VFD efficiency over 100%: A discovery of a perpetuum mobile?
April 1, 2022
Variable frequency drives (VFD) help to save energy on a system level. Also on component level the VFD has very high efficiency. But could the VFD efficiency exceed 100%??
As today is April 1, the Fool Days, we ask this provocative question. Despite of that the post is quite serious and does not aim to fool you.
VFD efficiency over 100%? The answer is YES and NO.
Actually the answer to whether VFD efficiency can exceed 100% is both “yes” and “no”. In a specific measurement setup the resulting VFD efficiency may be over 100%. Unfortunately is does not mean that we have finally discovered a perpetuum mobile. It only means that we have chosen a measurement method that is not suitable for the purpose of the measurement campaign.
Energy conservation principle
A perpetuum mobile, i.e. a machine that produces more energy than it consumes, would deviate the energy conservation law. The VFD is no exception and cannot work in perpetuum mobile regime. While the frequency conversion is extremely efficient, the VFD still produces certain losses with largest portion being associated with inverter switching and conduction losses, followed by rectifier losses, losses in busbars, filters, damping elements or the own consumption of cooling and control system.
The VFD losses are optimized to improve the life cycle cost, achieve best utilization of equipment, limit the temperature rise of components etc. However, some losses will always be present and physically the VFD cannot reach or exceed an efficiency of 100%. Realistic VFD efficiency at full load ranges from 97.0% to 99.2%. The value depends on the type of VFD, on the operation conditions, on motor parameters etc. The low end values around 97% are valid for VFDs with integrated transformers where the VFD losses include the losses of the transformer. On the other hand, very high-power drives with external transformer may exceed 99% efficiency mark.
Efficiency determined by measurement of input and output power
The real VFD efficiency will always stay below 100%. Yet, the measurements may show results exceeding 100%. Unfortunately, it is not a new breakthrough in the research and development. Instead, an unsuitable measurement method was used.
Surprisingly, when asking people how to measure the efficiency of a VFD, most of them may answer a method that measures the input and output power. On a paper this may work fine. It also sounds very straigh-forward. However, this method has one enemy: measurement inaccuracy or tolerances, respectively.
Each measurement device has certain accuracy meaning that there is a small error in the measurement. The error can be minimized by using quality equipment with high precision and thorough calibration. Yet, some remanent error is always present.
As the VFD efficiency is very high, even a small error has quite significant impact on the calculated efficiency.
Precise current and voltage transformers (CTs abd VTs) may be of accuracy class 0.2 meaning a 0.2% error. Getting anything better than this would be extremely difficult and costly. Moreover, the real error is likely higher.
- 0.2% accuracy class is usually given just for narrow frequency range (e.g. 45 to 65 Hz) while the VFD may be measured for other inverter frequencies. In those operation points the measurement error likely increases and is not guaranteed by the manufacturer of the instrument transformer.
- The waveforms of the VFD are not perfectly sinusoidal. They always contain some amount of harmonic distortion. For some VFD topologies the distortion is negligible while for others it is significant.
In practice, the power measurement contains a combined error of the voltage measurement, current measurement, error on the phase angle between voltage and current, error due to harmonic distortion and eventually also an error caused by operating the CT or VT outside of its specification. Summing up all those factors we can expect a total measurement tolerance in the range 0.5% to 1.0%.
As per above expressions, the measurement tolerance significantly impacts the efficiency determined in this way. Even though 0.5% combined measurement error sounds like pretty good achievement (and it actually is) it has a drastic impact on the calculated efficiency.
Table 1: VFD efficiency determined from input and output power measurement
In Table 1 we show an illustrative example. It is a 5’000 kW VFD that has 70 kW of heat losses at its rated load. That corresponds to an efficiency of 98.6%. However, when determining the efficiency based on input and output power we need to consider the measurement error on the power measurement. The table indicates the results for 0.5% and 1.0% combined measurement error (note that this is not the same as the CT/VT accuracy class). With 0.5% error the maximum VFD efficiency reaches 99.6%. That is a huge difference compared to the “real efficiency” of 98.6%. Shall the measurement error be 1.0% we could theoretically get an efficiency over 100%. Just on paper, of course.
Right method to measure VFD efficiency
There is a better way to measure VFD efficiency, at least for liquid-cooled VFDs. This method is based on the calorimetric principle, i.e. by measuring the losses evacuated via the liquid cooling circuit. IEC 61800-4, IEC 60146-1-1 and IEC 60146-1-2 refer to a segregation loss method that is proven to be much more accurate than determination based on input/output power measurement.
This method allows much better accuracy when determing VFD losses and efficiency. Even if the meausrement of VFD losses includes 10% error, the determined VFD efficiency is still much more accurate than calculation using input and output power. Segregated loss method will be described in our future post.
Take away
When determining an efficiency of a device or system it is important to choose a suitable method. In general, efficiency determination based on input and output power is inherently inaccurate due to the measurement tolerances. The higher the real efficiency the more this method suffers from the tolerances. The obtained results may be quite far from the reality – either much lower efficiency than the reality or in the other extreme an efficiency exceeding 100%. Real field experience confirmed that VFD efficiency over 100% can be measured with the input/output power method. Especially for VFDs with less good output waveform this method will fail to provide trustworthy results.
References
[1] What efficiency can you expect from your drive system? https://mb-drive-services.com/energy-efficiency-part-5/
[2] Inverter current higher than input current – perpetuum mobile? https://mb-drive-services.com/inverter-current-higher-than-input-current-perpetuum-mobile/
[2] Advantages of liquid-cooled drives, https://mb-drive-services.com/advantages-of-liquid-cooled-drives/
1 Comment
How to measure VFD efficiency - MB Drive Services · April 9, 2022 at 12:35 pm
[…] reported in [1], many people have the measurement of input and output power in mind when thinking of how to […]
Comments are closed.