Harmonic Distortion

[danmsnyder]

I am a facilityt engineer in Newtown, Pa. After moving into our NEW building, many 277v, T8 electronic ballasts were found to be bad, about 15 out of 300. Since September, about 20 more have gone bad. It's a commercial office building, 30,000 sq. ft., with about 50 computers. There is a separate server room on a K-rated transformer. There are 2 TVSS surge suppressors installed, one on the feed to the server room and one at the MDP. Several personal UPS devices have failed or malfunctioned. I had a survey of the Main Distribution Panel conducted, but it only found about 2% THD, and about 25% current distortion, the only construction issue was bonding in the service entrance to MDP. We also have 4 air handling units each with it's own VFD, 2-10 hp and 2-15 hp motors. They are experiencing about 110% harmonic current, most 5th and 7th. Wondering what to do next?

Dan

 

[SphincterBoy]

Hmmm, the VFD's could be the source of your headaches.

Attach a Dranitz voltage/current/power/THD recording
device at each power panel, and make a recording of what
is happening.

My hunch is the lighting ballasts are bad due to infant
mortality (poor quality). If there is any problems caused
by harmonics, the source of damaging harmonics is most likely caused by the VFD's. The UPS's may either be bad due to infant mortality, or poor quality. We have
had defective out-of-the-box equipment many times.

It's too early to tell w/o the Dranitz readout.

You may have a periodic, slow rise overvoltage from your utility....the Dranitz will uncover this as well, and can
be used as evidence for corrective action (the utility will
install or adjust voltage regulating equipment).

 

[llangford]

I'm not an expert in this field but I have done some reading on topics similar to this. My recommendation would be to hire a professional electrical engineer or consultant to check your ground and neutral systems. As I understand it, the grounds and neutrals have to be tied to separate common points and in some cases routed in a particular fashion to eliminate the harmonic distortions caused from the electronic devices present on the system. Filters placed ahead of the VFD's might cure the problem but a professional would be better able to recommend this also.
Good Luck.

 

[busbar]

One simple way to detect harmonics is to measure voltage/current with a (known accurate) true-rms meter and an average-responding meter. If there is significant difference in the readings, there are probably harmonics present. Also, on 3ø 4-wire feeders using the same setup to check neutral currents is a method to help decide if more serious action is needed.

 

[dpc]

Dan,

You might check the voltage on each phase of the 480/277V system. If possible, leave a meter connected to check min/max readings. High voltage could be a factor. Also, make sure you have 277V ballasts.

The ballast manufacturer should be asked to explain the failure mode. I agree that you could just have a bad batch of ballasts. There is a lot of price competition on these and sometimes QC can suffer.

I'd be surprised if your harmonic distortion is the cause of the failure, but keep in mind that these electronic ballasts create plenty of harmonics all by themselves.

 

I run an energy services company. We install lots of T8 and electronic ballasts. We stick with name brand ballasts, because poorer quality ballasts can have extremely high failure rates, say 15%, and also produce large amounts of harmonics. What ballasts are you running? Check manufacturer data against design specs for the project, maybe an innappropriate ballast was used, and you can get a low cost solution. There are some awful smart people in these forums, but checking the simply things first is my strategy. Other problem appears to be small UPS failures...how many failures in how many units? These problems may not be related at all. What are your VFDs? Are they pulse width modulated (best in my book for harmonic control, as a rule of thumb), or are there isolation transformers on the VFD. Check the project submittal or O&M data books to get the VFD THD numbers. If you change ballasts, pick a great ballast/lamp combination. I don't have the physics backup, but some conbinations show especially long life on both ballasts and lamps.