Chiller on step-up transformer

[BigMike]

On a project, we had two small 480V chillers. The MCA for each was 41A. The service to the building was 208V, so we used a step-up transformer to obtain the 480V. I used an existing 112.5kVA step-up transformer thinking I would be OK. The chiller that showed up was a larger single chiller with a MCA of 141. The existing transformer can only handle 135A on the 480V side. The chiller is on a VFD and has two stages. I can't get the FLA from the chiller manufacturer. I think I'm on the ragged edge of using the 112.5kVA. Even though it's on a VFD, I'm concerned about being that close to the capacity of the transformer. Any transformer size recommendations would be appreciated. I'm a little out of my comfort zone since it's a step-up transformer and don't know if there are some traps that I need to avoid.

Thanks.
Mike

 

[kwired]

It basically is not code compliant, but that said it is not likely to draw full capacity. If it did it would do so on the hottest day you could ever imagine - as that is when it would be running at 100% capacity for any length of time, and that would kind of depend on other factors like how much undersized it may or may not be for the application, or if there is other abnormal heat gain adding to the demand.

It is likely to never give you trouble, but is possible under the worst conditions.

 

[petersonra]

what is not code compliant? I don't recall any transformer sizing requirements in the code.

 

[kwired]

what is not code compliant? I don't recall any transformer sizing requirements in the code.

So are you suggesting this would work just as well on a 15kVA transformer code wise?

 

[Besoeker]

On a project, we had two small 480V chillers. The MCA for each was 41A. The service to the building was 208V, so we used a step-up transformer to obtain the 480V. I used an existing 112.5kVA step-up transformer thinking I would be OK. The chiller that showed up was a larger single chiller with a MCA of 141. The existing transformer can only handle 135A on the 480V side.
Thanks.
Mike

If the transformer is 480V and 112.5kVA, why the limit of 135A?

 

[Sahib]

MCA of 141A includes 25% for motors to size the circuit conductor. So the actual load on the transformer may not overload it.
In any case what is transformer OCPD for?

 

[petersonra]

So are you suggesting this would work just as well on a 15kVA transformer code wise?

I am suggesting it could be code compliant. it might not work. but the claim was that it was not code compliant.

 

[Jraef]

The MCA has to do with conductor sizing and as mentioned, would be based on 125% of the largest motor FLA + the sum of all other FLAs. If the chiller is centrifugal, as most are, then the VFD will not permit the current to exceed 110% for 60 seconds, well below the 125% if the chiller motor is the only load (probably is, but check). But remember, your CONDUCTOR SIZE to feed the VFD must be sized for 125% of the VFD input current, not the motor current. So you need to know if that MCA value you were given is inclusive of the VFD, or if that was done separately BEFORE the VFD was added to it. If it was separate, then all you really need to know to size the conductors is the VFD maximum input amp rating.

Or you could just go to the next size transformer larger than what you have, assuming no other loads. So a 150kVA xfmr is rated for 180A, size your conductors for that and you are good to go no matter what.

 

[StarCat]

Chillers in General

Chillers in General

VFD's on chillers are actually a fairly new application, although possibly coming into more widespread usage.
The types are Reciprocating, Scroll, Centrifugal and Screw. Centravacs are always larger tonnage machines, generally 200 tons and above.
I'd like to know what the make is. A MFG Engineering dept. should in all cases provide Compressor RLA figures for a Technician who has a need to know.
There is an " old " convention whereby some outfits did not like posting RLA on the machine for what amount to " wrong ", and " outdated " reasons. It was a ridiculous notion that Techs' were attempting to set the machine charge by " getting " it to RLA by adding gas which no savvy Mechanic would ever do.
Also its not possible to troubleshoot the compressor without knowing RLA. When its not published you have to call someone.
Air conditioning generally always has RLA on the nameplate.
Its definite that the HVACR Tech working on said machine or running startup needs to know what the compressor is rated for. AS thus I would start making demands on several fronts.
Depending on the application, chillers can get close to RLA and as noted the load will increase in the heat typicall even if the refirgeration load is constant due to increase the condensing medium temperature.

 

[kwired]

I am suggesting it could be code compliant. it might not work. but the claim was that it was not code compliant.

Maybe I should have just said it may be overloaded, by just a smidgeon instead of claiming it is not code compliant. But then again the MCA is probably 125% of largest motor plus all other loads so actual rated full load is likely going to be under the 135 amps secondary rating.

That said it was nice of whoever did this to spring up a different unit then the OP was already planning for. Had he known this unit was going to be installed he may have went with a 150 kVA transformer instead of pushing this one so close to max capacity.

 

[tesi1]

transformer size

transformer size

did you look at the degree rise of the transformer, if it is a 115 degree rise it can stand a 15% overload, and if it is a 80 degree rise it can stand a 30 % overload normally.

 

[texie]

On a project, we had two small 480V chillers. The MCA for each was 41A. The service to the building was 208V, so we used a step-up transformer to obtain the 480V. I used an existing 112.5kVA step-up transformer thinking I would be OK. The chiller that showed up was a larger single chiller with a MCA of 141. The existing transformer can only handle 135A on the 480V side. The chiller is on a VFD and has two stages. I can't get the FLA from the chiller manufacturer. I think I'm on the ragged edge of using the 112.5kVA. Even though it's on a VFD, I'm concerned about being that close to the capacity of the transformer. Any transformer size recommendations would be appreciated. I'm a little out of my comfort zone since it's a step-up transformer and don't know if there are some traps that I need to avoid.

