Overview

Why water is now a data center siting constraint

Data centers use water mostly for cooling: directly in evaporative systems, and indirectly through the power plants that supply them. As rack densities rise and campuses grow, municipalities, regulators and neighbors are looking harder at water draw, especially in drought-prone regions.

  • Evaporative cooling consumes water that doesn't come back to the local system.
  • Closed-loop and liquid cooling cut consumption but still need makeup water and heat rejection.
  • Municipal capacity and discharge permits can cap what a site is allowed to use.
  • Community concern about aquifers and local supply can slow or block approvals.
Decision Framework

What a project team should evaluate

01

Quantify demand

Start with cooling architecture, seasonal demand, water quality and operating profile.

02

Understand local exposure

Municipal capacity, drought, discharge limits and community concerns can affect site viability.

03

Evaluate supplemental supply

Atmospheric water generation and recovery can be evaluated where climate, scale and economics fit.

04

Integrate with cooling

Water decisions affect heat rejection and energy use, so they should not be modeled in isolation.

05

Plan resilience

Storage, redundancy and source diversity can reduce dependence on a single water source.

GTG Approach

How GTG approaches data center water

GTG atmospheric water generators produce fresh water from humidity in the air. Individual 10,000L units can be combined into clustered substations sized to campus demand, and output can be estimated by temperature and humidity with the AWG production calculator. Water vapor recovery can return part of what cooling systems lose.

Because water decisions affect cooling energy, GTG models them alongside onsite power. Background reading: why data center water is now critical infrastructure and the water source case study.

Common Questions

Frequently asked questions

How much water does a data center use?

It varies widely with cooling architecture, climate and operating profile. Evaporative designs can consume large volumes, while closed-loop and liquid-cooled designs use far less. GTG starts from the project's own demand profile rather than an industry average.

Can atmospheric water generation supply a data center?

It can where climate, scale and energy economics fit. Output depends on ambient temperature and humidity, and GTG units can be clustered into substations sized to campus demand. A site-specific estimate is the first step.

What is water usage effectiveness (WUE)?

WUE is a data center metric that divides annual site water use in liters by IT equipment energy in kilowatt-hours. Lower is better. It's a useful way to compare cooling and water strategies across sites.

Does onsite water help with community approvals?

It can address one of the most common concerns raised about new data centers: competition for local water supply. GTG's open letter to communities and operators sets out that case.

GTG Data Center Knowledge Center

Related guides

Alternative Data Center Power

Onsite, microgrid and hybrid options when the utility can't deliver.

Read guide →

Data Center Time-to-Power

Measure the gap between your target date and the utility's.

Read guide →

Onsite Data Center Power

Bridge and long-term generation on the campus itself.

Read guide →

AI Data Center Power

High-density loads, liquid cooling and phased capacity.

Read guide →

Data Center Infrastructure Gap

The framework that ties power, water and schedule together.

Read guide →

Bring GTG the site, load and target date.

We use MW requirement, water demand, utility status, resiliency requirements and project schedule to frame a project-specific infrastructure conversation.

Start Assessment