The Complete Guide to Magnetic Bearing Chillers

What Is a Magnetic Bearing Chiller?

A magnetic bearing chiller is an oil-free centrifugal chiller that replaces oil-lubricated metal bearings with a MagLev support system, using an electromagnetic field to levitate the compressor rotor and eliminate the oil circuit, oil pump, and all routine oil maintenance. Commercially available units span capacities from roughly 80 tons to more than 2,000 tons across both air-cooled and water-cooled platforms.

As the U.S. Department of Energy’s Federal Energy Management Program summarizes it: digital, internally controlled magnetic bearings reduce friction and eliminate the need for oil lubrication. The compressor itself is usually delivered as a separate component to a chiller packaged equipment manufacturer (more on that in H2-4), which accounts for the underlying core technology being available from several different chiller equipment brands.

Whether an oil free magnetic bearing chiller in a water-cooled configuration or an air cooled magnetic bearing chiller for a smaller footprint, these units sit within the broader industrial chillers category and are promoted, at the package level, for energy efficiency and low maintenance costs versus their oil-lubricated counterparts. What this summary glosses over, of course, is what an engineer or facilities buyer of a water chiller most needs to know: How is this active magnetic suspension controlled in real time? What mechanically happens if active suspension control is lost? And how does its claimed performance compare to oil-based machines. This paper proceeds, layer by layer, through these questions.

Inside the Levitation Control Loop, How Magnetic Bearings Actually Work

The 5-Layer Levitation Control Stack

This section covers the magnetic bearing chiller working principle in mechanical detail, not just the marketing summary from H2-1.

Here’s the Five-Layer Levitation Control Stack we’re describing for the MagLev suspension in a magnetic bearing chiller: rotor position sensors continually update position data (position update to controller), a micro-controller makes calculations as to what’s needed, high current drivers then apply those commands to electromagnetic coils that establish the Levitation fields (control signal to power amplifier), a start-up safety system that requires the rotor to be successfully levitated before the rotor (drive motor) will engage, and a set of auxiliary, contact type bearings, that the rotor contacts if and only if the Active Levitation is lost (mechanical auxiliary bearings as failsafe). This isn’t hyperbole – an actual patent claims the exact same five-layers of protection. The patent documentation that claims the underlying control system actually covers the technology up to the power amplifier driver-i.e., at the component level – while complementary patents cover the physical bearing assemblies, even down to such issues as lamination and permanent magnet design and mold configurations to provide electrical insulation to coil windings.

The distinction with “the” auxiliary bearings warrants some clarification. Active system control in a magnetic bearing system typically doesn’t consist of either “dual-or-better” hardware with independent pathways-e.g. two controllers, two position sensor streams -nor does it consist of duplicated command controllers for each axis. On loss of magnetic support and levitation – be it from a sudden loss of power, failure of the active controllers, or a control algorithm-related error signal – the compressor simply contacts auxiliary metal bearings and decelerates. It’s a legitimate safety device, not an endless loop-avoiding system, and buyers should consider it from that point of view, not as an infinitely failsafe guarantee.

Since the Danfoss Turbocor compressor is the best selling and most OEM-ed magnetic bearing compressor on the market, much of what a buyer “interacts with” as to how a magnetic bearing chiller is “supposed to work” is actually how Turbocor does a Five-Layer Levitation Control Stack – which is just another reason that the OEM environment for H2-4 is more complicated than a superficial view suggests.

Magnetic Bearing vs. Screw vs. Oil-Lubricated Centrifugal, Choosing the Right Compressor Technology

Buyers often frame this as a magnetic bearing chiller vs centrifugal question, though the more precise framing (and the one this section actually works through) is magnetic bearing vs. screw vs. oil-lubricated centrifugal.

Magnetic bearing centrifugal chillers come out strong on part-load efficiency and oil-free maintenance value – not necessarily when it’s put up against the entire field of compressor technologies and delivering absolute lowest usage in all operating conditions – though modern variable-speed screw chiller technology has many applications where magnetic bearing chiller efficiency measurements can be achieved or surpassed. It’s a refreshingly honest presentation (better than what almost any vendor’s marketing material out there tells you), built on a side-by-side comparison on site rather than a published efficiency number.

