Do Data Centers Use Too Much Water and Raise Consumer Energy Costs?

Data Centers Use Too Much Water and Raise Consumer Energy Costs Data centers become more controversial by the day.

Are data centers a unique threat to communities, or is the concern exaggerated bullshit?

As artificial intelligence and cloud computing expand across the global economy, data centers have emerged as a lightning rod for community pushback. Critics argue that these sprawling facilities consume disproportionate amounts of electricity and water, drive up utility bills for everyday residents, and inflict unnecessary environmental harm.

Calls for moratoriums or outright bans on data center construction have gained traction in local, state, and federal government.

However, energy analysts, economists, and technology researchers argue that broad bans miss a critical reality: data centers are not merely energy consumers—they are driving unprecedented corporate investment in renewable energy, spurring power grid modernization, and advancing water-recycling technologies.

Unprecedented Gains in Computing Efficiency
A primary claim made by opponents is that data centers represent a wasteful burden on the electrical grid. While total electricity demand from the sector is rising, projected by the International Energy Agency (IEA) to double globally by 2030, this growth occurs alongside dramatic improvements in energy efficiency.

Over the past decade, cloud providers have transitioned from dispersed, inefficient enterprise servers to centralized “hyperscale” facilities. According to studies by the Lawrence Berkeley National Laboratory, hyperscale centers operate at a fraction of the power overhead of traditional server rooms. Industry efficiency is measured through Power Usage Effectiveness (PUE)—the ratio of total facility energy to the energy delivered to computing equipment. While older enterprise facilities often operated at PUE ratios of 2.0 or higher (meaning half of the energy went to cooling and power distribution), modern hyperscale facilities average around 1.1 to 1.2.

Furthermore, the IEA notes that energy consumption per individual computing task is falling rapidly. Concentrating computational workloads into ultra-efficient data centers uses significantly less energy than powering millions of individual, unoptimized physical servers across decentralized offices.

Catalyzing the Clean Energy Transition
Rather than hindering the transition to clean energy, major technology companies have become the largest corporate buyers of renewable power in history.

According to data from the U.S. Department of Energy (DOE), tech hyperscalers accounted for roughly 40% of all corporate Power Purchase Agreements (PPAs) for renewable energy globally in recent years. By signing long-term power purchase contracts, data center developers provide the guaranteed revenue that energy companies need to secure financing for new wind, solar, and battery storage projects.

Moreover, because data centers require “firm” power that operates 24 hours a day, 7 days a week, tech operators are funding next-generation clean energy technologies that might otherwise lack commercial backers. These include:

Advanced Nuclear Energy (SMR Off-Take Agreements). Tech hyperscalers are entering major agreements to fund and deploy Small Modular Reactors (SMRs). For instance, Utility Dive reports on Amazon’s funding of SMR projects in Washington state and Google’s agreement with Kairos Power to deploy a pipeline of advanced reactors by 2030.

Enhanced Geothermal Systems (Deep Geothermal Partnerships). Tech companies are driving commercialization of next-generation geothermal energy. Details on Google’s world-first partnership with clean-energy startup Fervo Energy to supply always-on carbon-free power to data center grids can be viewed directly via the Google Cloud Infrastructure Blog.

Long-Duration Battery Storage (Solar/Wind Co-location). Large tech operators rely heavily on corporate solar-plus-storage Power Purchase Agreements (PPAs) to smooth grid variable output and meet round-the-clock carbon-free energy commitments. Insights into how tech companies leverage long-term renewable PPAs and storage strategies are outlined in Introl’s analysis of AI data center renewable energy strategies.

Ratepayer Protection and Utility Structure
Critics frequently allege that data center expansion directly drives up utility rates for residential households by forcing utilities to build costly new generation plants. While rapid load growth presents planning challenges for public utility commissions, industry experts point out that the financial mechanism depends entirely on regulatory policy.

State regulators across the U.S. are increasingly adopting specialized rate structures, such as Clean Transition Tariffs and large-load connection fees. These tariffs mandate that large commercial users cover the full cost of dedicated transmission upgrades and new generation assets rather than passing those costs onto residential customers.

When properly regulated, large industrial energy users can actually lower rates for residential consumers by expanding the overall utility billing base. Fixed grid maintenance costs are spread across a larger volume of electricity sales, reducing the per-kilowatt-hour burden on individual homeowners.

Evolving Water Management Technologies
Concerns surrounding water usage stem primarily from older evaporative cooling systems, which rely on evaporating potable water to dissipate server heat. However, water-intensity metrics across the industry have shifted significantly in response to environmental scrutiny and resource scarcity.

Modern data center designs are moving away from freshwater cooling in favor of closed-loop systems, direct-to-chip liquid cooling, and air-cooled chillers that consume little to no water during normal operation. When water is required, operators increasingly utilize non-potable water, such as industrial wastewater or treated municipal greywater, preserving local drinking water supplies.

According to a report from the Environmental and Energy Study Institute (EESI), data centers account for less than 0.1% of total public water withdrawals in the United States—a fraction of the water consumed by traditional heavy manufacturing, thermal power generation, or agriculture.

The Risks of Broad Bans
Proponents of data center expansion caution that banning these facilities would carry significant economic and technological consequences without solving underlying energy challenges.

General Economic and Technological Consequences. Industry analysts and research groups note that placing moratoriums or broad prohibitions on computing infrastructure risks stifling tech innovation and weakening broader digital services. A comprehensive overview of how data centers underpin national infrastructure and output can be found in the PwC / Data Center Coalition Economic Contribution Study, which evaluates the industry’s $900B+ contribution to U.S. gross domestic product and millions of supported jobs.

Digital Infrastructure Breakdown. The essential role of server facilities in maintaining real-time societal operations—ranging from digital banking and medical health networks to emergency logistics—is detailed in the World Resources Institute (WRI) Analysis on Data Center Community Impacts.

Geographic Leakage. Energy researchers highlight that localized bans or restrictive supply limits often result in “carbon leakage”—where computational workloads simply migrate to neighboring states or international regions with weaker environmental protections or dirtier fossil-fuel power grids. This dynamic and its impact on regional energy planning is explored by the Stanford Woods Institute for the Environment.

Economic Stagnation. Data centers generate substantial local property and utility tax revenues that directly support municipalities, fire districts, and public school systems without adding significant municipal burdens (such as increased student enrollment or local traffic). Detailed economic evaluations document how local jurisdictions receive millions in net tax surplus, directly offsetting residential school funding burdens.

While it’s reasonable to have concerns about the rapidly expanding infrastructure needed to support new data centers, we find the claims about energy costs and water crisis to be mostly bullshit

Claims about data centers are mostly bullshit
Home » Do Data Centers Use Too Much Water and Raise Consumer Energy Costs?