Estimating harmonics at the design stage

[bfuchs]

When you are designing a low voltage installation, how do you estimate, or what precautions do you take to make sure (if possible), that harmonics are not going to be a problem.<br>
Harmonics from equipment like :<br>
- variable speed drives<br>
- fluo lamps<br>
- office equipment<br>
- etc...<br>
Harmonics affect capacitors, currents in the neutral (triplens add up there unfortunately !), shorten the life of transformers, ... <br>
<br>
When do you oversize, when do you add filters, isolation transformers, double neutrals, ... How to determine the best solution ?<br>
<br>
Any ideas recommendations ?<br>
Thanks,<br>
<br>
Boris<br>
<br>

 

[jbartos]

List of precautions:<br>
1. Analyze voltage waveform for harmonic distortion by the harmonic analyzer. If it exceeds industry standard, IEEE Std 519-1992 limits (more than 5% in the total harmonic distortion and/or 3% in single harmonic distortion) then utilize harmonic mitigating devices listed below.<br>
2. Harmonic mitigating equipment<br>
2a. Passive harmonic filters<br>
2b. Active harmonic filters<br>
2c. Phase shifting transformers<br>
2d. Active power line conditioners<br>
2e. Motor-Generator sets (M-G sets)<br>
2f. Some combination of the above items<br>
3. Analyze voltage waveform again to make sure that it meets the IEEE Std. 519-1992

 

[bfuchs]

Thanks for your tips, but I was thinking more along the lines of prevention instead of cure. How to design (initially) your system to reduce or eliminate the effects harmonics will have.<br>
What solution is best in which conditions ...<br>
<br>
When to use active filters, passive filters, etc ... And more importantly, when NOT to use them !<br>
<br>
Thanks for any feedback on this.

 

[jbartos]

3/3/00<br>
Try to avoid switching devices of abrupt nature which create harmonics, e.g. IGBTs, SCRs, IGCTs, Triacs, MCTs, BRTs, etc.<br>
Apply oldfashioned variable transformers, magnetic balasts not electronic, electric or manual typewriters instead of computers, switching belts on electric motor pulleys to change speed of motor, overrate transformers, overrate motors, etc. This will minimize harmonics without mentioned passive and active harmonic filters, phase shifting transformers, etc.

 

[somanind]

bfuchs and jbartos, your postings on harmonic distortion made interesting reading.<br>
jbartos, were you replying in a lighter vein when you advise to use manual typewriters, switching belts etc etc ? !<br>
I am facing a situation where about 100 VFDs are to be installed at a location. Total non-linear would be about 2500KW and linear load would be about 1500KW.<br>
Line chokes of about 5% will be used with each VFD. By using a Square D Harmonic Calculator software, the estimated TDD (current distortion) works out to about 10.5% with 12% allowable limit as per IEEE 519-1992.<br>
My question is, how accurate could this software be ? What are the experiences from the field on the reliability of such an estimation ? Is it necessary to conduct a full-scale Harmonic Analysis Study by employing experts in Harmonic distotion ?<br>
Regards,<br>
somanind.

 

[bfuchs]

The estimation based on the calculation is really based on the input parameters, and how the problem is modelled.<br>
Is the harmonic spectrum of the VFDs correctly modelled, are they always running on max, can you identify the real worse case scenario ?<br>
If you are looking at current distortion, is it possible for all the linear loads to be down at any given time, leaving only the VFDs running (theory, or reality ?). If you only have non-linear loads running, then your current distortion will be extremely high ! Maybe worse than the 'calculated' value.<br>
It's very hard to work on such value, the only true method is to measure what's in the field, and from experience, start toget a gut feel of an installation and how much filtering to introduce, and where !<br>
<br>
Good luck !