Transcript
Introduction to Supraharmonics
Welcome to today’s Ask a Pro webinar. My name is Landon Rhodes. I am an engineer here at Power Monitors, Inc. and I’m going to be talking about my white paper on supraharmonics.
On the grid as EV chargers, photovoltaic inverters, variable frequency drives, LED lighting systems have come onto the market, have been developed and are increasing in popularity, we have seen an increasing rise in a new power quality issue called supraharmonics.
This is defined to be any quasi-stationary distortion on the voltage or current waveform that happens in the frequency range of 2 kilohertz to 150 kilohertz. The quasi-stationary clause excludes transient behavior because transient behavior is different, and it should be treated as a different cause with a different effect, with different sources.
This sits in between the bands of the traditional harmonic distortion measurements and the CISPR 16-1-1 frequency band B. That’s the bound that this two kilohertz to one hundred and fifty kilohertz is defined between. Traditional IEEE 519 records usually only go up to about two or three kilohertz, and this frequency band B deals a lot with conducted emissions, and this band covers 150 kilohertz to 30 megahertz, which is the next band up.
Properties of Supraharmonics: Lumped Element Model
One of the main things that makes supraharmonics interesting is that an electric feeder in an electric grid, you can’t treat it as a lumped element. The wavelength of a wave is given by lambda equals the speed of the wave divided by the frequency of the wave. I am assuming that the speed of the wave is two-thirds C. It varies somewhere between two-thirds C and C, the speed of light. So this is about between two hundred million meters per second and three hundred million meters per second.
If a wire is less than lambda over ten, it is less than one-tenth of the wavelength, then it is a very reasonable approximation to treat it as a lumped element model, where every single point on the wire is assumed to be the same potential, same voltage potential. If the wire is greater than lambda over ten, that is not a reasonable assumption anymore.
I have a table here that gives frequency, wavelength, and the lambda over ten lumped element cutoff. So for sixty hertz, the lumped element cutoff is three hundred and thirty kilometers, which is very, very, very long compared to any feeder length. So at sixty hertz, an electric feeder is essentially a short circuit. There’s no difference in potential across the whole feeder.
I have multiple rows: sixty hertz, two kilohertz, six kilohertz, eighteen kilohertz, fifty kilohertz, and one hundred and fifty kilohertz. And the table describes how the lumped element cutoff drops dramatically from ten kilometers at two kilohertz, all the way down to one hundred and thirty meters at one hundred and fifty kilohertz. So this range, an electric feeder in the small number of kilometers range or shorter, these numbers start to matter a lot.
Traveling Wave Demonstration
I have a Desmos graph here that describes a traveling wave. T is my time, and as you see, the wave is traveling to the right at a constant speed. I am showing two wavelengths of the period, and you can see the whole sine wave across this X-axis. W describes the number of wavelengths that I’m seeing.
If I show one wavelength, you still see the full waveform. Now, if I’m gonna do one-tenth wavelength as recommended, if I stop the time, you still see some difference from the start to the end, but it’s relatively low difference. So for example, here, the end ends at one, whereas it starts at around minus point seven five, and it ends at minus one. So it’s relatively consistent across the whole range.
This is about the cutoff where wave effects start to stop mattering quite as much. If I decrease this to one one-hundredth of the wavelength, then you see it is really, really constant across the full range. And this is what a sixty hertz waveform would look like on a feeder. You would have the start here, and you have the end of the feeder here, and they’re essentially the same potential everywhere along the feeder.
Whereas if you show only two wavelengths, this is what makes supraharmonics different than regular harmonics. Because the assumption that every place on the feeder is the same, because that assumption breaks down in the supraharmonic range, you start to see a lot more effects like reflections, standing waves, interference, refractions, attenuations. You start to see some weird wave physics effects. You have to treat this as a wave all along the feeder, and you can’t treat it as just because there’s a single potential at one point in the feeder, there must be the same potential everywhere else on the feeder.
Sources of Supraharmonics
The next section in my paper talks about where supraharmonics come from, and there’s two main sources. The first is electronic power converters, and the other main source is electromagnetic and power line communications.
Electronic Power Converters
The first source is electronic power converters, where you have a switching element that switches at somewhere between fifty kilohertz to maybe a hundred kilohertz or a little bit higher. It varies. It depends on the manufacturing, depends on the device.
