Beware: high-efficiency motors may not be reliable

ABB is calling upon the electric motor industry to be more transparent with its technical motor data by making available information that gives a clear indication of motor reliability - such as running temperature at full power and full speed - alongside efficiency levels and motor noise. While Steve Ruddell, ABB's general manager for motors and drives in the UK, acknowledges that the main reason why motors fail is through winding breakdown, closely followed by bearing failures, he believes that, all too often, customers accept these failures without understanding the root cause.

"In our experience, we believe that many of these winding and bearing failures are a direct result of motors running too hot," says Ruddell. "For example, you may be told that your bearing has run dry. While in some instances this may be down to a poor re-greasing regime, it is also possible that the motor was simply too hot and the grease degraded prematurely."

While efficiency classifications, as defined by Eff1, Eff2 and Eff3, have helped customers recognise the difference between a poor energy-efficient motor and a high energy-efficient one, ABB is concerned that the classification is also being assumed to be an indicator of the reliability of a machine. Today Eff1 is perceived as high quality; the belief is that you pay a higher price for a higher efficiency motor and that the certification also implies high quality and high reliability.

"This is just not the case," argues Ruddell. "Our experience shows that there are some motors which achieve Eff1 status at the cost of significant drawbacks. These manifest themselves in many ways - increased running temperatures and excessive noise being a couple of examples."

Ruddell proposes that 'reliability' could be defined as the sum of efficiency plus temperature use. He argues that each of these elements directly affect each other and, therefore, the quality and reliability of a motor. And MTBF data is not much help, as this probably assumes the motor is running in 'ideal' conditions, which is seldom the case in real-world applications, where premature failures can easily be caused by, for example, not increasing the bearing re-greasing frequency in line with increased running temperatures.

Misleading efficiency claims

Eff1 is easy to achieve in all but the smallest frame sizes by simply increasing the amount of active material in the motor - more copper in the slots and smaller air gaps in the design. Yet the challenge is that the IEC 34-2 standard sets tolerances for efficiency that are quite wide. ABB fears some manufacturers are declaring efficiencies that are at the top of the tolerance band, while delivering motors close to the lower tolerance level - or, at worst, outside this band.

Meanwhile, the temperature use of a motor can also be higher than would be expected from an Eff1 motor. "Hold a thermal camera up to the motor and you can see how hot it is running."

In the case of the windings, higher temperatures degrade the winding insulation more rapidly; for the bearing end shields, higher temperatures can lead to premature degradation of the grease and an increased need for re-greasing.

"Clearly, the cooler the running temperature the better," says Ruddell. "Lowering the temperature by just 10 to 15 degrees C can double the life of the winding, yet most catalogues do not give this specific information."

ABB has observed that some cheaper motors are generating more heat. The normal frame surface temperature in a high-reliability motor running at full load can be as low as 60 to 80 degrees C. "Yet lower reliability motors often run in excess of 90 degrees C and have even been recorded at well over 100 degrees C."

The importance of temperature

Temperature affects the re-greasing intervals of bearings. ABB assumes that bearings will run at 80 degrees C based upon an assumed ambient temperature of 25 degrees C. Should the bearing temperature increase by 15 degrees C, then the re-greasing interval should be halved. If the temperature decreases by 15 degrees C, the interval can be doubled. "The problem is that this data is not always highly visible, so many engineers are probably not aware of this fact" says Ruddell. "As a result, if a catalogue states 10,000 hours and the temperature increases by 15 degrees C, then the bearing would need re-greasing in 5000 hours. But how is a user meant to know this? End-users with continuous process applications should ask their supplier to provide the winding and bearing temperature criteria from the type test reports."

Temperature can lead to problems in other areas too. More active materials usually mean more heat being generated. To keep the motor within stated temperature limits, larger fans are employed to provide more cooling air. And larger fans mean more noise.

"Remember, a 3dB increase in noise level is equivalent to a doubling of the audible noise of the motor," says Ruddell. "So if, say, a 200kW motor is showing a 77dB noise rating against one showing 70dB, then the 7dB increase in noise equates to the motor being about four times noisier. This should set alarm bells ringing. Higher noise levels could mean that the temperature is higher, which affects the overall motor reliability."

ABB says that getting the right balance between efficiency, temperature rise and noise will go a long way towards lowering life cycle costs through reduced running costs and increased overall reliability.

"If you use motors in a 24/7 continuous process, then the last thing you need is a hot motor, which eventually fails. The cost to production can be immense. We believe that these users should have access to temperature rise information and should understand the importance of the noise levels. This would help them choose a motor with greater reliability."

Remember, however efficient a motor is when it is running, a premature failure - and the resultant unplanned downtime - could be so costly as to make the question of efficiency immaterial.

Ruddell and his team at ABB are presently gathering statistics and data and are keen to hear from any end-users that believe that motors have failed prematurely through temperature issues, or from users interested in learning ways to identify a 'reliable' motor.