IEEE 519 Standard – what do I need to know?  Where do I start? 

So one of the questions we get a lot is, obviously, how does IEEE 519, the standard that deals with harmonics, how does that relate to me, and what do I need to know about it?

Basically, let's take a quick overview of what's important here. First thing is that this is a recommended practice. So the intent is that it's supposed to guide you. It's not supposed to be very, very, very strict rules. And you'll notice that if you talk to utility companies, a lot of times they're typically not enforced unless required by neighboring customers and things that cause problems.

It's important to understand the point of common coupling, the voltage limit table, the current limit table, THD versus TDD, and why do we have statistical limits? And so all those things are important with IEEE 519. The 2014 standard, by the way, came out and replaced the 1992 standard, which, obviously, hadn't been changed in a long time. But really, there were minimal changes over the years, except for the fact that the document went way down in size and got very, very focused on important things.

And the important things are really that it establishes the goals for the design of electrical systems that includes both linear and nonlinear loads or harmonic-producing loads. It addresses steady-state limitations. So basically, some transient limitations, meaning things can change. But you have to be aware that there's some statistical-- that's where the statistical part comes into play.

The limited values are, again, recommendations, and they should not be considered binding. That's important. It's not so strict that you can't get around some of these numbers. And some conservatism is present, but it's not necessarily important in all cases. So as we look through this, it's important to understand, again, what the guideline is telling us.

Now the PCC or the Point of Common Coupling is really critical to understand. That's the point where another customer could be served. And the idea of this standard, by the way, is really for-- it makes the utility the mediator for when things go wrong. So if I as one customer create harmonics and you as another customer get the benefit of taking all the distortion that I've created and causing problems on your load, it's not so much a benefit. But if you have issues, the utility has to be the mediator.

So this point of common coupling is really important to understand. And again, it's typically where the utility can serve another customer. We're usually looking at the secondary if multiple customers are fed or the primary if you're the only customer on that transformer. A lot of times people ask us the question, what if I have to check my voltage distortion and current distortion on my transformer but I don't own the transformer, and the utility wants me to check in on the primary?

Well, if you pass, it turns out that if you pass the distortion levels on the secondary, you're going to pass on the primary. And the reason is because remember, harmonic currents come from the load. Unless your neighbor is screwing up the system, if you are putting harmonic currents on the system, the current is going to go through the transformer, except for some of the third harmonics-- Triplens. But most of the currents are going to go directly through. So the current on the primary and secondary should be about the same in terms of distortion.

The voltage distortion will be less on the primary than on the secondary. So if you pass on the secondary, you're going to pass on the primary. And that's because the impedance of the system that we're talking about, the transformer, is one of the bigger parts of the impedance.

Now, what if you said, I'm going to protect all my loads by correcting the harmonics on the loads themselves. Well, it turns out that if you could guarantee that all your loads would comply with 519-- if that's really even a thing, and it isn't because IEEE 519 is a system approach-- would your system comply? It's difficult to do because you would have to guarantee that every LED light, every computer, everything that you create, variable frequency drives, anything that creates harmonics would have to have a harmonic control on it.

If a very high percentage of your loads are drives and other large rectifiers and other things, it's possible to minimize the effect of harmonics with load-based solutions. But if you have a blended load throughout the system, it's going to be difficult to minimize all those things. So again, thinking of how to comply from a system level versus individual load level.

Managing the harmonics in a power system is really considered the joint responsibility of you as an end user and, for sure, the owners and operators, meaning the utility company. Some level of voltage distortion is generally acceptable, meaning you're going to have some background distortion because, again, harmonics are the new normal. We're going to have current flowing around the system. How much that screws up the voltage is really dependent on the impedances in the system.

Really, we both have to work, meaning end users and utilities have to work cooperatively to keep the actual voltage distortion below levels that are going to cause problems with loads. And the underlying assumption here is that by limiting the current injected loads from the system, meaning if you have VFDs and so forth, if we limit that current injection, the voltage distortion will be kept below a reasonable level that's going to cause problems.

And then the other thing is that this standard should not really be applied to individual pieces of equipment on your system. That's not the way that 519 was written. A lot of people will take and use that as a statement in design guides and things like that. But the guide really intended to have the point of common coupling as the place where these limits were applied.

So let's take a look at the limits. This is the voltage limit table, and you can see that it varies based on the voltage level in terms of less than 1,000 volts, medium voltage, and high voltage systems. But the main change that occurred between the '92 standard and 2014 standard was that the Total Harmonic Distortion or THD in voltage went from 5% to 8%. And that was really a function of people understanding that the loads weren't typically as affected at a lower level of voltage distortion.

The current distortion table is a little bit more complicated and detailed. What the left-hand column is saying is, I short circuit over load. If your ratio is small, then your TDD-- Total Demand Distortion on the right-hand side of the table-- also has to be small. What that means is if I have a weak system or a high impedance system, I can't have as much current distortion. If I have a lot of current distortion, I will cause voltage distortion.

And a lot of people get caught up in the percent THD and TDD and the current distortion levels. But in fact, this standard is, really, again, all about controlling the voltage distortion at the point of common coupling. Current distortion within your facility is going to cause heating. It's going to cause other operation issues and things like that. But really, at the point of common coupling, we're more concerned with voltage distortion. So this table really guides us in that direction.

The other thing here to notice is TDD versus THD. TDD versus THD, the difference in the equation is really what's in the bottom of the equation-- the denominator. The I 1 is the fundamental current, 60 hertz current like right now. And the I L is the demand current or the 60 hertz current at the peak demand on the 15 minute window, for example, on your system. So normally, that current's going to be higher, which makes the TDD value lower, which is, again, easier to comply with, so it's a little bit of a benefit.

So why is current distortion, THD, TDD important? I often show this graph. And if you had 100% distortion on your current, look what happens. Your RMS current would be 100 squared plus 100 squared, and that would be 141-- let's say it was amps-- 141 amps of total current, total RMS current. If your fundamental current was 100 amps and your harmonics were 20 amps-- so again, 20% distortion-- you'd end up with 102 amps.

And the reason I bring this up is, again, it's not so much about current distortion because if your system worked at 100 amps but didn't work in 102 amps, I would say that your system was designed marginally to begin with. And again, people will argue that harmonic currents cause more heating. But in fact, again, the difference between 100 amps and 102 amps in terms of actual system issues is insignificant. So using percent current distortion, even at 20%, I'm not overly concerned with that. And that's what the 519 standard also guides us to because, again, it goes back to voltage distortion.

Now, lastly, here, we'll talk about statistical methods. And why do they put statistics in here? Because harmonics are all about long-term effects. Typically, you're going to have-- it's like touching a hot stove here in the picture. If you touch a hot stove and you pull your hand away, you're likely not to get burned unless you touch the burner. But if you touch a hot stove and you know it's hot, you pull your hand away. It's a very short time.

Harmonics on a system in the same way don't create a lot of problems if they're only there for a short period of time. But what kind of loads can do that? Resonant effects, transformer inrush, other types of things like that, soft starts. But then there may be issues related to that. So there are times when we can't exceed certain levels, even during those short periods of time.

So again, summarizing. With IEEE 519, what's important to recognize-- voltage tables, current tables, statistical methods, the point of common coupling, and the fact that the IEEE 519 standard is intended to be a guideline so that one neighbor doesn't affect the other neighbor, and the utility is the mediator to guide how much harmonic current and voltage distortion is allowed to flow on the power system.