An Oak Ridge National Laboratory evaluation performed under the U.S. General Services Administration’s Green Proving Ground program placed a variable-speed screw chiller and a magnetic bearing centrifugal chiller on the same water loops at the same facility, the Sidney Yates Building in Washington, D.C. Both were rated at an identical IPLV of 0.33 kW/ton on their datasheets. In the field, measured consumption came out to 0.623 kW/ton for the screw chiller versus 0.699 kW/ton for the magnetic bearing chiller, roughly 11% lower measured energy use for the screw machine on average, though measurement uncertainty widened that result to a range from 24% lower to 4% higher; a weather-normalized analysis found a smaller 3.4% screw-chiller advantage . Both units were independently rated more than 35% better than applicable FEMP efficiency criteria, this isn’t a story about one bad chiller, it’s a story about two efficient technologies performing closer to each other in the field than their marketing narratives usually suggest.

Where magnetic bearing technology does hold a clearer, more consistent edge is oil-free operation and part-load behavior, and the widely-repeated “friction-free equals dramatically more efficient” narrative oversells that edge. Independent industry commentary on oil’s actual thermal energy impact puts it at roughly 0.1-1% of heat-transfer efficiency depending on load, not the dominant factor marketing copy often implies . Screw compressors, meanwhile, avoid the surge instability that can affect centrifugal compressors at low part-load, and ASHRAE’s own handbook chapter on liquid-chilling systems notes screw technology can be economical in applications suited to its particular operating envelope .

In short: when operation on oil-free machines, part-load duty cycles and low vibration are paramount to your application’s success, and they’re not always about the most efficient rating on the datasheet, that’s when your call should lean toward a magnetic bearing technology. Always consider total cost of ownership over peak efficiency to make the decision between these two technologies.

Who Builds Magnetic Bearing Chillers? The OEM Compressor Landscape

Most magnetic bearing chiller brands don’t design their own compressor; they buy it from a small number of specialist OEM makers and integrate it into their own packaged chiller. Danfoss’s Turbocor line is the most widely adopted example: Smardt Chiller Group, Airedale, Heinen & Hopman, and Climaveneta all build around Turbocor compressors rather than designing their own.

That OEM Compressor Handoff means the “who builds the compressor” question matters as much as “who builds the chiller” when evaluating a spec sheet. Danfoss’s own OEM customer page confirms the handoff for those four brands, and Smardt’s smardt magnetic bearing chiller lineup is one of the more visible product lines built this way.

Not every manufacturer takes the OEM Compressor Handoff, though. Daikin explicitly positions itself as one of the few manufacturers that independently designs both its compressors and its refrigerants, and LG’s own two-stage inlet-guide-vane compressor design follows the same in-house path. On the newer end of the field, a 2022 Chinese-origin patent (CN114876825A) describing a sealed induction-motor chiller assembly with dual magnetic bearing assemblies and a high-speed semi-hermetic direct-drive motor confirms active, independent engineering investment in this compressor category outside the established Western OEM ecosystem , consistent with the broader manufacturer footprint in this category.

From the buyer’s point of view, what this means is that two chillers on a cabinet are identical under the hood-same basic compressor engineering, same underlying operating principle for the Five-Layer Levitation Control Stack, same basic failure modes-just have different brands on the box; the buyer might want to inquire into what’s under the paint when comparing cabinets. Koven Air’s own product range, magnetic bearing chiller, conveniently publishes a 25-model, model-by-model spec grid for the line, rather than a blanket efficiency rating range-for the very purpose that a buyer has actual numbers by which to comparison-shop.

Reading the Efficiency Numbers, AHRI 550/590, COP, and IPLV Explained

AHRI 550/590 IPLV is a weighted average of a chiller’s efficiency measured at four part-load points, 100%, 75%, 50%, and 25% load, not a single full-load number, and the weighting is fixed by the standard rather than tailored to any specific building’s actual load profile. ANSI/AHRI 550/590-2023 is the current I-P edition of the standard, with AHRI 551/591-2026 as the newer SI-unit equivalent, confirming AHRI actively maintains this rating methodology into 2026 rather than it being a legacy metric.

