By Brian Workman | CEO & Founder, ExhaustFlow Technologies | 28 Years in Mechanical Engineering and HVAC Design
Let me tell you about a meeting I have been in more times than I can count.
An operations executive is standing in front of a slide showing energy costs. The numbers are up — again. The team has already replaced two chillers in the past three years. There is a capital request pending for another one, plus a generator upgrade. Maintenance costs have gradually been increasing, and, nobody has a clean explanation for why.
What nobody in that room is talking about is the air outside the building.
That is usually where the real problem is.
In the typical overcrowded chiller plant, some percentage of energy spend, capital budget, and maintenance spend is addressing something that was never intentionally selected, and is rarely recognized: the expense of hot discharge air entering the condenser intakes. The equipment battles conditions it was not designed for. It consumes more energy, provides less cooling capacity, wears out sooner and is replaced…only to have the same conditions meet the new equipment and the cycle repeats itself.
Nobody budgeted for this. It is not a line item. But it is real, and in many facilities it has been compounding since the day the plant was commissioned.
| The most expensive operational problems are the ones that get absorbed into the baseline rather than identified as the cause of it. |
What Is Actually Happening
The physics are simple. When you cram air-cooled chillers into a limited mechanical footprint — which is how most data centers built in the past ten years were constructed — there isn’t much place for the hot air they’re blowing to go. It goes up, it stagnates, it flows back down. Some of it is sucked into the condenser inlet of an adjacent machine.
That neighboring unit is now operating in air that is warmer than the actual ambient. Not a little warmer. We measure recirculation effects of 10 to 30 degrees Fahrenheit above ambient on a regular basis. At 20 degrees above design, a chiller that was sized to carry a specific load may only be able to carry 70 percent of it. The rest of the capacity is gone. Not because the chiller is broken. Because the air around it is wrong.
Here is the part that tends to land hard when I walk through it with a CFO or VP of Operations.
The chiller has no way to know the air is wrong. It cannot compare its current inlet temperature to what the manufacturer assumed when the performance curve was drawn. It just responds to what it sees. And what it sees drives every downstream outcome: how hard it works, how much power it draws, how much cooling it delivers, how long it lasts.
Change the inlet air and you change all of it. That is the whole argument.
Can You Find the Hidden Budget Cost?
No? Neither can we and given the size of this cost, most leaders are surprised. The cost of air recirculation does not show up as its own line. It hides inside categories that already exist and already look normal.
Energy costs are the obvious one. A plant running at 20 percent below its designed efficiency is burning roughly 20 percent more electricity to deliver the same cooling output. For a medium-scale data center that might be several hundred thousand dollars a year. The bill arrives every month and gets paid without anyone connecting it back to an airflow problem that has been present since day one.
Capital expenditure is the one that really stings, though.
When a chiller plant cannot meet load, the standard response is to add capacity. Another chiller unit. A larger generator. More electrical infrastructure. All of it goes through capital planning, gets approved, gets installed. And in many cases, not all, but many, the capacity shortfall was not a demand problem. It was a recirculation problem that made existing equipment underperform. Unfortunately, the new chiller gets added to the same environment that plagued the earlier units. The plant has more nameplate capacity but the thermal situation has not changed.
I have walked plants where this has happened two or three times over the life of the facility.
| Budget Category | How Recirculation Inflates It |
| Energy costs | Up to 20% excess consumption from efficiency penalty |
| Chiller capital | Capacity added to solve a performance gap that was never a sizing problem |
| Generator sizing | Up to $29,000 per chiller in excess capital from inflated electrical demand |
| Electrical infrastructure | Wire, switchgear, and overcurrent protection sized to inflated load |
| Maintenance and service | Accelerated wear from thermal stress; more trips, more calls |
| Equipment replacement | Shortened service life from operating outside design conditions |
That money has already been spent, in most cases.
That money has already been spent, in most cases. It cannot be recovered. But it illustrates what the ongoing version of this problem is costing in energy, maintenance, and the next capital cycle.
| Some of what looks like a demand problem is actually a performance problem. The plant was not undersized. It was operating in the wrong conditions. |
Why It Stays Invisible
Three reasons, mostly.
