together, these developments reflect a persistent tension: while pressure for greater transparency continues to build, many states remain cautious about adopting requirements perceived as undermining their competitiveness in attracting data center investment.
… Meets an Administration’s Prioritization of AI
President Trump has prioritized positioning the United States as the “AI capital of the world,” and issued an Executive Order (EO), “Ensuring a National Policy Framework for Artificial Intelligence,” directing federal agencies to identify and challenge state-level AI regulations deemed inconsistent with national objectives and to evaluate potential federal preemption of certain reporting frameworks. Although the EO excludes “AI compute and data center infrastructure,” its broader emphasis on limiting state-level disclosure requirements introduces uncertainty regarding the durability of certain state transparency initiatives affecting AI-related facilities. State water-use disclosure mandates could face preemption arguments if framed as AI oversight rather than traditional water resource management.
At the same time, federal agencies are taking parallel steps to facilitate data center development and related infrastructure. The U.S. Army Corps of Engineers is finalizing its five-year overhaul of the Clean Water Act Section 404 dredge-and-fill permitting program, including revisions to the nationwide permit (NWP) framework. While most existing NWPs were reissued largely unchanged, the Corps included a notable modification to streamline permitting for data center and AI-related facilities. Coverage under an applicable NWP can shorten permitting timelines by avoiding individual permits. Early evaluation of NWP eligibility and notification thresholds can materially affect schedule assumptions for projects involving wetland fill, pipelines, or transmission infrastructure.
Congress has likewise signaled interest in accelerating infrastructure deployment. The Promoting Efficient Review for Modern Infrastructure Today (PERMIT) Act, which passed the House of Representatives and was referred to the Senate Environment and Public Works Committee, includes provisions aimed at reducing permitting risk and delay for infrastructure projects, including data centers and associated energy facilities. The bill would narrow the scope of “waters of the United States,” constrain Section 401 authority, and limit Clean Water Act litigation timelines for infrastructure projects.
Taken together, these federal actions reflect a clear push to streamline approval pathways for AI-related infrastructure, even as state and local governments continue to grapple with water availability, disclosure, and cumulative resource impacts at the project level.
Strategic Water Planning Considerations for Data Center Developers
Data center developers increasingly treat water as a core determinant of project feasibility. At the site selection stage, project teams now incorporate water use intensity analysis alongside traditional power demand forecasting. This includes evaluating cooling systems, modeling system performance under varying climatic and load conditions, assessing the seniority, transferability, and reliability of available water rights, and anticipating how future AI-driven capacity expansions could materially alter water demand. Stress-testing proposed designs against drought scenarios, curtailment risk, and supply interruptions has become an essential step in ensuring that early design decisions remain viable over the facility’s operational life.
In practice, water supply agreements have evolved from standard utility service contracts into negotiated frameworks allocating drought, curtailment and expansion risk. Utilities and developers now routinely address priority-of-service provisions, interruptibility terms tied to drought declarations, volume guarantees and ramp schedules aligned with phased compute buildout, and cost-sharing arrangements for pipelines, pump stations, or recycled water infrastructure. These agreements also address reclaimed-water quality standards and contingency backup supply requirements. In jurisdictions where reclaimed or recycled water is not yet recognized as a firm or transferable supply, developers are engaging counsel early to structure reclaimed-water arrangements that achieve firm-supply treatment under state law while preserving operational flexibility.
Comprehensive assessments—including aquifer-specific modeling, sustainable yield analysis where applicable and evaluation of groundwater management plan consistency—can meaningfully reduce permitting friction and community opposition by addressing concerns that municipal staff and residents now routinely raise during entitlement proceedings. In some jurisdictions, water resource analysis is already becoming a condition of zoning or site plan approval, making early coordination with watershed districts and technical agencies, including relevant USGS Water Resources Mission Area Science Centers, an increasingly important component of project development.
These structuring considerations are increasingly embedded in site diligence, utility negotiations, and entitlement strategy across high-growth markets.
Conclusion: Aligning Water Strategy with Development Certainty
For data center developers, water is no longer a secondary utility input—it is a threshold feasibility and financing variable. As infrastructure expands into regions with increasingly constrained supplies, successful projects are those that treat water availability, water rights, power procurement and permitting strategy as integrated components of a single development framework.
Best practices increasingly include:
- Conducting parallel legal and hydrologic diligence during site selection, including review of priority, transferability, place-of-use limitations and curtailment history;
- Structuring water supply agreements to allocate drought and interruptibility risk in a manner aligned with phased buildout;
- Evaluating reclaimed and recycled water not only from a technical standpoint, but for long-term regulatory durability and firm-supply recognition;
- Stress-testing cooling system design against shortage scenarios and basin-level cutbacks; and
- Engaging early with utilities, groundwater sustainability agencies, irrigation districts and state regulators to reduce entitlement friction.
Projects that proactively address long-term supply reliability, priority risk, cumulative basin impacts and evolving disclosure expectations are better positioned to maintain schedule certainty and financing confidence.
Pillsbury’s Environmental & Natural Resources team works alongside data center developers, utilities and investors to structure firm water supplies, acquire and transfer water rights, negotiate drought-resilient service agreements, design reclaimed water strategies, and navigate permitting and environmental review. We are actively advising on these issues across multiple jurisdictions and will continue to monitor state and federal developments as the regulatory framework evolves.
[1] Mandatory reporting measures in California and New Jersey passed both legislative chambers but were ultimately vetoed amid concerns about regulatory rigidity, economic competitiveness, and the potential exposure of proprietary information. Similarly, Virginia—currently home to the largest concentration of data centers globally—passed legislation that would have authorized local governments to require site-specific assessments of water resource and land-use impacts before approving new facilities, only to see it vetoed by the governor.
Gravel2Gavel Construction & Real Estate Law Blog