That fixed-weighting mechanic is the most common source of buyer confusion: a chiller with an excellent published IPLV isn’t guaranteed to deliver excellent real-world savings in a building whose load profile spends most of its hours away from the 50-75% band the formula weights most heavily. A high efficiency headline number is only useful in context. Cross-checking a magnetic bearing chiller’s part-load curve against the specific facility’s actual bin-hours data, not just its headline IPLV number, is the more reliable way to size an expected savings claim, which is exactly the gap the GSA/ORNL field study in H2-3 highlights when rated and measured numbers diverge. Jackson L. Ball, Group Vice President at Arctic Chiller Group, makes the same point from a sizing angle: the real efficiency advantage shows up at loads of 85% and below, so he recommends deliberately oversizing a unit (for example, a 350-400 ton machine for a 300-ton load) to keep it running in that part-load sweet spot rather than sizing tight to the peak load.

Where Magnetic Bearing Chillers Make the Biggest Difference, Data Centers and Beyond

The sweet spot for magnetic bearing chillers is facilities running at part-load conditions for most operating hours, with tight limits on vibration and noise, and an owner motivated to reduce energy use, cut building energy costs, and eliminate oil-based O&M spending — a profile disproportionately represented in data centers, hospitals, and cleanroom-adjacent manufacturing.

Data center cooling demand in particular has shown itself to be another concrete driver in increasing adoption – cloud providers and data center facility operators combined will commission more than 120 new hyper-scale and colocations across 2024 through 2026, a direct construct on how the keyword search for magnetic bearing chillers shows to trend flat to growing through the same period, instead of the expected boom and bust based on trends and keywords.

A district-energy best-practices case profile of the University of Cincinnati, distributed via the International District Energy Association’s resource library, documents exactly this pattern: three 2,400-ton magnetic bearing chillers installed at the campus’s East Utility Plant reduced annual electric operating cost by 28% and delivered 22% more efficient chilled-water production versus the plant’s historical performance, averaging 0.44 kW/ton (0.27 kW/ton at peak) and producing chilled water at 0.35 kW/ton over 52% of operating hours, a first-year summer energy reduction of 29% (8,284,500 kWh), as self-reported by the university’s own utilities team in the case profile a campus-scale example of the part-load-heavy duty cycle where magnetic bearing technology’s efficiency curve, not just its oil-free maintenance story, does most of the work. Facilities running closer to constant full load, by contrast, get proportionally less benefit from a technology whose real edge shows up specifically in the 50-75% load band the IPLV formula weights most heavily (see the worked example in H2-5).

New Build or Retrofit? The Replacement Decision Window

The question of whether to retrofit an existing plant with a magnetic bearing chiller, or to instead hold out for a new-build/end-of-life replacement, remains a conditional choice rather than an all-or-nothing decision, and the federal case data backs up both sides of that statement in the same dataset. DOE’s own Navy Techval program tracked three case-study sites: San Diego (a compressor-addition retrofit) made 40% in energy savings at 8.4 yr payback, Jacksonville, Fla. (a compressor retrofit, this one with condenser water reset) came in at 41% in savings at 7.0 yr payback, and Newport, RI (where a new chiller, not a retrofit, was installed) actually recorded best case results of 65% saving at a 3.8 yr payback. Retrofitting was successful in 2 of the 3 cases but required longer payoffs than the new-build project.

A different GSA/PNNL testbed evaluation for their site concluded, by contrast, that the economic justifications for retrofit failed since, given its age, their specific unit still had plenty of service life remaining. Viewed in isolation from one another, the cases aren’t necessarily mutually exclusive but are simply the same rule restated: the decision comes down to condition/lifespan of the existing evaporator and condenser, and ability of the plant’s capital horizon to accommodate either a 3.8yr or ~7-8 yr payback.