In order to comply with traditional harmonic distortion metrics, what manufacturers have done is bump the switching frequency well into the tens of kilohertz. When you push the switching frequency well into the tens of kilohertz, that doesn’t show up in the normal harmonic distortion metrics where you only are concerned up to about the 50th harmonic of a 60 hertz waveform. So traditional distortion looks really clean, but you end up with distortion even though you’re not traditionally measuring that. That’s what the supraharmonics measurement is all about.
A switching power converter can either show up as a stable frequency peak where the device is switching at a particular frequency all the time. It can also look like a narrow band spectrum of peaks where the device will change its switching frequency over time in order to spread the switching energy across a frequency range. It can change the switching frequency in a continuous way so that the peaks don’t show up quite as substantial. They can spread the energy out over a frequency range. That’s what this frequency dithering is all about.
Power Line Communications
The other main source of supraharmonic effects come from power line communications. Substations can communicate with different devices along a feeder, like a recloser, like smart meters, things like that, cap banks. And the way they do that is they superimpose a communication signal on top of the 60 hertz waveform.
According to this paper, on waveform distortion and the frequency range of two kilohertz to 150 kilohertz, switching converters are generally much lower intensity than power line communications, but are often continuously present. They don’t change in amplitude. Whereas PLCs, power line communications, they often have a small burst where they’re high intensity, high distortion, and then it’s a long time where there’s nothing happening. You often get more damage because of power line communications, but you often get more distortion because of the switching converters, because they’re on more of the time.
So supraharmonics come mainly from switching power converters. That’s EV chargers, photovoltaic inverters, variable frequency drives, LED lighting systems, things that take a 60 hertz waveform and convert it to DC or take a DC and convert it back to a 60 hertz waveform.
Effects of Supraharmonics
My main source is this paper, the effects of supraharmonic distortion in medium voltage and low voltage AC grids. Here are some negative effects that they list:
- Power losses in conductors
- Aging of insulators
- Aging of capacitors
One particular effect is the skin depth is inversely proportional to the square root of the frequency. So as your frequency rises, your skin depth drops. This describes where most of the energy in the current is only in the small outer shell of a wire. Instead of the whole wire conducting energy, it’s only really the outer shell of the wire. The skin depth starts to matter at supraharmonic frequencies. You start seeing wires that are effectively higher resistance because only a small fraction of the wire is actually conducting the current.
You start to see higher resistance. You start to see breakdowns in insulating materials. You see breakdown in capacitor dielectrics, like in cap banks. You see terminations between conductors in different clips that can degrade faster and fail earlier when you have supraharmonics present.
EV Charging Beat Frequency Example
This paper references a case where several EVs were charging from a single circuit, and there’s an audible beat frequency. A beat frequency is where you have multiple sine waves that are very close in frequency, and you would hear the difference in the frequency. So if one is 50 kilohertz and the other one is 50.5 kilohertz, you would hear a 500 hertz tone happening. That’s what this described when several EVs were charging from the same circuit with similar switching frequency characteristics.
Categories of Interference
I referenced a section in the on waveform distortion paper by Sarah Ronberg. She describes four different categories of interference:
- A device is occasionally not operating as intended
- Failed operation or damage
- Interference with power line communications
- Audible noise
I talk about a 77.5 kilohertz radio control signal that is messed up. A copy machine had underperforming effects, like it couldn’t communicate. Some light dimmers were underperforming or misoperating. This is a serious problem. There are serious effects that supraharmonics can cause.
The first category with device misoperation is the most widely reported. There’s not as many examples of device failure yet, but device misoperation where something receives the wrong control signal. A light dimmer usually works through something like pulse width modulation, where it turns on and off the signal at a certain frequency, and then it can be confused by something like supraharmonics. Medical equipment can misoperate. It can receive incorrect signals. Electricity meters can be confused or stop working correctly. Super harmonics can be a serious issue.
Why Existing Instruments Don’t Observe Supraharmonics
That essentially has to do with the IEEE 519 recommendations for harmonic distortion measurements only recommending measurement up to the 50th harmonic, which is about three kilohertz. And that is recommending measurement based on the IEC 61000-4-30, which is electromagnetic compatibility, testing and measurement techniques for power quality measurement methods.
