Telecommunications Building - Current Harmonics

[gepedo]

Hello All....

A power quality analysis was completed at my workplace (telecommunications building). The final report suggests that we have high levels of current harmonics in our power system (measured at the main service entrance switchboard). It also notes that the level of voltage harmonics is within IEEE standards. The consultant is suggesting that further power quality testing be carried out at individual loads to identify specific harmonic producing equipment.

It it correct to assume that SCR based rectifiers have a tendency to pollute the load side with voltage harmonics and AC side with current harmonics ?

If this assumption is correct, these finding make sense as there are no filters currently in place on the AC side of these rectifiers and since all DC equipment is backed up via large DC batteries (which act as filters), our electronic equipment have had no problems.

I'd like to know if any one has encountered similar problems and if limiting the current harmonics produced by AC-DC converters via filters helped there situation.

Thanks in advance

 

[dpc]

Voltage harmonics are created by current harmonics. It is normal for current harmonic levels measured as a percentage of the fundamental to be much higher than the voltage harmonics where you have a large amount of non-linear loads. Your telecommunications facility is likely full of switching power supplies that have a highly non-linear current demand and create a high level of current distortion.

If you reduce the current harmonics caused by the non-linear power supplies, the current and voltage distortion will go down.

The first question to ask is whether or not these harmonic levels are creating any actual problems? Or is the serving utility complaining about harmonics (unlikely if your voltage distortion is within IEEE 519 limits).

There are a lot of people out there trying to sell solutions to power quality "problems" that are not really causing any real problems.

 

[alehman]

What kind of numbers are you seeing at the service?

 

[gepedo]

The numbers at the service are as follows:

-> maximum voltage total harmonic distortion is typically less than 4% (with occasional spikes of 6-8%).

-> maximum current total harmonic distortion is typically between 10-15% (with occasional spikes of 18-25%).

 

[Skogsgurra]

That is a very low current distortion. You usually see a lot higher values.

On the other hand, the voltage distortion is rather high. As a rule, the voltage distortion is about ten times lower than current distortion (yes, can be twenty times lower too, depends on mains impedance/short circuit rating).

To me, it seems that your mains impedance is very high - or that the current distortion measurement is wrong.

Anyhow, the voltage distortion (4 percent) shall not be a problem at all.

Gunnar Englund

http://www.gke.org

 

[cuky2000]

I concur with Gunnar & DPC since appear that the level of distortion is not at critical level.

For reference purposes, below is a sample of voltage and current distortion that I found in the net with uncompensated load harmonic data as follow:

Harmonic Voltage Current
Number Distortion [%] Distortion [%]
3rd 0.14 3.85
5th 3.85 15.21
7th 1.02 3.87
11th 1.22 4.41
THD 4.24 16.51

 

[jghrist]

The voltage distortion is not necessarily caused by plant loads. The utility voltage with no plant load will have some level of distortion from cumulative harmonic currents of all customers. This is why the voltage distortion limit of IEEE-519 is the responsibility of the utility and the current distortion limit is the responsibility of the customer.

 

[easpe]

If its not broke don't fix it. As long as there are no problems found caused by harmonics there are no postives to spending money on this issues. As a learning tool you might have a power meter with waveform and harmonic capture installed at the transfer switch load side. Then you could monitior it and shutoff different pcs of equipment and see what happens. I have about 60 t-com sites and harmonic readings are all over the place depending on types and makes of equipment. As long as no problems pointing to harmonic are found or overheating problem are found. Its fat, dumb and happy.