DIRECT-DRIVE TFM MOTORS FOR DATA CENTER HEAT REJECTION SYSTEMS

Data center heat rejection is under pressure from four directions at once. Water restrictions are tightening. Energy budgets are scrutinized down to the third decimal of PUE. Liquid cooling is changing what the heat rejection plant has to do. And noise has become one of the leading sources of community opposition to new sites.

Our new white paper, Direct-Drive Transverse Flux Motors for Data Center Heat Rejection Systems, looks at what these pressures mean for the fans that move the air, and for the motors that drive those fans.

The water and energy tradeoff

Evaporative cooling consumes one to two liters of water per kilowatt-hour of IT energy. Dry cooling removes that water use, but air alone carries far less heat, so a dry system has to move 3 to 5 times more air to reject the same load. More air means more fan power, and that power counts directly against PUE.

Why fan size matters

The paper shows that the most effective way to move that much air is with fewer, larger, slower fans. A larger fan at lower speed moves the same air at lower pressure. It wastes less energy as discharge velocity and makes less noise, because fan sound power rises steeply with tip speed. Fewer fans also means fewer motors, drives, wiring runs and maintenance points.

The barrier has been torque

Large, slow fans need very high torque at low speed. Until now that has meant either a gearbox, with its efficiency losses, maintenance and tonal noise, or a conventional direct-drive motor that becomes large, heavy and costly. This is the constraint that has kept fan cells small.

What a transverse flux motor changes

A transverse flux motor (TFM) separates the electrical and magnetic circuits, which allows a high pole count without giving up copper area. The result is direct-drive torque at low speed, at one-half to one-third the weight of conventional machines.

What it means for a 1 GW campus

The paper compares heat rejection architectures for an illustrative 1 GW data center. Against a dry-cooled baseline of 9-foot geared fans:

• Moving to 24-foot direct-drive TFM fans reduces annual cooling energy cost by 43%, from $15 million to $8.6 million.
• The number of fan drives falls by more than 70%, from 608 to 175.
• A 30-foot TFM configuration matches the PUE of evaporative cooling with zero operational water use, with a larger heat rejection footprint as the tradeoff.
• Gearboxes are eliminated in every TFM scenario.

The paper also covers the fan affinity laws behind these results, reliability and redundancy with fewer fans, noise sources and ordinances, and the motor physics that make TFM a fit for continuous fan duty.