According to those standards, most ordinary distortion is synced to the 60 hertz waveform, which means that the 60 hertz waveform causes some effect. Very little distortion happens outside of that when the source of the distortion is the 60 hertz waveform interacting with some nonlinear load. Most of the harmonics usually die out by the 50th, and so traditional standards only recommend measurement up to the 50th.
So that’s how most power quality measurements record, and if you only record up to the 50th, then all of the switching devices, like all of the inverters and all of the variable frequency drives, which behave independently to the 60 hertz waveform, all of those are sort of invisible.
The Seeker+
At Power Monitors, we have been working on a device and we recently released it, which is called the Seeker+. This is our newest product, newest power quality meter. This measures supraharmonics, and it measures it according to NXC and NXD of the IEC 61000-4-30 standard. NXC talks about functional design for measurements in the two kilohertz to nine kilohertz range, and NXD talks about nine kilohertz to 150 kilohertz.
We follow these standards. We sample the voltage at one megahertz, and we continuously compute supraharmonics from those samples in 200 hertz wide bins between two kilohertz up to 150 kilohertz. So we have one centered at two kilohertz, one centered at 2.2 kilohertz, one centered at 2.4 kilohertz, et cetera, all the way up to 150 kilohertz.
Spectrogram Display in PQ Canvass
This data is available in PQ Canvass as a spectrogram. A spectrogram is a cool graph format where you have time on the X-axis, and this is a 24-hour recording roughly, and the Y-axis is frequency. We start at near DC, and then 20 kilohertz, 40 kilohertz, 60 kilohertz, 80 kilohertz, 100 kilohertz, 120 and 140 kilohertz on the Y-axis. X-axis is time, Y-axis is frequency, and then inside the graph is intensity demonstrated by color. Brighter colors mean more prevalence, and darker colors mean less prevalence of that frequency.
You can see in this recording we have a broadband signal in the upper range between 130 kilohertz and 150 kilohertz, so that’s kind of broadband and low energy signal across the whole recording. And then we have several distinct bands. We have one band at 80 kilohertz, one band at 60, one band at probably 55, one band at 50, one band at 45, and one band at 40. And then we’ve got a few lower energy bands below that.
These are very consistent from 40 kilohertz to 60 kilohertz. Those frequencies are very, very consistent in this recording. The one at 80 kilohertz is oscillating up and down a little bit. It changes in time. And then the one centered at 150 kilohertz, which is the broadband low energy signal, that’s very consistent over the whole recording. And then we see a distinct valley at 100 kilohertz, so there’s almost no 100 kilohertz energy, and there’s almost no energy below 20 kilohertz in this recording. So this is what the Seeker+ can do. It can record supraharmonic data, and it can display it in a cool spectrogram format.
The Need for Supraharmonic Measurement
Because of all of the power electronics that we’ve put on the grid, like EV chargers, like inverters, also servers, computer chargers, iPhone chargers, all of those contribute a little bit of this supraharmonic energy to the grid. Because these have grown in popularity, emissions in this band have also grown. But the tools that we need to understand and observe and make corrective actions have not kept up, partly because it’s more difficult to characterize signals that aren’t consistent over the full feeder. If you measure something at one point in the feeder, measure it at a different point in the feeder, they’re necessarily gonna be different at these frequencies.
It is a sort of a challenge to see what the proper standards are, and they’re starting to become more and more of a problem, and the analysis tools and the standards will eventually present a cohesive standard. But while this is taking place, we need more data, and the Seeker+ gives you that option to measure supraharmonic frequencies, and it does it in accordance with the IEC 61000-4-30 Edition 4 standard that was recently published in October of 2025. We follow the standard for measurement and for the sampling and the binning and everything.
While the compatibility levels and emission limits are still being considered and decided upon, now is the crucial time to make frequent supraharmonic measurements throughout the grid and contribute to the understanding of this effect. The Seeker+ allows users to measure, to analyze, to understand supraharmonic data that’s happening on your grid.
Thank you for attending this afternoon’s webinar. You can submit questions to support@powermonitors.com, and we will get back to you as soon as we can. Have a great day.