non Linear loads

[Dennis Alwon]

A friend called and asked some questions about a job he may do.

A residence with a remote building where the remote building has a 200 amp single phase service. The owner wants to install many, many computers in the building at 800 watts each. He is not sure how many but it may be 50 or so.

The question is if you have a load on the neutral of 1600 watts for one circuit (2 machines), and 1600 watts on the neutral for the other phase, then is load on the neutral 1600 watts. If I think about the service as a multiwire branch circuit then the neutral load would be "zero" but I don't think it works that way.

Secondly, once the neutral is calculated and the non-linear loads are more than 50% then how do we size that neutral ?

Thirdly, do we need to even worry about this on single phase since the code mentions harmonics for a 3 phase wye and not single phase.

 

[kwired]

A friend called and asked some questions about a job he may do.

A residence with a remote building where the remote building has a 200 amp single phase service. The owner wants to install many, many computers in the building at 800 watts each. He is not sure how many but it may be 50 or so.

The question is if you have a load on the neutral of 1600 watts for one circuit (2 machines), and 1600 watts on the neutral for the other phase, then is load on the neutral 1600 watts. If I think about the service as a multiwire branch circuit then the neutral load would be "zero" but I don't think it works that way.

Secondly, once the neutral is calculated and the non-linear loads are more than 50% then how do we size that neutral ?

Thirdly, do we need to even worry about this on single phase since the code mentions harmonics for a 3 phase wye and not single phase.

I won't claim to be an expert on harmonics, but we are always told they are additive in the neutral. So something to chew on a little before someone that knows harmonics better chimes in - for single phase multiwire applications the effects would not likely be as bad as they are for three phase multiwire applications as the additive effects of the harmonic currents is only coming from two ungrounded conductors instead of three. What I don't know is if the harmonics are still additive when there is 180 degree phase angle or if it is only additive because of the 120 degree angles of a wye system.

 

[ron]

It is just a multiwire circuit, so not considering the harmonics, the neutral conductor will have 0A.

Harmonics are load based and will differ depending on the kind of power supplies are in the computers.

3rd harmonic and all of its integer multiples (collectively called triplen harmonics) generated by 120^o phase-shifted fundamental waveforms are actually in phase with each other, that is why on a 4 wire feeder, there is a concern of excessively high current that may be more than the individual phase conductors. This led to years of oversized and double neutrals.

The triplens are not additive for 120/240V with the 180^o phase-shifted fundamental waveforms

Most will just run a single full sized neutral to accommodate either one circuit full on and the other off, or both on and significant harmonics on the neutral.

 

[Dennis Alwon]

Ron, I don't think a calculation of a neutral conductor is calculated that way. The branch circuits are not multiwire branch circuit so if you calculated the load of the neutral as a service with just those 2 circuits I don't believe the load would be 0

I guess you would have to size the neutral to the largest load on one phase as you could have one phase off

 

[kwired]

Ron, I don't think a calculation of a neutral conductor is calculated that way. The branch circuits are not multiwire branch circuit so if you calculated the load of the neutral as a service with just those 2 circuits I don't believe the load would be 0

I guess you would have to size the neutral to the largest load on one phase as you could have one phase off

In a two wire (120 volt) branch circuit the current (linear as well as non linear) is same in both circuit conductors. You are concerned about additive effects of harmonic currents in a neutral conductor of a multiwire circuit - which your feeder and service conductors would have this apply to them. But as Ron mentioned the linear load on the neutral of the single phase source is going to be zero (the linear balanced load) So you are starting out with less current on the neutral to begin with before any harmonics start to come into play on such a system, is kind of what I think he is saying. Then as I said earlier you don't have a third phase to add even more harmonics to the neutral.

 

[Ingenieur]

Grow lights

 

[Dennis Alwon]

Grow lights

No pc's-- lots of them. My friend didn't know why he had so many but he thought it had something to do with bitcoins

So it sounds like we really don't have to worry about the neutral load at all... He will be happy to hear this.

 

[GoldDigger]

Grow lights

Many computers usually means Bitcoin mining, so minimizing power cost is critical.
For a single phase MWBC, all even order harmonics would be additive.
But to get even harmonics would require a current asymmetry between positive and negative half cycles, such as a single half wave rectifier input.

Sent from my XT1585 using Tapatalk

 

[Jraef]

"Bit coin mining" sounded like something fun and interesting... until I looked at the details. BOOOOORING! It's basically an incestuous process of setting up computers to help process and verify other bitcoin transactions from all over the world as part of a decentralized network based system, for which you are rewarded with... bitcoins.

The thing about the Power Supplies will boil down to this: Are the Switch Mode Power Supplies "power factor corrected" or not? If they are, then the harmonics is insignificant. Everywhere else in the world EXCEPT the US require SMPS to have PFC, we do not. But indirectly we generally end up with the PFC versions anyway, because manufacturers don't really want to bother making different versions and our domestic marketplace is not big enough for a US mfr to stand alone and compete against suppliers selling all over the world.

 

[junkhound]

There is probably a delta-Y somewhere in the system.

Circa 1992, company installed about 200 new computers in the offices. Not PFC corrected inputs, straight rectifiers in those older power supplies.

About a week later the poco transformer on the roof smoked due to all the triplett harmonics.

OP should check on the type power supplying the new computers, they likely have a PFC circuit in the front end, so no problem. Easy to check, slap on a CT and look at the current waveform.

 

[drcampbell]

...
OP should check on the type power supplying the new computers...

With luck, they haven't been purchased yet and there's time to specify that power-factor correction will be implemented.