What Happens When Levitation Fails, Reliability, Backup Bearings, and Field Reports

If a magnetic bearing chiller loses active magnetic control, the rotor coasts down onto mechanical auxiliary bearings rather than the compressor experiencing an uncontrolled failure — that’s the fallback layer described in H2-2’s Five-Layer Levitation Control Stack, and it’s a genuine engineering safeguard. Whichever way the new-build-or-retrofit decision from the previous section lands, this failure behavior is one more reliability factor worth weighing against the remaining service life of the existing unit.

What’s less settled is exactly how those auxiliary bearings hold up under repeated real-world power-loss events: some field discussion among practitioners describes recurring board and TXV failures on multi-unit deployments and describes condenser-water-temperature sensitivity as a common installation mistake, but this comes from forum-level field reports that couldn’t be independently re-verified this round, treat it as a directional signal from practitioners, not a hard engineering spec.

That last point deserves more attention than it usually gets in chiller buying guides. DOE notes plainly that all IoT-enabled devices, and many current chillers ship with network connectivity for demand-response and remote monitoring, can introduce exposure to potential data breaches, and that security can almost never be added to a networked building-control device after the fact. For a facility already evaluating a magnetic bearing chiller’s control architecture in H2-2, it’s worth asking the same due-diligence question about its network architecture before the purchase, not after.

On maintenance specifically: even without oil to change, magnetic bearing chillers still need scheduled attention, sensor calibration checks, condenser/evaporator tube cleaning, refrigerant charge verification, and periodic auxiliary-bearing inspection remain on the maintenance calendar. “No oil” doesn’t mean “no maintenance,” and vendors that imply otherwise are overselling the technology’s real advantage.

Industry Outlook, What’s Driving Adoption Now

Data center cooling demand and a hard regulatory deadline on refrigerants are doing more to drive current magnetic bearing chiller adoption than market-size headlines alone suggest. On the regulatory side, EPA’s own Technology Transitions HFC Restrictions table sets a hard compliance date for chillers used in comfort cooling: refrigerants above a 700 GWP limit (rather than a low-GWP refrigerant alternative) can’t go into new chiller products or new chiller system installations on or after January 1, 2025, a hard procurement-timing constraint that applies regardless of whether a given facility is weighing a retrofit or a new install (see H2-7). DOE’s own chiller-purchasing guidance describes the same transition in broader terms and flags that buyers should confirm exact compliance dates for their specific equipment subcategory rather than assume a single blanket date, since EPA’s rule sets different dates for different chiller applications (comfort cooling vs. industrial process refrigeration vs. data center cooling, for example). Continued patent activity, including a 2022 Chinese-origin filing describing new sealed induction-motor magnetic bearing chiller assemblies, suggests the underlying compressor engineering is still being refined rather than sitting still as a mature, unchanging technology.

“Chillers, when a stand-alone product… Effective January 1, 2025, chillers for comfort cooling using a regulated substance, or a blend containing a regulated substance, with a global warming potential of 700 or greater.” — 40 CFR 84.54, the EPA regulation underlying the HFC restrictions table above

On market sizing, for background magnitude only: one industry aggregator, Dataintelo, suggests the global magnetic bearing chiller market is in the high single-digit billions of dollars with a high-single-digit-percent compound annual growth rate through the early 2030s, presented here as one named source’s rough order of magnitude, not a load-bearing fact or a cross-report consensus figure, since market-research aggregators commonly disagree on both precise numbers and segment scope for this niche category. The more concrete, verifiable driver behind current adoption and real energy savings on the ground is the data center buildout referenced in H2-6 and the refrigerant-transition deadline above, not any single market-size estimate.

Buying Checklist, What to Ask Before You Sign the Spec

Prior to even submitting the order specs, and with the adoption drivers covered in the previous section in mind, a buyer ought to not simply respond to efficiency specifications and price, and must inquire into compressor design considerations, field part-load operation, time-tables for complying with refrigerant regulations, and cybersecurity or network-security features of the proposed system or devices to be sourced. The checklist below is built to be copied directly into a request-for-quote so a vendor’s answers can be compared apples-to-apples.

FAQ