Conducting a power quality site analysis

There are three locations where you can determine the extent of harmonic interaction: the nonlinear load, the next upstream transformer, and the point of common coupling with the utility.Because we hear and read so much about the extent of harmonics, we might be concerned as to how this new disturbance may cause problems on our facility's electrical lines. We need to remember what harmonics consist

There are three locations where you can determine the extent of harmonic interaction: the nonlinear load, the next upstream transformer, and the point of common coupling with the utility.

Because we hear and read so much about the extent of harmonics, we might be concerned as to how this new disturbance may cause problems on our facility's electrical lines. We need to remember what harmonics consist of and what types of loads produce them. Only then can we begin to hunt this culprit down.

The basics of harmonics

Nearly all new electrical devices have power supplies that may cause a disturbance to a power source's current sine wave. The characterization we are looking for is the current spectrum: the actual demand made upon our source of energy. This spectrum is defined in terms of integer multiples of 60 Hz, which is known as the fundamental or 1st harmonic. The multiples are then 2, 3, 4, 5, 6, 7, etc. The ways in which the load device asks for its current needs, in different amounts of these multiples, determine the extent to which the sine wave is distorted. If the proportion of multiples is high compared to the value of the 60 Hz, we say the distortion is large; if the proportion is low, we say the distortion is small.

Since the spectrum we speak of is composed of multiple sine waves (three, five, or seven times the 60 Hz frequency), we know that all those requested components will come from the 60 Hz energy source. The power company has a large capacity of sine wave currents, and is able, up to a reasonable limit, to send those different frequencies to the load as it makes its demands. An example would be you going to a buffet line for a slice of roast beef. When you request your portion, you ask for a well-done piece. However, several in line behind you request the same. At some point, the server has to advise the next person in line that no more well-done meat is available.

In the same way, we are concerned that the growing demands of harmonic currents on all electrical systems will result, at some future time, in their incapability to support our needs without some modification of the sine wave shape (i.e. distortion for all of us).

So much for where these current requests come from (nearly every modern load). Let's examine what effect harmonic currents can have on our distribution systems. These currents are "thieves" running on our systems, robbing us of electrical capacity. They need this capacity in order to run through the system from the supplier to the load device requesting them. They also steal from the power company system if they are of sufficient quantity, and they can cause voltage changes or distortion on entire busway systems as the distorted currents pass through the output impedance of a transformer, making a voltage drop at each harmonic order in the spectrum.

This "thievery" is similar to the stealing of system capacity that occurs when we experience poor displacement power factor (PF). For example, with poor PF, we must carry more "charging" or "filling-up" energy into our windings, coils, and motors just to get them ready to produce work. When we add another nonworking component,this time,high frequencies (180 Hz, 300 Hz, etc.), we further increase the size of the system needed to sustain the work output. Fig. 1 shows the total size that the system must now be; this size is the length of the colored diagonal vector in the six-sided box. If we had only the PF problem, the vector would be in the plane of the kW-kvar axis. But, now we have the two thieves: displacement and high frequency distortion, represented by the isometric drawing. To run the same work, we need more kVA.


Our challenge is to keep both thieves under control and maximize the work for this size of electrical system. When we do this, we achieve high total PF.

In the case of the high frequency distortion, we know it's not going to do any work, but it will consume capacity and space on our conductors; it will cause overheating of apparatus; and it will cause overloading and failure of transformers that are not rated to handle the heating effects.

How to look for harmonics

As we enter the power system to look for the high frequency culprit, we can do some preliminary investigation to see if the problem is present and how bad it is. We do this with our existing meters, before we buy or involve expensive spectrum analyzers or harmonic measuring devices. Let's go through this first approach.

Using a true rms meter and a peak reading meter, we measure the current at a load device. If it's a 100%, 60 Hz sine wave current, the ratio of its peak to its rms value is 1.414:1. This value is called the current waveform's cress factor and is always this ratio for a pure sine wave.

