Cooling-water availability is a fundamental consideration in data centre site selection and infrastructure planning. Cooling is one of the largest non-IT resource draws in a facility, though cooling configurations vary. Assessing water consumption versus withdrawal clarifies site constraints alongside grid power as campuses scale into the tens and hundreds of megawatts, informing local utility requirements and infrastructure planning reviews.
How much water does a data centre use?
Consumptive water use is defined by the U.S. Geological Survey as the portion of withdrawn water that is evaporated, incorporated into products, or otherwise removed from the immediate water environment. In data centres, net water consumed differs from total water withdrawn from utilities, which can include potable, recycled or other water sources. Evaluating cooling demand requires examining operator reporting boundaries, peak seasonal draw, ambient climate conditions and site power dedicated to heat rejection.
Across an entire market, aggregated figures can mask local utility exposure. Where facilities cluster within a single catchment, commercial assessment depends on whether regional networks can sustain peak intake alongside treatment capacity and seasonal availability. Due diligence requires examining local utility constraints, source water types such as potable or recycled supplies, and the specific reporting boundaries applied across operating facilities.
Direct and indirect water: two figures that matter
A site's water footprint involves direct and indirect components, requiring clear operational reporting boundaries. Direct water covers what a facility withdraws or consumes on-site to cool equipment, such as when cooling towers evaporate water to reject heat. This intake can include potable, recycled or other industrial water supplies. Indirect water is consumed upstream during electricity generation across thermal or hydroelectric facilities. This indirect volume can exceed on-site consumption, and assessing both requires tracking utility generation data alongside on-site meter readings.
This distinction matters when comparing designs. The site WUE metric measures annual on-site water use against IT equipment energy use, while other variants may include source water. Because site WUE covers on-site operations, its reporting boundary excludes indirect grid-level consumption. Comparing designs requires evaluating local intake and grid-generation intensity, as systems with low on-site evaporation can require additional electrical power to reject heat.
Why dry regions face the hardest trade-off
Cooling choices involve trade-offs between site power demand and water intake under local climate conditions. While evaporative heat rejection can lower facility power consumption in dry conditions, site assessment must confirm whether municipal or regional supplies can support seasonal draw. Decisions established during planning consent define operational infrastructure early, prompting technical scrutiny of intake requirements, heat rejection designs and long-term utility commitments before equipment selection is finalised.
How water shapes where data centres get built
Water access forms a core feasibility consideration alongside power connection during site evaluation. Feasibility reviews question whether local utility capacity exists, which water sources are permitted, and how regional catchment policies apply. Examining utility supply agreements, planned intake volumes and environmental planning documentation clarifies whether a facility's cooling strategy is viable before capital commitments are made.
Evaluating site viability requires confirming how regional planning frameworks review cooling-water sourcing and utility infrastructure. Assessment processes typically examine proposed heat rejection methods, peak withdrawal rates, seasonal consumption profiles and supply resilience. Technical evaluations also review whether cooling systems rely on potable, non-potable or recycled water networks, ensuring that utility connection requests and operational intake models align with local catchment conditions.
Can operators cut their water use?
Several cooling approaches manage on-site water use through varied mechanical configurations. Closed-loop systems recirculate fluids to limit evaporative losses, while air-assisted systems can reduce direct water withdrawal, often altering electrical demand. Operators can also evaluate alternative sources, including recycled, industrial or treated non-potable water, alongside direct-to-chip designs. Evaluating each option requires checking whether reduced on-site withdrawal changes overall facility efficiency, system complexity or upstream power consumption.
Each cooling design presents commercial and technical trade-offs. Air-based systems can increase power requirements, influencing indirect water consumption depending on grid fuel sources. Utilising recycled water depends on local utility distribution infrastructure and quality standards. Evaluating sites involves testing how climate conditions, power availability and utility infrastructure align with the selected mechanical cooling specification.
Following the public record
Cooling infrastructure requirements are established during early facility design, meaning records of a site's water strategy appear across planning and utility documentation. Water authority filings, planning approvals and environmental impact assessments provide verifiable information on proposed intake volumes, source types and trade waste arrangements. Reviewing these records alongside power connection studies clarifies technical feasibility when assessing, acquiring or developing data centre assets. This evidence-based review helps capacity buyers, operators and investors verify whether water utility capacity, permitted discharge terms and environmental allocations match the operational requirements of the intended facility.
Frequently asked questions
How much water does a large data centre use?
Water intake and net consumption vary by cooling architecture, local climate and operator reporting boundaries. While evaporative systems consume water through active heat rejection, closed-loop circuits and air-cooled plant reject heat differently, requiring verification of actual evaporation against total utility withdrawal across seasonal peaks.
Why do data centres use water at all?
Data centres use water to reject heat generated by computing equipment. While evaporative cooling towers consume water through evaporation, other systems withdraw water for heat exchange before returning it. Operators may utilise potable, recycled or industrial sources, or select air-assisted designs to reduce direct on-site consumption.
What is the difference between direct and indirect water use?
Direct water is withdrawn or consumed on-site for facility cooling. Indirect water is consumed upstream during electricity generation. Because the site WUE metric measures direct on-site use within its reporting boundary, evaluating total water impact requires confirming the operator reporting boundary and grid generation source.
Is data centre water use a problem in Australia?
It can represent a significant consideration in supply-constrained catchments. Planning reviews and utility evaluations typically scrutinise a proposed facility's peak withdrawal rates, projected consumption profile and water source selection, requiring clear evidence on whether intake relies on potable, recycled or alternative supplies across seasonal operating peaks.
Where can I find a site's water information?
In public planning and utility records. Utility connection filings, environmental impact assessments and planning applications detail proposed cooling systems, water source types and projected consumption. Reviewing these documents alongside power connection records helps verify whether a facility's cooling requirements align with local utility capacity.
