Specifying 18 Pulse Variable Frequency Drives


18-Pulse VFDs: To specify, or not to specify…

by: Vann Barden

We recently got the opportunity to quote on several large VFDs for a proposed raw water pump station.  We have a long history with the water treatment plant where the new pump station is to be built, having repaired several large ABB and Altivar VFDs for this customer and having actually replaced some of the same VFDs because their age (10 to 12 years, on average) made the customer somewhat fearful about future access to the components we’d need to affect additional repairs that would certainly have been needed as the VFDs continued to age.

The new VFDs we installed, like the older VFDs they replaced, are conventional 6-pulse PWM units.  Since there was no history of problems resulting from excess noise or harmonics at the facility, we saw no need to propose 18-pulse VFDs even though they had been turning up with increasing frequency at neighboring facilities and we might have been able to use their burgeoning popularity to convince our customer to spend at least twice as much money as was actually necessary.

Imagine our surprise, then, when the VFD portion of the specifications for the building of the new raw water pump station came out requiring 18-pulse VFDs.  I spoke with the maintenance supervisor at the water treatment plant to find out how this had happened, and was told that the engineering consulting firm that had been hired to write the specifications had made the decision on its own, not, apparently, as a result of potential
harmonics problems discovered by virtue of a site harmonics survey performed by an independent power quality consultant or the local utility, but as the result of employing boilerplate water treatment plant specifications, which, by default, specify 18-pulse VFDs.  The maintenance manager had been able to make a dent in the boilerplate, though.  The consulting firm had generously agreed to include as an approved source, the manufacturer of the VFDs we had previously installed at the plant—the VFDs the customer wanted for the new raw water pump station—but even then, only by addendum.

Why, we asked ourselves, would an engineering consulting firm specify for its customers products which are over engineered, and thus overpriced (in this case, by a factor of 2) for an application?  If we had chosen to give in to cynicism, an obvious answer might have been that the profits made by engineering consulting firms on these kinds of projects are typically directly related to the monetary size of the project.

Having a natural distrust, however, of answers that seem too simple, we decided to do some research on 18-pulse VFDs to discover why they might logically be specified for an application.  VFDs create harmonic distortion because they are non-linear loads, which draw non-sinusoidal current from the power source.  The fundamental frequency and integral multiples of it are combined by the waveforms that give rise to this non-sinusoidal current.  Problems may be caused throughout an electrical system when non-sinusoidal loads make up a substantial portion of it.  Elevated RMS current, nuisance fuse blowing, overheating of transformer and distribution equipment, random tripping of breakers, equipment malfunctions and shortening of the life spans of motors and other expensive components may be a direct result.  As a very broad rule of thumb, if the

capacity of an electrical system is greater than twice the demand placed on it by non- linear loads, harmonic distortion is low enough that there is little danger of appreciable interference with other electrical equipment.

However, there is more at stake in regard to harmonic distortion than the equipment in the immediate environs of the user’s non-linear loads.  The potential for a user to degrade the voltage of a utility, and by extension that of the many other users the utility serves, by virtue of the excessive harmonic distortion it produces via non-linear loads, made it necessary to establish a standard to limit harmonics.  The latest version of the resulting standard IEEE 519-1992 limits acceptable harmonic voltage distortion of power systems 69 kV and below to 5% total harmonic distortion (THD) and each harmonic to 3%.  In systems 69.001 kV through 161 kV, THD is limited to 2.5% and each harmonic to 1.5%.
In systems 161.001 kV and above, THD is limited to 1.5% and each harmonic to 1.0%.  Limits on current harmonics vary according to the short circuit strength of the system into which they are injected.  Simply put, the more harmonic distortion the system is able to handle, the more the user is permitted to inject.

The daunting complexity involved in applying IEEE 519-1992 in a one size fits all manner to the legions of disparate users it was meant to serve is already obvious from just these few percentages culled from a document that is over 100 pages long.  And herein, we realized, was probably the real reason a consulting engineer might specify 18-pulse VFDs where history and experience would dictate otherwise: fear of the unknown.  It’s the justifiable fear that, due to the product recommendations they, the consulting engineers, have made to a user, harmonic distortion limits may be exceeded, possibly resulting in penalties being levied on the user by the utility, and the extrapolated effect on the reputation and, by extension, the financial health of the consulting engineer.

So, when in doubt, over engineer?  Granted, with a well balanced 3-phase supply voltage,
an 18-pulse VFD performs in an exemplary fashion (except, of course, for the fact that the high losses caused by the required phase shifting transformers lower the efficiency of the VFD), easily meeting the limits established by IEEE 519-1992.  However, with a voltage imbalance of as little as 1%, the 18-pulse VFD is much less effective.  And, since voltage imbalances of 1% or more are not unusual, chances are the two extra 6-pulse rectifier bridges per VFD that the user has paid at least twice the money for may effectively amount, much of the time, to little more than expensive paper weights.

