% Current and Voltage THDs in In-Plant Industrial Systems

[ThePunisher]

Hi all,

I read the IEEE 519-2002 and the standard have set %THD limits based on Table 11-1 but this is for utility application at the PCC between the industrial plant and utility interconnection (on the metering point).

However, it may be overly conservative if the same limits are applied to the downstream ditribution system. I was warned in the past that IEEE 519-2002 was catered for Utilities not for Industrial Plant downstream distribution (worst case, 20-8Y/120V, 600V, 4.16 kV systems). Assuming no resonance, the % THD is usually greater on where the non-linear loads are connected (mostly on the 600V and 4.16 kV systems in an industrial plant setting).

However, there are tables 10-3 and 10-4 in IEEE 519-2002 which sets the current limits on each harmonic frequency based on Isc/IL ratio and system voltage. Are these tables suitable enough (not overly conservative) to establish harmonic distortion limits for downstream distribution systems in within the Industrial plant power system?

If not or they tend to be overly-conservative, what are good basis for setting limits then for industrial power distribution system?

 

[X49]

At a minimum, I would recommend keeping harmonic voltage distortion below 5%, as recommended by IEEE 519. The current distortion requirements in IEEE 519 are more stringent and don't necessarily need to be met in order to have a healthy system.

 

[Skogsgurra]

An actual case illustrates that one can sometimes get away with rather high THD levels.

A few years ago, I visited one of the larger toilet paper mills (hygenic tissue) and had a chat and some coffe. These guys had only recently learned about THD and related stuff and were wondering where they were on a scale from 0 to 100%. I brought in a Dranetz 4300 and made a quick check. THD was 14% with 5th and 7th dominating.

That was a lot more than the (then) recommended 4% but no one had noticed any problems at all. Then, they decided to install harmonic filters and when they switched the filters on, the fuses blew before the filter guys had a chance to do any THD measurements. They never connected the filters again and I suspect that they still run with their 14% THD.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[magoo2]

Gunnar,

The 14% THD sounds like you were measuring current rather than voltage THD. It sounds like it's a reasonable level - if it is current. Do you know what the stiffness was - i.e., the Isc/Iload?

 

[ThePunisher]

Per Table 10-3 of IEEE 519-2002, 14% acceptability on the 5th and 7th Harmonic Current at 120-69,000 V is based on Isc/IL >1000.

Since Table 10-3 is having range of 120-69,000 volts, could this mean that I can use this as basis for downstream sysems (up to 120V) for current THD and use 5% for Voltage THD?

 

[Skogsgurra]

magoo, that remark is rather insulting. But I forgive you. It was voltage, not current.
I wanted to share this to illustrate that voltage THD may be a lot worse than the official limits say without anyone noticing it.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ThePunisher]

This is actually my dilemna. Some plants may have been operating with % Voltage THD or harmonic distortions without anybody noticing or any equipment damage or maloperated. And I don't find any standards that tells me the reasonable THD limits on downstream buses.

If the the total harmonic voltage distortion (root mean square summation) + calculated voltage (sending voltage-voltage drop) on a particular bus is calculated...is the sum going to be compared with the cable and bus insulation and use it as a maximum allowable basis (of course, the utility % THD at PCC per IEEE 519 should still be maintained)?

 

[Skogsgurra]

My view is that insulation levels have very little with this to do. Harmonic distorsion that is so high that it causes insulation problems have caused a lot of other problems before that.

The main problems with high THD is higher losses in transformers and (in extreme cases) busbars (due to skin effect), higher motor losses, higher current and possible resonance in PFC capacitors, flickering light (not so much any more) and a few other minor problems.

The most serious problem is, in my view, the transformer and motor losses. Resonance in PFC capacitors can be avoided but extra thermal load of the capacitors can be a problem. Modern capacitors are designed to allow a certain THD and that can usually be read from the nameplate.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[waross]

Most motors run below nameplate rated current. Some loads vary, but the RMS of the loading is often below the rated load. This gives some headroom to accept some higher harmonic heating.
eg: A motor which operates all day at 100% of rated HP will be more at risk from harmonic heating than a motor which normally operates at less than 100% of rated HP.
The real issue where the IEEE recommendations may come into play is when one plant has motors that are suffering from harmonic heating caused by a nearby plant which is causing high harmonic distortion on the distribution feeders.
The other issue which may become more common in the future is where a utility charges a penalty for excess distortion power factor at the PCC.
Nothing in this post shall be taken as in any way contradictory to any statement made by Skogs. Grin

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[LionelHutz]

The distortion level which causes an issue is different for each bus. You basically have 3 options. You can make a conservative plant standard which will require extra spending to purchase and maintain the harmonic mitigation equipment. You can wait until there are issues and address them at that time. You can basically do a power system study and address the real possible issues you find. It's a tough situation. If you are too conservative then you're the guy that costs the plant a bunch of money. If you are not conservative enough then you're the guy that cause the equipment failure.

