The Data Centre Water Gap: Why Cooling Water Now Decides Where You Can Build

Key takeaways

  • Water is now a gating factor for data centre projects. Cooling-water supply affects siting and permitting, not just operating cost.
  • Efficiency gains do not close the gap: Global Water Intelligence forecasts that on-site cooling water use will still rise by just over 50% by 2030 while capacity doubles.
  • Treated wastewater (STP water) is a practical local source, but lower-quality water usually means more chemicals and more blowdown.
  • Treating water to remove scale-formers before they reach the heat exchangers, without chemicals, allows higher cycles of concentration and less water loss. CET Enviro reports typical payback of under two years.

Why is water now delaying data centre projects?

Data centre projects are increasingly held up by a question that has little to do with power or capital: where will the cooling water come from?

In the first quarter of 2026, at least 75 major projects worth more than $130 billion were delayed or cancelled, with organised local opposition cited as a contributing factor. In the second quarter, Data Center Watch counted 45 US projects worth about $68 billion blocked or delayed. Water use sits alongside electricity cost among the concerns communities raise most often.

The pattern is not confined to one country or one climate. Wherever a campus competes with households, agriculture, or other industry for the same supply, cooling water becomes a permitting question before it becomes an engineering one.

This post looks at why cooling is where the water goes, what an under-used source (treated municipal wastewater) can and cannot do, and how the choice of water treatment method changes both the design and the economics.

Why does data centre cooling use so much water?

Most on-site water use in a data centre is heat rejection. Evaporative cooling is attractive because evaporation moves a great deal of heat for very little energy, but it consumes water to do it. A Guardian analysis, as reported by ConstructConnect, notes that large facilities can need up to 5 million gallons a day for cooling.

Two things have turned that from an operating cost into a siting question:

  • Where projects are going. The same analysis found that 517 of 809 planned US data centres are in areas that experienced drought over the past year.
  • How the numbers are trending. US data centre water demand was about 17 billion gallons in 2023, and the Guardian cites projections that could take it toward 73 billion gallons annually by 2028. Projections vary by scenario, so treat the upper figure as an indication of direction.

Efficiency gains help but do not close the gap. Global Water Intelligence forecasts that on-site cooling water use will still rise by just over 50% by 2030, even as capacity doubles. It also notes that the more water-efficient cooling technologies need higher-quality water.

The obvious alternatives trade one resource for another:

Approach Effect on on-site water Trade-off
Evaporative cooling towers Highest Energy-efficient, but needs a reliable water supply and treatment
Free-air cooling Low Climate-dependent; often ineffective in hotter regions
Mechanical chillers, dry cooling Low Higher electricity demand, which moves water use upstream to power generation
Closed-loop, liquid cooling Substantially lower Higher upfront capital and tighter integration; still needs water-quality management for scaling and corrosion

Table draws on Pillsbury’s review of water reuse in data centre development and the GWI note above.

The practical conclusion is that water use cannot simply be designed out. The more useful question is where the water comes from, and how well it is managed once it is in the loop.

Can data centres use treated wastewater for cooling?

Municipal sewage treatment plants (STPs) already produce a steady, local stream of water that is discharged after treatment. For a cooling tower, which has no need for drinking-water quality, that stream is an obvious candidate.

Operators have started to use it:

If it is so available, why is it not the default? Three reasons come up repeatedly.

  1. Water quality. Treated effluent generally carries more dissolved minerals and organic matter than potable water. In a cooling loop that concentrates the water through evaporation, those solids drive scaling, corrosion and biological growth. Reuse systems need treatment capable of meeting strict quality thresholds to protect the equipment.
  2. Rules differ widely by jurisdiction. In the US, for example, California requires disinfected tertiary recycled water where cooling towers are involved, while Arizona’s rules prohibit direct reuse for evaporative cooling. The concern behind both is aerosolisation and public exposure.
  3. Chemical dependence. The usual way to run a tower on lower-quality water is to dose more scale inhibitors, biocides and corrosion inhibitors, and to discharge more blowdown. That erodes part of the water saving the reuse was meant to deliver, and adds a chemical stream to manage.

The third point is where the choice of treatment method starts to matter.

How can cooling water be treated without chemicals?

A conventional cooling water programme keeps dissolved minerals in solution with chemical inhibitors, then bleeds water off before concentrations get too high. Treatment and water loss are linked: the more the water is chemically managed, the longer it can cycle, and the less it needs to be replaced.

