Energy efficiency – Part 4

Just two important reminders before starting with the topic:

1) Always keep in mind that the energy savings come due to variable speed control. The load consumes significantly lower power when the speed is reduced and the process is such that it does not require full speed all the time.

2) Compare only apples with apples. The efficiency shall be clearly defined to make a fair and true comparison possible.

Variable frequency drives and drive systems generally have high efficiency. Yet, there are some differences. One should be careful when comparing the efficiency and losses of two different drive systems. Here are some rules:

Focus on system efficiency rather than efficiency of individual components

Comparing just system components might be misleading. And such comparison is also not much relevant. For the user the system efficiency is what counts in the end. Let’s illustrate it on three simple examples:

Example 1:

One drive system has a VFD with output sine filter, the other VFD has just a small dv/dt filter. The first has additional losses in the sine filter and might need a higher switching frequency as well. No surprise that the efficiency is somewhat lower. The other VFD has marginally better efficiency. However, the first one has a clean output waveform. The second one has more harmonics causing additional losses in the motor, i.e. the motor will have lower efficiency. Which system looks better from system efficiency view? That is difficult to generalize. But you probably got the point. Comparing just those two VFDs would not be correct.

Example 2:

One drive system consists of a multi-winding transformer supplying a diode rectifier based converter. The other drive system has a transformer-less (direct to line) converter. The first system has approx. 1% losses in the input transformer. The second system eliminates the transformer, but requires an active rectifier and probably also an input reactor for decoupling from the grid. The additional rectifier losses and losses in the input reactor could be of similar magnitude like the input transformer.

Example 3:

One drive system is conventional with 4-pole motor and speed increasing gear. The other drive system is gearless high-speed. The high-speed motor has a bit lower efficiency due to higher friction losses and typically lower power factor. Also the VFD supplying the high-speed motor tends to have somewhat lower efficiency. So the outcome would be that high-speed motor drives are less efficient? No, this is not quite true. The high-speed drive eliminates the losses in the gearbox so on a system level both solutions will be very similar from efficiency point of view.

Check how the efficiency is defined. Use the same definition when comparing two drive systems

As the efficiency is generally high, the different definitions of efficiency impact the results.

–  Does the efficiency include auxiliary power consumption (cooling fans and/or pumps, control power) or not? If the auxiliaries are not included, the efficiency improves by 0.1-0.3%, in extreme cases even more.

– Is the efficiency calculated based on project rating or based on VFD capability? Most VFDs are standardized in specific frame sizes. If your drive is rated for 7’500 kW you might get 10’000 kVA or 12’000 kVA VFD. Does the efficiency refer to 7’500 kW motor rating (with corresponding apparent power) or does it refer to max. VFD capability? It will make a slight difference as some losses are (almost) constant regardless of the actual loading.

– Are the harmonic loses included in the calculation or not? This is a question for the transformer and motor. In the IEC world the transformer losses are commonly given as losses for sinusoidal waveform. The effects of harmonics are of course considered for dimensioning, but the no-load and load losses in the datasheet refer to a sine wave. The background is that the harmonic losses usually cannot be measured in the test field (no source of harmonics). For motor the practice is similar. Sometimes the additional harmonic losses are provided in the datasheet as informative value or the manufacturer states two efficiency values – for sinusoidal supply and for inverter duty. Therefore, if you get a figure of drive system efficiency, you shall know whether the harmonic losses have been considered or not.

The best way is to make a clear definition, e.g. guaranteed input power for given load points. In such case it is clear that the power consumption must include all losses including harmonic losses. If the measurement is only done in medium voltage then the figure is without auxiliary consumption. If low voltage consumption is measured as well then the figure shall account for all losses including auxiliary consumption.

Guaranteed versus typical values

If efficiency matters to you then you shall always ask for guaranteed values. ‘Typical’ gives lot of room for different interpretation. Also the manufacturers have different approaches. Some of them are conservative and tend to provide lower efficiency than what is finally measured (you may ask for a sample test report of comparable equipment already manufactured). Others take more risk and promise very high values. Penalty for excessive losses shall discourage less honest manufacturers. In case of performance measurements the accuracy of these measurements shall be considered and high accuracy class measurement equipment shall be used.

Ask for efficiency figures for multiple load points / for operation range

VFD will not operate at one fix speed otherwise it is not a VFD (exceptions always exist, but those cases are very rare). Therefore the user is interested in efficiency figures for various load points. Ideally these load points correspond to the expected operation point or guarantee points. Having the efficiency figures for the load points, the OpEx, energy savings and payback time can be calculated more precisely.

Verify the efficiency by measurements

If technically feasible you can verify the efficiency figures by site measurements. For liquid cooled VFDs such measurement is relatively simple using the segregated loss method, for air cooled VFDs it is less accurate. However, having such measurement in the contract will again discourage dishonest manufacturer and helps to you get what you have specified, ordered and paid for.