Feeding Chiller Module

[synchro]

If you did use a transformer to increase the source impedance, you'd be better off having it boost the voltage up to 480V since the chiller manufacturer has a version for 460V and it still has the same 10 kA SCCR as for 208V.

If you broke it up into two 4 stack modules and had a 225 kVA 208/480V transformer for each one, the online calculator below gives a ~9 kA short circuit current with 3% transformer impedance and ~5.5 kA with a 5% impedance. This apparently assumes a zero source impedance, and also no impedance from circuit conductors has been included. I haven't tried to determine how accurate this calculator is, but I'm assuming it's at least in the ballpark.

I agree with the others that if you can get the manufacturer to supply an SCCR of 65 kA that would be best.

https://voltage-disturbance.com/power-engineering/power-engineering-impedance-of-transfromer/

 

[electrofelon]

If you did use a transformer to increase the source impedance, you'd be better off having it boost the voltage up to 480V since the chiller manufacturer has a version for 460V and it still has the same 10 kA SCCR as for 208V.

If you broke it up into two 4 stack modules and had a 225 kVA 208/480V transformer for each one, the online calculator below gives a ~9 kA short circuit current with 3% transformer impedance and ~5.5 kA with a 5% impedance. This apparently assumes a zero source impedance, and also no impedance from circuit conductors has been included. I haven't tried to determine how accurate this calculator is, but I'm assuming it's at least in the ballpark.

Granted I am a bit out of my league here with this sized equipment and fault current, but is using a transformer(s) really an acceptable solution? If I was writing the checks, I would want to know if there was really no other option than all the capital costs and losses costs of transformers.

 

[synchro]

Granted I am a bit out of my league here with this sized equipment and fault current, but is using a transformer(s) really an acceptable solution? If I was writing the checks, I would want to know if there was really no other option than all the capital costs and losses costs of transformers.

That's why the first word of my post was "If".

 

[electrofelon]

That's why the first word of my post was "If".

Just out of curiosity, have you run into a situation where you needed to use a transformer just to reduce available fault current? If so was that because there was just no equipment available that could handle it, or would equipment that could handle it actually cost more than the transformer route?

 

[synchro]

That's why the first word of my post was "If".

It would be interesting to see a comparison of the capital costs, because multiple 260' runs of 500 kcmil 3-phase is not exactly cheap either. A 225 kVA 208/480V transformer from Maddox is $ 8,443. I think it would be good, as you had mentioned, to look at the overall costs including the cost and constraints of the installation, and then make a decision based on that.

 

[synchro]

Just out of curiosity, have you run into a situation where you needed to use a transformer just to reduce available fault current? If so was that because there was just no equipment available that could handle it, or would equipment that could handle it actually cost more than the transformer route?

In my work for a large company I never had deal with this particular situation, but I'm sure that there are others in this blog that could chime in about it.

 

[electrofelon]

In my work for a large company I never had deal with this particular situation, but I'm sure that there are others in this blog that could chime in about it.

Yeah seems like it would be a fairly common occurrence in places with network distribution. Plus I have heard NYC frequently will ONLY supply you with 208, no 480

 

[topgone]

Yeah seems like it would be a fairly common occurrence in places with network distribution. Plus I have heard NYC frequently will ONLY supply you with 208, no 480

Tried punching your numbers in using a current-limiting fuse (800A), The best let-through amps is 23kA from a prospective current of 128kA. With that fact, the best option there is would be to lengthen the power cables if you're comfy with an interposing transformer.

 

[Tainted]

Tried punching your numbers in using a current-limiting fuse (800A), The best let-through amps is 23kA from a prospective current of 128kA. With that fact, the best option there is would be to lengthen the power cables if you're comfy with an interposing transformer.

You can't use let through current for determining the available fault current.

 

[tortuga]

If you did use a transformer to increase the source impedance, you'd be better off having it boost the voltage up to 480V since the chiller manufacturer has a version for 460V and it still has the same 10 kA SCCR as for 208V.

If you broke it up into two 4 stack modules and had a 225 kVA 208/480V transformer for each one, the online calculator below gives a ~9 kA short circuit current with 3% transformer impedance and ~5.5 kA with a 5% impedance. This apparently assumes a zero source impedance, and also no impedance from circuit conductors has been included. I haven't tried to determine how accurate this calculator is, but I'm assuming it's at least in the ballpark.

I agree with the others that if you can get the manufacturer to supply an SCCR of 65 kA that would be best.

https://voltage-disturbance.com/power-engineering/power-engineering-impedance-of-transfromer/

Great idea but why stop at 480 if they offer a 600 volt model? Just go for 600V (575V motor nameplate) and enjoy more savings on copper, not much difference in breaker / disco / working space requirements with 600.