I think that using the IEEE519 current distortion limits isn't a bad idea but you may want to decide how and when you apply them. For example, you may not need to be worried about a MCC containing only VFD's when the MCC load is relatively small part of the plant's power system.

magoo, that remark is rather insulting. But I forgive you. It was voltage, not current.

Instead of being insulted perhaps you should have been clear with what you wrote and actually specified that you were giving a voltage distortion figure. I would have expected you to know that saying "I measured 14% THD." is being vague.

 

[Skogsgurra]

I wasn't insulted. I forgave immediately. Should have added some smiley.

The fact that I presented 14% THD as a high THD number was, I think, indication enough that it was voltage THD.

That number would not even be worth mentioning if it had been current THD. Then, 14% would be unusually low.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ThePunisher]

14% Current THD unusually low? From where did you based this on. I am interested on how you arrive with this assessment as IEEE 519-1992 limits was pretty lower than 14% THD (or TDD). Please bear with me as I am still trying to understand the bases for such limits.

 

[ThePunisher]

IEEE 519-1992 section 10.1 also defines the PCC WITHIN AN INDUSTRIAL PLANT as "the point between the non-linear load and other loads".

Does this mean that Table 10-3 limits can be used at the a bus where the non-linear loads are connected(e.g. 4160 V MCC bus where several 4.16kV VFDs are connected)?

IEEE 519-1992 section 10.4 states that " ideally, the harmonic distortion caused by a single consumer should be limited to an acceptable level at ANY POINT in the system and the ENTIRE system should be operated without substantial harmonic distortion anywhere in the system. The harmonic distortion limits RECOMMENDED here ESTABLISH THE MAXIMUM ALLOWABLE CURRENT DISTORTION for a consumer..."

I would interpret that the limits indicated in Table 10-3 will be harmonic current limits at the "industrial PCC" which is the BUS where the non-linear load and other loads are connected (e.g. 4.16kV MCC bus feeding both 4.16 kV VFDs and other loads). In the case of an MCC dedicated to VFDs, the PCC can be the switchgear feeding the 4.16 kV "VFD" MCC. The ISC/IL ratio will still be based on the PCC between the consumer and utility (metering point).

My understanding on the above further supports that maintaining the harmonic injection limits on the valid buses within the industrial plant PCC would somewhat ensure that the voltage THD at the Utility-Consumer PCC is within 5%.

Is my understanding valid enough?

 

[Skogsgurra]

Punisher

This was "one of the larger toilet paper mills". And paper mills have lots of drives without PFC. So, the current THD is usually a lot higher than 14%. 30 or 40% is not unheard of. Sometimes more, especially if there are heavy IR dryers with phase control.

I do not base the number on any standard but on real life measurements.

It will be very difficult to find "the point between the non-linear load and other loads" in any paper mill or steel works because there is usually a mix of controlled drives (non-linear) and DOL motors (linear) connected to a typical busbar. I see the PCC as the busbar itself and not a point between two types of consumers.


Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ThePunisher]

Based on what I gather here and what I understood from IEEE 519-1992

It is then appropriate for me to say that IEEE 519-1992 limits ONLY pertains to the PCC between Utility and Plant (Consumer) particularly on the metering point of the utility

With regards to downstream industrial buses, the % Voltage and Current THDs increases as you go nearer to the bulk of non-linear loads and the calculation results (especially the ones using software simulations) is being used as an estimate to see if problems are eminent to occur. In the absence of standard limits, evaluating these downstream THDs is subject to engineering judgement and experience.

However, downstream calculations should also account for any possible RESONANCE between any installed capacitors and the system impedance (which is mostly inductive). I have two questions related to the determination of resonance and voltage distortion:

1. In assessing Resonance, is it practical to include capacitance by large medium-high voltage power cables?

2. In calculating voltage distortion (volts), are we re-calculating the system impedance (at the point of interest) in such a way that the inductive reactance component is re-calculated on each harmonic frequency (since frequency is proportional to inductance)and then harmonic voltage distortion (at each harmonic frequency) would be the RMS sum of the function of each harmonic current passing through the re-calculated system impedance?

I thank you in advance for any guidance and comments to my understanding.

 

[LionelHutz]

1. It could be used but it generally doesn't seem to be a problem. I did deal with a site where they claimed it is the cabling capacitance that's causing a 60th harmonic and the harmonic is so bad there is 100% voltage THD at times.

2. Yes.

 

[Skogsgurra]

My experience is that the capacitance of medium-high voltage cables do not cause much trouble when it comes to THD.

There shouldn't be much voltage THD when you are talking medium and high voltage. I rarely see more than a couple percent on 10 kV and there's probably less when you get to 110 kV and above. I never touch those voltage (no pun) and do not know much about what distorsion levels to expect there.

I had one case, though, where large CSI inverters on dedicated transformers connected to a widespread 10 kV cable distribution system caused resonance between transformer impedance (mostly L) and cable capacitance.

The resonance was so bad that the dry type transformers (Gonella, Italy, now defunct) were subject to PD, which created lots of ozone in insulation voids so that the insulation degraded and the transformers eventually self-destructed. We found pieces of the red epoxy insulation more than twenty meters from the transformer.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.