That is the wrong drawer, and it is worth opening the ledger properly before anyone signs.

Three columns

Every immersion energy claim can be sorted into one of three columns. Ask which column a number belongs to and most of the confusion disappears.

Column one is where the kilowatt-hours are. Column two is real but modest. Column three is where budgets go to die. At Glacier Coolant we sort claims this way on purpose, because it is the fastest way to find out whether an energy case has been built or merely asserted.

Column one: what the architecture gives you

Three decisions, and they are not fluid decisions.

Fans gone. Air cooling spends most of its energy pushing air: fans on the components, on the racks, in the room. Immersion removes the component-level fans because the fluid picks up the heat directly. That removal is the single largest lever in the whole argument, and it is a change in how heat leaves the component, not a property of the coolant.

A warmer supply temperature. Air is a poor heat carrier, which pins air cooling to a low supply temperature. With fluid doing the pickup, the loop can run warmer and still keep the hardware happy. A warmer loop buys hours in which the site rejects heat straight to ambient instead of running the chillers.

A sealed volume. An immersed rack is a closed, dust-free box. The room no longer has to condition air for components that now live inside tanks. Less filtration, less conditioning, less fighting dust.

Those three lines are where the saving comes from. Two of them are decided before a fluid is ever specified, and the third is decided by the room plan.

Column two: what the fluid honestly gives you

The fluid is not irrelevant to energy. It is just not the headline, and pretending otherwise is what produces the disappointment later.

A fluid with needlessly high viscosity makes the pump work harder for the same heat moved. A fluid that ages early forces a change-out, and every change-out costs energy to drain, refill, reprocess and re-commission. A stable, correctly matched fluid avoids both.

Those are the fluid’s genuine energy contributions. Modest, defensible, and nothing like a brochure percentage.

They are also the two contributions a supplier can actually be held to. Ask what the viscosity is at your operating temperature and you get an answer with a unit on it. Ask how stable the fluid is across a long duty and you get a trend you can check against your own samples. Both are testable, and both survive an audit.

The architecture levers are larger, but they are also diffuse. Nobody sends you an invoice for the fans that did not have to be bought, and the free-cooling hours you won do not appear as a credit anywhere. That absence is exactly why the fluid ends up carrying a saving it did not produce: it is the only line on the page with a price attached.

Column three: what nobody should be selling you

This is the column worth reading twice.

A fixed saving percentage. Any “save X%” depends on what it is compared against, on the local climate, and on the system being replaced. It is a direction, not a figure. A supplier who will not say what the baseline was is not making a claim, they are making a slogan.

A dielectric claim as an energy benefit. Insulation is a property of the enclosed system, set by the equipment maker. It has no bearing on your kilowatt-hours, and we do not publish it.

A premium price presented as an energy purchase. Paying more per litre does not by itself change a single watt. The architecture delivers the saving. The fluid should be bought to support the architecture, not to stand in for it.

Every one of these fails the same test: ask which column it is in, and there is no honest answer.

Free cooling is a site number, not a brochure number

The free-cooling band is where most of the immersion energy argument actually lives, and it is the least transferable figure in the conversation.

It depends on where the site is, what the ambient does across the year, and how warm the loop is allowed to run. A design that wins a great many free-cooling hours on a temperate coast will not win the same hours inland or in the tropics. Same tank, same fluid, different answer.

This is why we ask for the climate and the permitted supply temperature before discussing grade. Without those two, any energy figure is a guess about someone else’s building.

Ask your design review for the band in hours per year, at your ambient, at your supply temperature. If that figure is not on the agenda, the energy case is still a direction rather than a number.

The same logic applies to the load profile. A hall that runs flat-out all year gets a different answer from one that peaks in bursts, because the free-cooling band only pays out during the hours the plant is actually rejecting heat. Two identical sites with different workloads will not bank the same hours.

This is uncomfortable for anyone who has already written a percentage into a business case. It is also the reason those cases tend to be revised downwards in the first year of operation, quietly, and usually after the supplier has been paid.

FAQ

Does immersion cooling save energy?

Yes, mainly by removing component fans and by allowing a warmer loop with more free-cooling hours. The saving is architectural rather than a fluid property.

What actually reduces PUE in immersion cooling?

Fan elimination, a higher permitted supply temperature that wins free-cooling hours, and a sealed enclosure that lightens the room HVAC load. Those three do the work.

Is the fluid responsible for the energy saving?

Not directly. The fluid’s energy role is pump power and change-out avoidance. Both are real, and neither is the headline.

Should I pay more for a fluid expecting energy savings?

A premium fluid does not by itself change your kilowatt-hours. Match the fluid to support the architecture, and buy the architecture for the saving.

Does a dielectric or insulation property save energy?

No. That behaviour belongs to the enclosed system and the equipment maker’s specification. It has no bearing on your energy number.

How does fluid viscosity affect energy?

Viscosity sets the work needed to circulate the loop. A fluid with needlessly high viscosity makes the pump work harder to move the same heat.

What is the energy cost of changing fluid?

Real across the life. An unstable fluid forces repeated drain, refill and disposal, and each one carries an energy bill of its own.

How should I compare immersion energy across vendors?

Compare architectures first — fans, free-cooling band, sealed volume — then the fluid’s viscosity and stability. Ignore fixed percentages that depend on someone else’s baseline.

One question that settles it in a meeting

You do not need a spreadsheet to sort this out. You need one question, asked at the right moment:

“Which of those three columns is this number from, and who is going to deliver it?”

Asked of an architecture claim, the answer comes back fast and with a name attached. Asked of a fluid claim, it comes back as pump power and change-out avoidance, which is honest and small. Asked of a percentage, it usually comes back as nothing at all, and the room moves on.

That question is the whole audit. It costs nothing, it takes ten seconds, and it is the difference between an energy case that survives contact with the first year of operation and one that does not.

The eight questions that stall a coolant decision covers the questions that come earlier. How to choose a secondary refrigerant by temperature does the same job for the loops outside the tank. The product range lists the families by temperature band.

Open the ledger before you sign

The fluid’s honest energy role is small and worth having. Low viscosity keeps pump power down. Stability keeps change-outs rare. Neither is a headline, and both matter over a decade of running.

The saving is the fans you did not buy, the chiller hours you did not burn, and the room air you no longer have to condition. All three are decisions, and every one has an owner.

Glacier Coolant supplies the fluid and the lever list. Send us the duty band, the enclosure plan and your site’s ambient profile. We will return an immersion grade matched to the architecture you are building, and the three-column ledger to take into your design review.