Transcript

Introduction to Ferroresonance

Good afternoon everyone, and welcome to today’s white paper webinar. Today, we are gonna talk about ferroresonance. This is a very interesting issue. It can be a power quality issue, it can be something that can damage equipment very quickly, and it’s a tricky one because the damage can happen without anyone measuring it.

It’s hard to prove after the fact that ferroresonance is indeed what caused the problem. So it’s one of those things that you’re left with something that is broken or failed. You got fuses blown, you got transformer damaged, but you don’t really have any signs as to what that problem was.

What Is Ferroresonance?

When we talk about ferroresonance, what we’re talking about is a resonance between external capacitance and the inductive core of a transformer. Now, the inductive core of the transformer is normally inducted with a fixed inductance, but if the transformer becomes saturated, that inductance will vary.

As that transformer saturates, it can form a feedback loop. So as that inductive changes, of course, you have an LC resonance formed by that transformer inductance and the capacitance. But if that inductance is varying with time, you’ll form a varying resonant frequency with time and you can get a positive feedback loop where you form an oscillator and that transformer starts oscillating.

Three Ingredients for Ferroresonant Oscillation

The ingredients that have to be there for a transformer to kick into ferroresonant oscillation are… You only need three things:

1. You need the iron-core transformer, whose inductance changes as the core saturates.
2. You need a capacitance in series to form the other half of that resonance.
3. You need no damping or very light damping.

With just a little bit of load or resistance on that transformer, that will provide the damping that keeps ferroresonance from either happening or getting out of control. But without that, you have the elements of a positive feedback loop, which causes oscillations.

The Effect of Resistive Load

It really doesn’t take much capacitive damping. Here, we see in this example circuit, we have just a little bit of resistive load, we can really dramatically change the picture. As that transformer saturates, you’ll have very high voltage that develops across the transformer winding, and this load resistance really knocks that down quite a bit. So it doesn’t take much load to damp out that ferroresonance.

Sources of Series Capacitance

So where does this capacitance come from? Generally, series capacitors cause this, and I’m gonna jump from the white paper to a different presentation too that gives you a little bit more detail on ferroresonance. The white paper really comes somewhat from this, but here we have a schematic of that sort of non-linear effect, where you have the transformer core and you have the series capacitance, and as that inductance changes versus time, you have that oscillation effect.

You have the frequency that varies versus time and that positive feedback loop. And you see here on the graph that’s shown in the lower right, phases A, B and C are the transformer output, and phase C is the normal 60 hertz voltage. Phases A and B are the transformer ferroresonance and we’ve lost 60 hertz altogether. We performed an oscillator where the oscillation happens between energy in the transformer core and this external capacitance, and it’s isolating as that resonant frequency changes.

So the peak voltage you can achieve here is four, five, six times the normal peak voltage, and this will persist until something fails. And again, similar to the paper, it doesn’t take much resistance to dampen this out. So in this model, there’s no resistance at all on that transformer core. The only resistance is in series with the capacitor.

Only a small amount of load resistance, in some cases just 1,000 watts or so on a transformer, will be enough to really dampen the voltage that’s developed across the inductor, which stops ferroresonance from happening. Here in this example, the numbers on this schematic, we drop down from 60 volts down to 10 volts with just a small amount of load resistance.

How a Transformer Enters Ferroresonance

What happens when a transformer goes into ferroresonance is, the transformer is lightly loaded, some sort of switching event happens. This could be when the transformer is energized one phase at a time, could be when you have a single-phase fault or single-phase open line recloser opens or single-phase fuse blows. Somehow, a switching event happens, and now you have a high-Q resonant circuit with the transformer inductance and the cable capacitance together and very little damping resistance.

So with the resistance low, it doesn’t really have a current limit in there, and so the voltage across the transformer increases. As that voltage increases, that saturates the core. So when you have that high-flux and the core saturates, that’s where you have that non-linearity that now turns into an oscillator. So now you have distorted waveforms, you have kinda chaotic voltage output when you have all these happen in that sequence.

Underground Cable and Cap Bank Capacitance

So you wanna have load, and again, this capacitance can come from underground cable capacitance. Underground cable has a lot more capacitance per foot than overhead lines, and if that’s directly in series with the transformer itself. Cap banks that are power factor correction cap banks are technically in parallel with the transformer, but they can appear in series, especially when you have one phase open on a transformer. Through some of the windings, those parallel capacitors can appear as series capacitances.

And there are some other more exotic ways of getting capacitance on high-voltage circuits. As I mentioned, with an open phase, parallel capacitance, like from a cap bank, which is almost certainly gonna be on a feeder, will look like series capacitance through some of the windings. And so that’s a way that you might not expect at first to get series capacitance. So you may not technically have series capacitance until a single-phase event happens and you got those two things happen at once. You’ve single-phased the transformer and you now have series capacitance.

Now, some utilities actually have guidelines for underground cable runs inside the transformer to limit the chances of ferroresonance. That can be helpful.

Damage from Ferroresonance

Going back to the paper, as the paper describes, you will have these severely distorted waveforms during ferroresonance that will usually cause immediate damage. It can damage a raster, it can damage the transformer itself.

Sometimes when you’re energizing a transformer one phase at a time, you may hear the transformer go into ferroresonance. And if you hear this, if it’s mild, you may be able to finish the connections. You know, quickly finish the other two phases and it’ll go out of ferroresonance once all three phases are there. If it’s severe, you’ll see the transformer actually vibrating. There are a lot of stories of transformers physically walking off the pads. That’s, of course, a damaging event.