 

[kwired]

There is probably a delta-Y somewhere in the system.

Circa 1992, company installed about 200 new computers in the offices. Not PFC corrected inputs, straight rectifiers in those older power supplies.

About a week later the poco transformer on the roof smoked due to all the triplett harmonics.

OP should check on the type power supplying the new computers, they likely have a PFC circuit in the front end, so no problem. Easy to check, slap on a CT and look at the current waveform.

OP has single phase service supplied by single coil with a center tap as far as we know.

Unless you want to call the transformer(s) back at the POCO substation part of "the system".

 

[infinity]

A friend called and asked some questions about a job he may do.

A residence with a remote building where the remote building has a 200 amp single phase service. The owner wants to install many, many computers in the building at 800 watts each. He is not sure how many but it may be 50 or so.

I wouldn't even worry about harmonics in this installation.

 

[junkhound]

Consider the 'remote building' and the 40 kW of computer loads into the discussion,
plus the discussions in other threads about how the poco prorate their distribution transformers (eg. feeding 5 different 200 A residential panels with a single 25 kVA transformer) (BTW, the OP's 800W computers are 'BIG' computers)

Say no PFC on computers, a guess at 0.75 PF. Guess at least another 2 kW load besides the computers, plus another 15 kW just for the air conditioner just to keep the computer room cool enough.

Better at least tell the poco what the loads are before smoking something ?

 

[Carultch]

I won't claim to be an expert on harmonics, but we are always told they are additive in the neutral. So something to chew on a little before someone that knows harmonics better chimes in - for single phase multiwire applications the effects would not likely be as bad as they are for three phase multiwire applications as the additive effects of the harmonic currents is only coming from two ungrounded conductors instead of three. What I don't know is if the harmonics are still additive when there is 180 degree phase angle or if it is only additive because of the 120 degree angles of a wye system.

They are only "additive on the neutral", when the harmonic multiplier of the fundamental frequency is divisible by the number of phases. So your fifth harmonics are not additive on the neutral, but your even harmonics in split phase, and your multiple of 3 harmonics in 3-phase, are.

 

[Sahib]

Whether harmonics are additive or cancel in the neutral, harmonics are present in the phases and they could cause maloperation of the very equipment producing them. Also, POCO may fine you if there is a harmonics penalty clause. So it is prudent to measure and ensure harmonics distortion is within limit

 

[Besoeker]

A friend called and asked some questions about a job he may do.

A residence with a remote building where the remote building has a 200 amp single phase service. The owner wants to install many, many computers in the building at 800 watts each. He is not sure how many but it may be 50 or so.

The question is if you have a load on the neutral of 1600 watts for one circuit (2 machines), and 1600 watts on the neutral for the other phase, then is load on the neutral 1600 watts. If I think about the service as a multiwire branch circuit then the neutral load would be "zero" but I don't think it works that way.

Secondly, once the neutral is calculated and the non-linear loads are more than 50% then how do we size that neutral ?

Thirdly, do we need to even worry about this on single phase since the code mentions harmonics for a 3 phase wye and not single phase.

I assume you mean a 120-0-120V system? I don't often deal with single phase non-linear loads on a system like that but here are my thoughts for what they're worth.

If the computers are all the same and are distributed equally on each half with a common neutral back to the supply then you might reasonably assume that, for the non-linear loads on each half, their input current waveforms would the same magnitude but in antiphase and cancel even if they are not sinusoidal.

The three phase case is different. If you have phase to neutral non-linear loads you get third order harmonics (three times the supply frequency) so they end up being in phase with other so add arithmetically in the neutral. That's why oversized neutrals are (sometimes) specified.

 

[Unbridled]

A friend called and asked some questions about a job he may do.

A residence with a remote building where the remote building has a 200 amp single phase service. The owner wants to install many, many computers in the building at 800 watts each. He is not sure how many but it may be 50 or so.

The question is if you have a load on the neutral of 1600 watts for one circuit (2 machines), and 1600 watts on the neutral for the other phase, then is load on the neutral 1600 watts. If I think about the service as a multiwire branch circuit then the neutral load would be "zero" but I don't think it works that way.

Secondly, once the neutral is calculated and the non-linear loads are more than 50% then how do we size that neutral ?

Thirdly, do we need to even worry about this on single phase since the code mentions harmonics for a 3 phase Wye and not single phase.

Be careful with MWBC.'s. Your must verify you are not using 2 phases of a 208/120V 3 Phase system. The neutrals to currents NOT cancel each other out like they do with a 120/240V Single Phase system.

208/120V 3 PH - using 2 of the 3 phases (3W MWBC)
1PH Nc= Highest current of the 2 phases

240/120V 1PH - Using both of the phases (3W MWBC)
1PH Nc= Sum of the two phases / 2 (in a balanced system it cancels each other out to 0 Amps.

See Mike Holt's Video's on his home page. " Calculating MWBC circuits."

 

[Besoeker]

Be careful with MWBC.'s. Your must verify you are not using 2 phases of a 208/120V 3 Phase system. The neutrals to currents NOT cancel each other out like they do with a 120/240V Single Phase system.

Dennis does state that it's single phase and I asked for clarification on that point.

 

[Jraef]

Dennis does state that it's single phase and I asked for clarification on that point.

It's a safe bet that it's single phase.

99% of single family residences only get a single phase service here. It takes a lot of money and cajoling to get a utility to run 3 phase out to a residential service, often including you having to pay the full cost for them to run the third wire from the nearest 3 phase connection point, which can be miles away.