Thanks.
Mike

I think this would be NEC compliant based on the MCA. Even though this chiller is VFD run I would be a little concerned that the transformer can handle inrush on startup and still maintain a VD acceptable to the manufacturer.
Also I'm wondering if this is really a step-up transformer or just a unit wired in reverse as that may be another issue. In my view you need to have a 208 delta primary with a 480Y secondary. Of course all the other code requirements would apply such as grounding and bonding of the secondary, GES, primary and secondary OCPD, etc.
Just wondering if you have considered issues other than sizing.

 

[kwired]

I think this would be NEC compliant based on the MCA. Even though this chiller is VFD run I would be a little concerned that the transformer can handle inrush on startup and still maintain a VD acceptable to the manufacturer.
Also I'm wondering if this is really a step-up transformer or just a unit wired in reverse as that may be another issue. In my view you need to have a 208 delta primary with a 480Y secondary. Of course all the other code requirements would apply such as grounding and bonding of the secondary, GES, primary and secondary OCPD, etc.
Just wondering if you have considered issues other than sizing.

I believe 2014 NEC doesn't allow reverse wiring a transformer unless it is specifically designed for that purpose - just a litte FYI for those interested.

 

[kingpb]

On a project, we had two small 480V chillers. The MCA for each was 41A. The service to the building was 208V, so we used a step-up transformer to obtain the 480V. I used an existing 112.5kVA step-up transformer thinking I would be OK. The chiller that showed up was a larger single chiller with a MCA of 141. The existing transformer can only handle 135A on the 480V side. The chiller is on a VFD and has two stages. I can't get the FLA from the chiller manufacturer. I think I'm on the ragged edge of using the 112.5kVA. Even though it's on a VFD, I'm concerned about being that close to the capacity of the transformer. Any transformer size recommendations would be appreciated. I'm a little out of my comfort zone since it's a step-up transformer and don't know if there are some traps that I need to avoid.

Thanks.
Mike

I don't see this as an NEC issue. The NEC is not a design guide. The question you have is based on engineering application.

Step 1:
Advise GC/Owner that there was a change in Chiller design. And there are 2 options.

Option 1: Utilize what you have and advise them that the transformer could fail earlier than expected, i.e. it may not last 25-30 years.

Option 2: Replace transformer with larger one.

Step 2:
Do what GC/Owner wants


The transformer will work. It will run hotter than originally planned, and maybe be a little noisier. The fact that the chiller is on a VFD is good for inrush on transformer when chiller starts because the chiller motor won't have a large voltage drop.

In my experience there is no way you are going to be drawing a overload amount of current since the transformer temperature rise actually allows for an increase (as stated earlier of as much as 30%) in current.

Provide all the technical data and I'll bet calculations will show its just fine.

 

[BigMike]

Thanks for all your responses. It is much appreciated.

I am going to proceed with the existing transformer and monitor it. If it's not performing well, then we will switch it out. Calculating the FLA puts me below the full capacity of the transformer, so now I hope the in-rush plays nice.

 

[Besoeker]

If the transformer is 480V and 112.5kVA, why the limit of 135A?

Aplogies.
I took it to be single phase.
Mea Culpa.....

 

[kwired]

Aplogies.
I took it to be single phase.
Mea Culpa.....

We will let it go, as you are not likely familiar with NEMA standard transformer sizes, 112.5 kVA is a standard three phase size but not a standard single phase size, at about any standard voltage as well.

 

[StarCat]

Compressor Starting

Compressor Starting

We will let it go, as you are not likely familiar with NEMA standard transformer sizes, 112.5 kVA is a standard three phase size but not a standard single phase size, at about any standard voltage as well.

AS noted above, starting a compressor on a VFD is quite soft compared to XL, this should be quite useful in your case. Also, if the absolute full capacity of the machine is not necessary, the drive settings " could " be moderated by the Startup Tech to optimize the RLA without sacrificing possibly too much.
As an example we have a European machine that utilizes a pump and we only have to run the drive at about 33% to

 

[StarCat]

Compressor Starting

Compressor Starting

We will let it go, as you are not likely familiar with NEMA standard transformer sizes, 112.5 kVA is a standard three phase size but not a standard single phase size, at about any standard voltage as well.

AS noted above, starting a compressor on a VFD is quite soft compared to XL, this should be quite useful in your case. Also, if the absolute full capacity of the machine is not necessary, the drive settings " could " be moderated by the Startup Tech to optimize the RLA without sacrificing possibly too much.
As an example we have a European machine that utilizes a pump and we only have to run the drive at about 33% to accomplish the necessary work required.
The charactersitics of the refrigeration load would determine this.

 

[drktmplr12]

AS noted above, starting a compressor on a VFD is quite soft compared to XL, this should be quite useful in your case. Also, if the absolute full capacity of the machine is not necessary, the drive settings " could " be moderated by the Startup Tech to optimize the RLA without sacrificing possibly too much.
As an example we have a European machine that utilizes a pump and we only have to run the drive at about 33% to

While modifying the settings is a viable option, he should take care to ensure that the frequency and voltage supplied to the motor is still within mfg. specs. OF course, asking the startup tech to make the changes is the right way to go.

Motors that operate outside their design curves don't do so well in the long run. For example, submersible centrifugal pumps typically cannot be operated at less than 30-35 Hz for long periods because the insulation on the windings will eventually break down. One must also consider that loading characteristics as running a motor off it's curve will cause excessive heat and premature wear. This is why the motor guys will load the motor and record work done and current draw during a startup.