A second approach works on the water before it reaches the heat exchangers. It precipitates the scale-forming salts in a controlled reaction step, so that what circulates through the condensers and chillers has far less tendency to scale. The design intent is that the water chemistry is corrected physically rather than held in check by dosing.

CET Enviro’s SBR is one implementation of this idea. It is designed as a chemical-free treatment, and it can operate on STP-quality water, which is the combination that matters for the sites described above. The company reports:

  • 2 to 3 times higher cycles of concentration, which means proportionally less make-up water and less blowdown;
  • No scale reaching heat exchangers, which protects chiller and condenser efficiency;
  • No scale-inhibitor or biocide programme to buy, store, dose and discharge;
  • Technology readiness level 9, with units in operation since 2018 and around 200 SBR installations (CET Enviro’s wider installed base is above 2,000 systems).

In the data centre sector, SBR™ has been installed at Sify, jointly with CET Enviro’s ACCS™ and AMOS systems.

Higher cycles matter more than they first appear. On a recycled-water supply, the cost of water is not only the purchase price. It includes the treatment upstream, so every cycle gained reduces both the volume to be treated and the effluent to be handled.

One caution applies to any technology in this category. Installation counts say little about fit. Operators should ask for operating data from loads that resemble theirs: continuous duty, high heat density, and tight uptime requirements.

What does the payback look like?

CET Enviro reports typical payback of under 2 years for SBR™ installations. That figure is a range of outcomes, not a promise for every site, so it helps to see what it depends on.

The annual savings come from several lines that move together when cycles of concentration rise:

Cost line What changes What to measure on your site
Make-up water Fewer litres replaced as cycles go up Current cycles of concentration, make-up volume, delivered water price
Blowdown and effluent Less water discharged and treated Blowdown volume, sewer or STP treatment charges
Chemicals Inhibitor, biocide and corrosion programme no longer needed Annual chemical spend, dosing and storage costs
Labour and compliance Fewer dosing checks, handling and discharge records Hours on water treatment, reporting obligations
Energy Cleaner heat-transfer surfaces keep chiller and condenser performance closer to design Approach temperatures, chiller kW per ton over time

Sites with cheap water, low blowdown and a modest chemical bill will see a longer payback than the headline range. Sites where water is priced at scarcity, or where STP effluent charges are significant, will see a shorter one.

For a data centre there is a second return that does not appear on a payback sheet. A design with lower make-up demand, no chemical discharge and a recycled-water source is easier to explain to a permitting authority and to a community that is already asking about water.

The takeaway

Water has joined power as a gating item for data centre projects. The sites that move through permitting smoothly are likely to be the ones with a credible water story: a low draw, a recycled source, and little to discharge.

Treated wastewater is already available near many planned sites. The harder part has been running cooling towers on it economically, because lower-quality water usually means more chemicals and more blowdown. How the water is treated decides how much of the saving survives.

If you are comparing cooling water options for a planned or existing site, CET Enviro can run a water balance and payback estimate against your own make-up, blowdown and chemical figures. Details are at cet-enviro.com.

Sources

Why Cooling Systems Waste High-Quality Water in Food & Beverage Plants (And How to Stop It)

Not all water carries the same value. Every industrial facility runs on an implicit hierarchy for its water, even if no one has written it down. Some water is essential and non-negotiable — it becomes the product, or it keeps the product safe. Other water exists to support operations, and how carefully that water is managed varies enormously from plant to plant.

In food and beverage manufacturing, this hierarchy is unusually visible — and unusually costly to get wrong.

Most of a F & B plant’s water is water it can’t compromise on

Water used in food and beverage manufacturing generally falls into three categories: process water (which touches the product directly or indirectly), cooling water (which removes heat from processes and equipment), and boiler feed-water (used to generate steam for sterilization and heating).1

Across the sector as a whole, close to three-quarters of the water used is drinking-quality water.1 That is the standard required wherever water comes into direct contact with food — and, in many plants, it’s also the standard applied more broadly than strictly necessary, simply because it’s the water that’s already on hand.

Cooling doesn’t need to meet that bar. Heat rejection from chillers, condensers, and process equipment has no food-contact requirement. And yet cooling systems routinely draw on the same high-grade water supply as the parts of the plant where drinking-quality water is genuinely non-negotiable.

The architecture problem hiding in plain sight

Part of the reason comes down to system design. Cooling architecture in food and beverage plants generally falls into a few categories: once-through systems (no recirculation), closed circulation loops, open circulation systems using cooling towers and direct-contact cooling.