First, the equipment checks out fine. When a service technician runs diagnostics on an underperforming chiller, they are looking inside the unit. Refrigerant charge, compressor operation, controls behavior. If those things are normal — and they usually are — the unit gets a clean bill of health. What is not getting measured is the temperature of the air entering the condenser coils under peak load conditions. That requires a different kind of attention, and it rarely happens during a standard service call.
Second, the baseline was never right. A facility that was commissioned with recirculation baked into its operating conditions has never seen what the plant should actually perform like. The energy costs that showed up in month one became the benchmark. Budget was built around them. Expectations adjusted. There is no reference point for how much better things could be because the plant has never run any other way.
Third — and this is the one that is hardest to fix — the people who understand the technical problem and the people who control the capital decisions are rarely in the same conversation. An engineer who suspects recirculation is the issue may not have a clear way to quantify it in financial terms. A CFO who sees the energy budget creeping up does not know to ask about inlet air temperatures. The gap between those two conversations is where the problem lives.
What It Looks Like When You Fix It
The solution we built at ExhaustFlow Technologies starts from a simple premise: if recirculated hot air is the problem, bring in air from somewhere it has not been recirculated. Clean ambient air, drawn from outside the operational influence zone of the chiller array, delivered directly to the condenser intakes.
That is what EFT’s patented integrated base system does. It does not modify the chillers. It does not require operational downtime. It works with any major chiller manufacturer and it applies to both new facilities and existing ones.
The performance outcomes are consistent.
| What Changes | Measured Result |
| Chiller efficiency | Up to 30% improvement in kW/ton |
| Cooling capacity from existing equipment | Up to 25% increase |
| Electrical draw per chiller | Up to 19% reduction |
| Generator capital requirement | Reduced — lower full-load kW draw reduces generator sizing requirements |
| System stability | Fewer nuisance trips, reduced thermal stress on components |
| Equipment service life | Extended — operating closer to design conditions |
For anyone evaluating this from a financial perspective, the question is payback period. And the honest answer is that it depends on the size of the plant, the severity of the recirculation, and the local cost of electricity. That is why we offer project-specific modeling. Give us the inputs from a real facility — layout, chiller count, load profile, energy rates — and we produce a documented financial projection. Not a range. An actual number tied to actual conditions.
That is the only way to have an honest conversation about return on investment.
One Thing Worth Doing Before the Next Budget Cycle
Most annual budget processes start with the assumption that last year’s cost structure was roughly correct. Energy costs what it costs. Maintenance runs what it runs. If capacity is constrained, you add to it.
Before that process starts, there is one number worth getting.
You need to gauge/measure the air temperature entering your condenser coils during peak load. Not the weather station reading down the street. The actual temperature of the air the chillers are pulling in, at the intake, when the system is working hardest.
Compare that to the design ambient your equipment was sized for.
If there is a gap of more than a few degrees, some portion of your energy costs and your capital history reflects that gap. It has been there the whole time. It just has not had a name.
Now it does.
| FOR OPERATIONS AND FINANCE EXECUTIVES
EFT offers project-specific site evaluations that quantify the financial impact of air recirculation on your chiller plant — and document what recovery looks like. Contact us at info@eflowt.com or visit www.eflowt.com to start that conversation. |
A Final Thought
I want to be careful not to oversell this. Not every underperforming chiller plant is a recirculation problem. Some facilities do actually need more capacity for a variety of reasons. Some have equipment issues that are exactly what they appear to be.
But after 28 years, I can say with confidence that a significant share of the capital and energy costs that get attributed to demand growth or equipment aging are actually attributable to an airflow condition that has been present since day one. That is a fixable problem. It just requires someone to look in the right place.
Most of the time, nobody does.
Brian Workman is the CEO and Founder of ExhaustFlow Technologies, a patented airflow management company focused on air-cooled chiller plants in data centers and high-density commercial applications. He can be reached at info@eflowt.com or through www.eflowt.com.