Suppose we measure something considerably different, let's say 3.0:1. This is an indication of a distorted wave shape, since it's so far from the 1.414:1 ratio. In fact, you could probably do all right even with a wave having a crest factor plus or minus 7% that of the true sine wave. As such, you can set for yourself a range around the 1.411:1 number. As long as your peak-to-rms ratio is somewhere between 1.3:1 and 1.5:1, you don't need to make a further detailed study. This is shown in Figs. 2 and 3. If you're outside this range, ask for help in getting a more detailed study and a more sophisticated instrument, as you are into distortion.


Let's say this first approach shows we have a high' crest factor, and our ratio is outside the 7% range. In fact, it's terrible, approaching the peaked waveshape in Fig. 3: 3.0:1. Now let's make the preparations to measure the actual current spectrum being asked for by this load device.

We can do this with a handheld indicating instrument, which will measure the rms, percent harmonic current in each harmonic order (3rd, 5th, 7th, etc.), actual amps at each order, and total harmonic distortion (THD). Different instruments will do this and more, showing waveshape, capturing the shape, showing bar graphs of the amount of each harmonic, etc. Two instruments we use regularly are battery-powered, handheld, and very easy to use. More are coming into the market and should be considered as the instrumentation gets more portable.

Where to look for harmonics

Where do we look in our facility? How should we judge what's a problem? Is it a problem at one point but not at another? These are certainly logical questions when starting to look for high frequency.

Remember, there are three zones of concern in your facility that should serve as a gauge for harmonic interaction events.

The proliferating device. Let's pick a variable frequency drive (VFD) as an example. Say we measure the 5th, 7th, 11th, and 13th harmonics, with a spectrum of 40%, 25%,13% and 8% respectively as the distortion for each harmonic in percent of fundamental. That is a fairly typical VFD spectrum for the common six pulse type of rectifier/converter. Our question is, does this level cause this drive a problem, or any other load equipment at this point in the facility? NO? Then there is enough inductance in the system to "swallow up" the high frequency current.

The next upstream transformer. Are these currents large enough to create a voltage distortion on the secondary bus so as to affect other devices fed from that bus? No? Then the size of the transformer and its relatively low impedance, compared to what must be low level of harmonic currents, has kept you from danger.

The point of common coupling with electric utility. This is usually the service point. Do you measure either harmonic currents or voltages that are in excess of the maximum stated in IEEE-519? No? Then you have passed the last of the three zones and do not have a problem. You have used your capacity, (wiring, transformers, bus systems, etc.) to dissipate the currents being asked for by the VFD. While you have no need to spend any money for a "fix" at the present, watch out that your management does not wish to suddenly use the so-called excess capacity to install more load. At that time, you may need to examine which is the least expensive alternative for the new capacity: adding harmonic traps to solve the harmonic mitigation needs, thus freeing up system capacity to do the new work, or investing in new switchgear and apparatus to provide the increased current capacity for the new work. The best answer will be to mitigate the harmonics, free up your system ampere capacity at a lower cost than new switchgear, and proceed to handle the new load at the lowest cost possible.


In Part 1, we examined a site for "mystery disturbances: those happenings that do not correlate to sources of power delivery disturbances. These occurrences (spikes, impulses, surges) are generally transitory in nature, here briefly and then gone. As such, we need to capture them with an analyzer in order to see them at all. We also looked at those disturbances connected with the power distribution system, both inside and outside of our facility. We are reminded to verify what kind of problem we are trying to solve in order to know where to look.

In Part II we were in more familiar surroundings, where the disruptions are more notice ably repetitive (sags, swells, momentary interruptions, etc.).

In Part 3 here, we look at other repetitive disturbances: The harmonic currents and voltages produced on the power system when nonlinear power supplies make demands of a linear, 60 Hz, sine-wave-producing power delivery system.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.