Of these, once-through cooling systems carry the highest water demand of any cooling architecture available.1 They’re simple and reliable, which is exactly why so many plants still run them — but simplicity here comes at a real and ongoing water cost, one that compounds every day the system operates.

Fouling adds a second, quieter cost

Even in a more water-efficient closed-loop or cooling-tower system, a second factor erodes efficiency over time: Fouling.

As scale and biological deposits build up on the inside of condenser and heat exchanger tubes, heat transfer efficiency drops. A fouled system can no longer reject the same amount of heat with the same water and energy it used when the tubes were clean — so operators compensate. That typically means more water flow, more frequent blowdown, and more makeup water, just to hold the same operating setpoint a clean system would reach more efficiently.

The usual response — chemical dosing or periodic mechanical descaling — solves the immediate scaling problem but introduces new ones: chemical handling and disposal, tube wear from repeated mechanical cleaning, and downtime for the cleaning itself.

A different way to keep condensers clean

CET Enviro’s SBR™ (Scale and Bio-Removal) system addresses fouling at its source rather than after the fact. Instead of dosing chemicals or manually descaling, SBR™ precipitates scale-forming salts out of the water in a dedicated reaction chamber, before they ever reach the heat exchanger surface. That keeps tubes clean continuously, without chemicals — and it allows cooling towers to run at significantly higher cycles of concentration, which is what actually drives the water savings.

In practice, that translates to 30–70% water savings and complete elimination of chemical dosing, typically with a payback period under two years.

For a food or beverage plant, the case is straightforward: water spent compensating for a fouled system, or flushed out to control scale buildup, is water that never needed to compete with what the plant needs for product and hygiene in the first place.

Where to start

If you don’t know how much of your plant’s water use cooling actually accounts for or how much of that is being spent fighting fouling rather than doing useful work, that is  the first thing worth finding out.

Learn more about SBR™ →

Request a free facility audit →

References

  1. Valta, K., Moustakas, K., Sotiropoulos, A., Malamis, D., & Haralambous, K.J. (2016). Adaptation measures for the food and beverage industry to the impact of climate change on water availability. Desalination and Water Treatment, 57, 2336–2343.
  2. https://academic.oup.com/fst/article/34/4/58/7818771

5 Ways to Clean the Condenser

A clean condenser keeps your operations steady and efficiency high. And lowering in the efficiency of the system can lead to considerable losses of time, energy and finances. That’s why regular cleaning is important.

Here are five methods to clean condensers:

Using Automatic Tube Cleaning System (ACCS)

As the name suggests, Automatic Tube Cleaning System is an advanced measure of cleaning condenser tubes. A dirty condenser faces many issues including fouling and depreciation of the tube’s material. The ACCS gets rid of all these issues and that too, in a very small amount of time. It optimizes the heat transfer performance and removes downtime. This helps you in getting effective results with maximum efficiency. If you’d ask us, we’d suggest going with ACCS.

Cleaning with a Brush

The oldest method of cleaning condensers is through brushes. In this method, you’d attach a brush on a long rod, making sure that the brush’s bristles are longer than the radius of the tube. Then you’d use the rods to clean the interior of the tubes after throwing in some water for lubrication. It’s an obsolete method and has been discarded by many organizations. Its main issue is it leaves a lot of residue inside the tube and takes up a lot of time and energy.

Cleaning with TCGs

TCG stands for Tube Cleaning Gun. These guns project high-pressure water in the tubes and target the dirt. Like the ‘brush method’, this method is also old and obsolete. The primary issue with this procedure is the consumption of time. And it has the risk of increasing the clog inside the tubes as the high-pressure water may not be able to get rid of all the dirt.

 

Cleaning with Chemicals

In this method, one adds acid solutions and circulates them through the tube bundles. This softens the scale deposits which are then cleaned with brushes. Chemicals are quite costly, and that’s why this method is also discarded in modern operations. Moreover, to perform this procedure effectively you’d need a highly trained expert.

Using COLTCS

COLTCS stands for Condenser Tube Cleaning System. In this method, specialized sponge rubber balls are inserted in the Condenser. They prevent the build-up of any scaling or fouling. After the passage, the system retrieves the balls and pumps them into inlets. It’s an automatic system. This means you won’t have to worry about its operation and efficiency. This is another favored method of multiple industries.

What’s your Choice?

Which method would you prefer for your condenser’s cleaning? Let us know through your Comment. We’d love to hear from you. You can find out more about ACCS and COLTCS.