Logically speaking then, if there is good reason to believe that excessive harmonic distortion may be a problem for the user, why not simply specify harmonic filters in lieu of 18-pulse VFDs?  In comparative testing, harmonic filters from MIRUS International (the LINEATOR) and TCI (the Harmonic Guard) have equaled the performance of 18-pulse VFDs in mitigating harmonic distortion, and they’ve done so while also maintaining the high efficiencies of the 6-pulse VFDs they’ve been integrated with.  And, as an added bonus, the excellent performance of these harmonic filters is basically

unchanged when encountering voltage imbalances of 1%.  Oh, and the price of a 6-pulse VFD outfitted with either of these filters?  Well, suffice it to say that it’s substantially lower than that of an 18-pulse VFD.

So, are the presently popular 18-pulse VFDs an over designed solution to a problem that
can easily be alleviated in a more sensible and cost effective manner?  Possibly.  How many times in the past decade alone have latests and greatests and other must haves withered on the vine under the unforgiving sunlight of reality, to remembered years later
only as curiosities?  Certainly 18-pulse VFDs should not be written into VFD specifications for water treatment plants by default.  But, given the current climate of confusion and fear as regards implementation of IEEE 519-1992, it’s probably safe to assume that it won’t likely be consulting engineers who may eventually consign 18-pulse
VFDs to the dustbin of automation history.  No.  As always, that decision will have to be made—and its ramifications enjoyed or suffered—by the owners of the electrical power using entities whose profits or losses will depend upon it.


1.   IEEE Standard 519-1992, “IEEE Recommended Practices and Requirements for
Harmonic Control in Electric Power Systems,” Institute of Electrical and Electronics
Engineers, Inc. 1993.
2.   D.J. Carnovale, T.J. Dionise, and T.M. Blooming, “Price and Performance
Considerations for Harmonic Solutions,” Power Systems World, Power Quality
2003 Conference, Long Beach, CA.
3.   “Interpreting IEEE Std 519 and Meeting its Harmonic Limits in VFD Applications,”
Tony Hoevenaars, P.E., Member IEEE, Mirus International Inc.; Kurt LeDoux, P.E.,
Member IEEE, Toshiba International Corp.; Matt Colosino, Crescent Power Systems,
4.   “What is Meant by Meeting the Requirements of IEEE-519-1992?”, Howard G.
Murphy, P.E., Hitachi America, Ltd.

About deltaautomation

President of Delta Automation, Inc. Working in industrial automation in excess of 30 years
This entry was posted in Delta Automation, Delta service advantages, Service stories, Tech Tips and tagged , , , , , , , , , , , , , , , . Bookmark the permalink.

3 Responses to Specifying 18 Pulse Variable Frequency Drives

  1. Rick says:

    Dear Mr. Barden

    Great article. My name is Rick I am currently working at MTE Corporation a fast growing passive filter company in Menomonee Falls, WI. We are currently working to prove to people in the industry that a 6-pulse drive with one of our Matrix AP Filters is a more cost efficient and better performing solution than an 18 pulse drive. Your article gives a little bit of insight into what possibilities lie ahead and we are looking forward to helping consumers make the best decision. We are currently running tests directly comparing 18 pulse drives to our 6 pulse filter solution. We are also in the process of gathering information on 18 pulse drives to show the price difference as well. Overall this article shows exactly where we believe the 18 pulse drive is going to go but the difficulty comes in convincing people that this alternative solution works.

    • Rick,
      Thanks for the response and feedback.
      Unfortunately Mr. Barden passed away last August.
      I will say though, I know that he would have been pleased with your comments as well.
      This is a message that we here at Delta Automation, have been trying to get across to the engineering firms tat specify these units.
      We have successfully convinced several clients, by replacing a failed 12 or 18 pulse unit with a new 6 pulse unit with no increases in noise or harmonics. These clients have passed this information on to their engineering firms, but so far the engineering firms are not willing to step up and try something new, even if it means saving tens of thousands of dollars for the end user.
      Thanks again,
      Bob Culley, president.

  2. Most users and engineers do not understand the extra required equipment that has to be installed in the VFD package for the 18 pulse design. Not only does it add cost but the extra equipment reduces the reliability of the equipment. I have recommended the passive filters which perform just as well at a reduced cost. If they fail, they can always be bypassed which you cannot do with a 18 pulse drive. The only concern with passive filters is when used with emergency generators. The capacitors have to be shunted out of the system until the motor comes up to speed

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s