There's no magic bullet for data center cooling. I learned this the hard way.
When I first took over purchasing for our facilities—managing orders for a mid-sized firm with a growing IT footprint—I assumed you just picked the biggest HVAC brand and called it a day. Turns out, the 'best' solution depends entirely on what you're working with. A retrofit in a 20-year-old building? That's a different animal than a new build with a blank slate.
Over the last few years, I've processed over 200 orders for cooling equipment across three different facility upgrades. I've seen what works, what doesn't, and where a specialist like Munters (whose data center cooling solutions we've used) makes sense versus where you're better off with a traditional chiller.
Let's break it down by scenario.
This is the most common situation I run into. You have an existing server room or small data center, originally cooled by a standard CRAC (computer room air conditioner) unit. But as you've added density—more servers, more load—that unit is struggling. Your hotspot map looks like a target board, and the facilities team is constantly dodging alarms.
In this scenario, munters data center cooling solutions based on indirect evaporative cooling or precision dehumidification can be a game-changer. Why? Because you don't have the ductwork or floor space for a massive retrofit. A Munters system can often be brought in as a sidecar, handling the latent load (humidity) while your existing gear handles sensible load (temperature).
Real talk: We had a situation in 2023 where a direct expansion (DX) unit was struggling to keep humidity below 60% RH. Our server vendor threatened to void the warranty. We looked at options: a new, larger CRAC unit would have required pulling new power lines and modifying the floor plan. The Munters dehumidifier HC-150 was a drop-in solution next to the existing unit. Did it take a month to dial in the settings? Yes. But it cost a third of what the full CRAC replacement would have.
When to go with Munters here: You need precise humidity control without ripping out existing equipment. You have moderate heat loads (up to ~100 kW).
When to look elsewhere: If your heat load is massive (300+ kW) and you need to rip and replace everything anyway, a modern chilled water system with variable speed drives might offer better efficiency. I wouldn't try to patch a Munters unit onto a system that's fundamentally undersized.
"The vendor who said 'this isn't our strength—here's who does it better' earned my trust for everything else." This applies to gear as much as suppliers. A good cooling specialist will tell you if your situation calls for something else.
This is the dream. You have a clean room, full control over the design, and a budget that was approved for a five-year plan. In this case, the conversation shifts from 'what can we patch in' to 'what's the long-term OPEX.'
This is where I've seen the energy efficiency of munters data center cooling solutions really shine. Their indirect evaporative cooling systems—like the Oasis™ series—can achieve a PUE (Power Usage Effectiveness) that beats traditional chillers in the right climate. But—and this is a big but—it's climate-dependent. If your data center is in Arizona, the evaporative effectiveness drops in summer. If you're in Seattle, it's fantastic year-round.
In our 2024 new build (400 employees, three locations), we spec'd a Munters hybrid system: an indirect evaporative cooler for most of the year, backed by a small DX coil for peak heat events. The projected savings vs. a full chiller plant? About 35% on annual cooling energy, based on a cost review I did in Q1 2024.
When to go with Munters here: You care about energy costs and PUE. You're in a climate with moderate humidity and decent 'dry bulb' temp drops (i.e., not a tropical rainforest). You have a facilities team that can handle some system complexity.
When to look elsewhere: You're building in a high-humidity market and need absolute reliability without any 'free cooling' window. Honestly, in places like Singapore or Houston, Stulz or a traditional chilled water system might be a safer bet. The Munters system works, but the number of hours in 'DX backup mode' eats into your ROI.
Let's be honest: not every situation warrants a six-figure engineering solution. Sometimes you just need to get a rack of equipment to stop overheating without calling your CFO for a capital request. This is where you end up thinking about things like a dewalt blower to move hot aisle air—which I've done in a pinch (it worked for 48 hours until the real unit arrived).
For these temporary or low-density situations, the conversation is different. You're not buying a cooling system—you're buying time.
I've used a Can-Am air filter rigged to a standard fan to create positive pressure in a dusty room. I've borrowed an electric snow blower (dead of winter, okay?) to blow cold outside air through a filter box into a server closet. These are MacGyver solutions. They're not best practices. But they kept systems online while the PO for a proper Munters unit went through approvals.
When to go with MacGyver fixes: Immediate emergency (hours), small closet, temporary load reduction.
When to absolutely not: For any production environment with live customer data. The risk of dust, voltage spikes from a blower motor, or thermal shock is not worth the $200 you saved.
This isn't about guesswork. Here's a simple litmus test I use before calling any vendor:
Look, I'm not a mechanical engineer. I'm an admin buyer who has ordered a few hundred thousand dollars worth of cooling gear. But after 5 years of managing these relationships, I've come to believe that the 'best' vendor is highly context-dependent.
For the situations where you need precise humidity control, energy efficiency, and a specialist who owns their niche—a Munters dehumidifier HC-150 or an indirect evaporative system is a solid call. For the times you just need airflow? A generic fan will do until the real gear shows up. Know your limits. That's the professional way.