 

[synchro]

Great idea but why stop at 480 if they offer a 600 volt model? Just go for 600V (575V motor nameplate) and enjoy more savings on copper, not much difference in breaker / disco / working space requirements with 600.
View attachment 2570618

Sounds fine. But it might be prudent to verify that replacement parts are readily available in the US.

 

[Tainted]

Great idea but why stop at 480 if they offer a 600 volt model? Just go for 600V (575V motor nameplate) and enjoy more savings on copper, not much difference in breaker / disco / working space requirements with 600.
View attachment 2570618

That's waay more work, I would need a transformer, then I would need ventilation in the room. Not to mention I need space for the transformer + it needs to be in a 2hour rated room. Oh, then you need primary and secondary disconnect switches, which adds more space that we need

 

[topgone]

You can't use let through current for determining the available fault current.

Putting the cart in front of the horse? Who said that's what I did?

 

[electrofelon]

That's waay more work, I would need a transformer, then I would need ventilation in the room. Not to mention I need space for the transformer + it needs to be in a 2hour rated room. Oh, then you need primary and secondary disconnect switches, which adds more space that we need

I think he was saying if you go the transformer route to lower AFC, might as well go 600V, it's probably the same cost.

 

[Knightryder12]

I will go ask the rep tomorrow... I hope it's possible to get 65k.

If not possible, I will tell my mech engineer to give me 3 stack + 3 stack + 2 stack modules.

For each 3 stack module I would need 1 set of 500kcmil with 260 feet of wire minimum to get the fault current below 10k. I would also be able to maintain 2% voltage drop.

260 feet is a lot but doable, they have a large garage I can just run it around the garage.

Most if not all chiller manufacturers can do up to 100K. I go through this a lot with my mechanical guys and their vendors. I think the last time we had to do 65K minimum and the adder was like $500. Not much money when the chiller itself was like $350K.

 

[Knightryder12]

In a 208V 3-phase building I am feeding chiller module. The available fault current at the service is 128,000 amps.

The chiller module has these specs:

LRA: 1359 amps
MCA: 930 amps
MOP: 986 amps
Recommended fuse size: 800 amps

The short circuit rating of the chiller module control panel is 10,000 amps. The chiller module will be 120 feet away from the power source.

How the heck do I get 10,000 amps from 128,000 amps? The length of the run would be astronomical with terrible voltage drop.

If I need a line reactor, which one do I get? Wouldn't the line reactor be extremely large? I'm also worried about the voltage drop too if the line reactor is installed. I am not even sure if this installation is feasible..

How far from the utility transformer is the mechanical equipment? Have you done a fault current calc. to see what your actual AFC is?

 

[Tainted]

Most if not all chiller manufacturers can do up to 100K. I go through this a lot with my mechanical guys and their vendors. I think the last time we had to do 65K minimum and the adder was like $500. Not much money when the chiller itself was like $350K.

Tbe chiller rep said they don’t have one at more than 10kA

 

[Knightryder12]

In a 208V 3-phase building I am feeding chiller module. The available fault current at the service is 128,000 amps.

The chiller module has these specs:

LRA: 1359 amps
MCA: 930 amps
MOP: 986 amps
Recommended fuse size: 800 amps

The short circuit rating of the chiller module control panel is 10,000 amps. The chiller module will be 120 feet away from the power source.

How the heck do I get 10,000 amps from 128,000 amps? The length of the run would be astronomical with terrible voltage drop.

If I need a line reactor, which one do I get? Wouldn't the line reactor be extremely large? I'm also worried about the voltage drop too if the line reactor is installed. I am not even sure if this installation is feasible..

I ran a quick fault current calc and got these results:

3 sets of 350kcmil cu = 47,120 AFC at chiller
4 sets of 250kcmil cu = 51321 AFC at chiller

So if you can get them to provide 65K you should be fine.

 

[topgone]

Although I'm late, I think you can achieve an AFC below 10 kA at the chillers if you are going to use a current-limiting reactor. My calcs tell me you need a 5%, 5 kVAr, 208V reactor (20uH) in conjunction with a 300 kVA transformer. But, it is a reactor that gives you a series impedance or a loss of 0.2%. The most insane solution I can think of is to loop your 3 X 3-conductor 500 MCM cable three times (for a total of 720 feet)and achieve 9 kA. But that's very expensive.

 

[electrofelon]

Thats pretty lame that the chiller people dont offer a unit higher than 10K. are they stuck in 1983 or something?