But you can also have the transformer insulation compromised. You may finish energizing it and it may hold once you’ve done that, but it may still have had degraded insulation inside that transformer. It can have mechanical damage and insulation degradation from the voltage that it’s seeing, and the physical vibration that happens when that transformer is in ferroresonance.

Transformer Core Design and Ferroresonance Susceptibility

The likelihood of a transformer going into ferroresonance depends somewhat on the construction of the transformer itself, how much coupling there is from one phase to another. And that depends on its core design. Some transformer cores, for example, the shell form core, the five-layer solid core, these have a somewhat tight coupling between the phases and so they’re more likely to go into ferroresonance than other types.

For example, if they’re three independent transformers or the triplex cores, equivalent of that, that’s the less coupling per phase. Although it’s still possible for a bank of three independent transformers to go into ferroresonance, it’s less likely than an integrated transformer with some of these core designs, where there’s much stronger coupling from phase to phase.

High Efficiency Transformers

Another issue are these high efficiency transformers. There are some DOE mandates for amorphous metal and high-grade silicon steel cores that are meant to achieve much less loss. For example, there is a low-loss, 25 kVA transformer that only has 17 watts of resistive loss inside that transformer, compared to 78 watts of no-load loss for the standard transformer design.

Now, that’s great for efficiency reasons, but that is not great for providing inherent dampening for ferroresonance effects. That 78 watts of no-load loss is basically resistive load in the transformer itself that helps dampen any sort of propensity to go into ferroresonance, that lowers the Q of the temporary resonance that’s formed during a ferroresonance condition.

So these low-loss transformers are a lot more likely to go into ferroresonance, all else being equal, than a standard core transformer. So be on the lookout if you’re using these super high efficiency transformers, that they’re more likely to go into ferroresonance than the standard transformer.

Energizing Transformers to Avoid Ferroresonance

When you’re energizing a transformer, if you’re doing it one phase at a time, the most reliable way to avoid ferroresonance is if you have a cable connecting to the transformer, energize the far end of the cable first. And if it’s a dead front construction, then energize the transformer, so that way your switching of it is between the capacitance and the transformer.

So going back to this schematic, for example, if this series capacitance is that underground cable, if the switching event is on the resistance side of the capacitor, you have your… You’re connected to the cable to the transformer first and then energizing the entire cable, that’s worse. That’s more likely to cause ferroresonance.

On the other hand, if you energize the cable first and then connect the cable to the transformer live, if you can do that safely with a dead front transformer, that puts the switching event between the C and the L, the cable capacitance and the transformer. That’s less likely to have the transformer go into ferroresonance. So if you can do this safely, as the paper recommends, it can be helpful to energize cable first and then connect that cable when it’s live to the transformer. Again, if you have a dead front construction that allows you to do that safely.

Ferroresonance in Photovoltaic Witness Testing

Now, another situation where you’re likely to see ferroresonance is in photovoltaic witness testing. Here we see in figure five, the utility’s onsite doing witness testing and you can see those current steps in blue as the inverter is stepped up in power level, but we also see ferroresonance tests here.

And I’m gonna jump to PQ Canvas to show this in a little more detail. This is a witness test and it includes the ramping in the inverter and also includes intentionally removing a phase, so dropping a phase to see if the system disconnects itself. This is the anti-isolating test. But the problem is, every time this test was attempted, the transformer would go into ferroresonance.

We can zoom in here, we can see this huge spread between the min and max voltage. During this time, the transformer’s in ferroresonance. We can look at these waveforms. They can get quite ugly, where 60 hertz phase A is much higher than the rest here and then the graph, we got a graph on the primary voltage side. And we can see they were going up about 2,000 volts higher than it should be on the primary side.

Here in this situation, we look at the RMS values, we see the phase B in red is the normal and 20 volts. Phases here A and B are much higher, going on 140, 145. You have voltage all over the place. For the vector diagrams, we can see phase angles are kind of all over the map for voltage.

So this was a ferroresonance event on two phases. And then it switched to three phases here shortly. So here, we’re going up to almost 160 volts on phases A and C in this graph, and then phase B was still varying quite a bit, but not enough damage in it. We’re going from 118 to, say, 122, 124 on a 120 volt basis on the PTs.

And here we have a similar situation in figure five, where we’re looking at primary side voltage. We’ve got the PT razors put in. And you can see this 2,000 volt increase and dropping a couple thousand volts. The voltage is really all over the place, unregulated at that point, before something would trip.

So this is basically impossible to pass the single phasing test, because we go into ferroresonance immediately. So witness testing is another situation where you’re very likely to see this ferroresonance, because you’re intentionally creating the conditions for that trigger. You’ve got a lightly loaded transformer and you’re giving it a single phase event.

Techniques to Avoid Ferroresonance

So basically to avoid ferroresonance, you want to avoid single phase events on transformers, you want to avoid single phase fuses on transformers. You wanna try to avoid single phase switching on them if you can.

If you’re energizing a transformer, especially if it’s underground, where the cable is underground, where you’ll likely have a lot of series capacitance, you wanna try to energize it with a load bank on it, perhaps, to dampen that. Utilities will use a load bank of a few kilowatts or even 1,000 watts of resistive load while they energize it, and then remove that once it’s energized to decrease the likelihood of ferroresonance. All those techniques can be used to reduce ferroresonance.

And again, it’s a tricky situation, because it’ll cause equipment damage very quickly, but it often leaves no signs except in something damaged as to what happened.

Contact Information

If you have questions about ferroresonance, give us a call anytime, 1-800-296-4120 or send an email to support@powermonitors.com.