Reactor Design and Fault Current Levels for a Chiller
Reactor Design and Fault Current Levels for a Chiller
(OP)
This was an interesting issue I ran into while doing a fault study for a hospital. We turned out to be about 3kA over the 35kA AIC rating of an existing chiller. This rating was mostly constricted by a couple parts within the starter , (CB and contactor) that were only rated at 35kA. I was looking into different ways to get this issue resolved...
1 - to reduce the available fault current seen by the chiller
2 - to increase the AIC of the chiller.
Our recommendation as a firm is going to end up being to supply new VFD starters for these chillers. After looking into using a line reactor to reduce the fault current seen by the chiller, an app engineer told me that these reactors can only be used when a VFD is existing and will reduce the seen fault current by max of 5%. I really have no feel for these reactors and what they do, and when you would want to use them by and large.
I have heard many times of line/load reactors. Is the premise of these to simply control the impedance seen by the source so as to allow for more power quality, and reliability? or to correct power factor etc on a load by load basis?
input appreciated
1 - to reduce the available fault current seen by the chiller
2 - to increase the AIC of the chiller.
Our recommendation as a firm is going to end up being to supply new VFD starters for these chillers. After looking into using a line reactor to reduce the fault current seen by the chiller, an app engineer told me that these reactors can only be used when a VFD is existing and will reduce the seen fault current by max of 5%. I really have no feel for these reactors and what they do, and when you would want to use them by and large.
I have heard many times of line/load reactors. Is the premise of these to simply control the impedance seen by the source so as to allow for more power quality, and reliability? or to correct power factor etc on a load by load basis?
input appreciated






RE: Reactor Design and Fault Current Levels for a Chiller
1) Current-limiting reactors are used to limit fault current to downstream equipment. This would seem an entirely appropriate application for the situation you have.
2) Reactors used on input of VFD to limit or smooth harmonic currents drawn by the VFD during normal operation and limit the harmonic impact of the VFD on the power system.
3) Output reactors on the drive to help control high voltages that can occur at motors when fed from VFDs, especially on long feeders.
At first glance, your proposed solution seems like a serious waste of a VFD, unless it is needed for some other reason.
Have you considered simply putting a current-limiting fuse upstream of the chiller starter? If there is a US-made circuit breaker in the chiller, there should be no problem in finding a UL-tested combination of upstream CLF and downstream breaker that will limit fault current to something the downstream breaker can handle.
A current-limiting reactor would be another option and it will definitely work, especially since you need only a slight reduction. But it will be big and you'll have to find somewhere to put it, deal with heat produced, etc.
RE: Reactor Design and Fault Current Levels for a Chiller
By the way, chances are really good that without current limiting fuses, that VFD is only going to have a courtesy 5kA rating anyway.
RE: Reactor Design and Fault Current Levels for a Chiller
What code covers this , and fault studies in general?
RE: Reactor Design and Fault Current Levels for a Chiller
As for codes, we'd need to know what country the project is in.
RE: Reactor Design and Fault Current Levels for a Chiller
RE: Reactor Design and Fault Current Levels for a Chiller
In this case, the fuse and the downstream breaker must be a "listed combination" which means the fuse and the breaker have been short circuit tested together. Most molded case circuit breakers have been tested with a variety of upstream fuses, so you should not have much problem finding a listed combination. Check with the breaker manufacturer.
RE: Reactor Design and Fault Current Levels for a Chiller
RE: Reactor Design and Fault Current Levels for a Chiller
Virtually any low-voltage fuse you can buy in this size range will be current-limiting.
RE: Reactor Design and Fault Current Levels for a Chiller
Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com
RE: Reactor Design and Fault Current Levels for a Chiller
As DPC saied:
2) Reactors used on input of VFD to limit or smooth harmonic currents drawn by the VFD during normal operation and limit the harmonic impact of the VFD on the power system.
Could you please send me some link about this application of reactor.
We have up to 40% 5-th harmonic in 690V AC system and we think it's because not good choice of reactor.
Regards.
Slava
RE: Reactor Design and Fault Current Levels for a Chiller
1) Is the 38kA total fault current (power system + chiller contribution)? If so, that is not the right number to look at. The CB and contactor in the chiller see current either only from the chiller or only from the system, and will not see total fault current.
2) Revisit your cable impedances. Where did you get your cable impedances from? If you go to 3 different resources for cable impedance, you will find 3 different impedances. Did you use a resource that is maybe too conservatively low?
3) Revisit your utility source impedance. Did you get a source / utility impedance that erred on the high current side?
4) Revisit assumptions of X/R for the utility and step down xfmr. However, revising this can make your calclulated fault duty rise, but you likely need to be sure you have them right.
RE: Reactor Design and Fault Current Levels for a Chiller
RE: Reactor Design and Fault Current Levels for a Chiller
RE: Reactor Design and Fault Current Levels for a Chiller
on current.
Regards.
slava
RE: Reactor Design and Fault Current Levels for a Chiller
I've always found these guys to be helpful regarding reactors or harmonic filters:
http://www.transcoil.com/
RE: Reactor Design and Fault Current Levels for a Chiller
Harmonic voltage is a function of harmonic current times source (including reactor) impedance, so the reactor will impact harmonic voltage.
This description may be an oversimplification. But in my opinion, I'd be looking at your drive rather than your reactor if you want to figure out where that 5th harmonic current is coming from.
RE: Reactor Design and Fault Current Levels for a Chiller
DPC:
Your solution 1) (at top) seems great. This is what my manager suggested looking into. But after digging and talking to the reactor manuf., it seems that he indicated that you can only limit fault current in this manner if you are feeding a VFD..
please expand if possible -
Theres supposed to be a LittleFuse rep calling me tomm so X fingers.. The fuse solutions going to be a lot cheaper than 2 new VFDs..
RE: Reactor Design and Fault Current Levels for a Chiller
The reactor will work, but I still think the CLF will also work. You really need to be working with the circuit breaker manufacturer as opposed to the fuse supplier. They can tell you if the breaker has been tested with upstream fuses. Fuses are all the same basically, despite what the Littlefuse guy is going to tell you.
RE: Reactor Design and Fault Current Levels for a Chiller
There are a couple tricks, though:
1) As previously mentioned, you NEED to make sure the fuses are LISTED for use with the devices they are intended to protect. See NEC 240.86(B).
2) Motor contribution (which is often eliminated when using VFD's) can complicate their application. A quick reading of NEC 240.86(C) might convince you that your vendor is right, that you can't use current-limiting fuses to protect your starter (aka, "series rating"). But when you realize that the motor is connected to the load side of all devices involved, you'll also realize that you're in compliance with (C).
The worst part may be 240.86(A). If I were you, I'd make sure your fuse vendors stamp went onto that required documentation rather than my own.
Of course, if I was really you, what I'd really do is tell your manager that HE'S the idiot and then proceed to wrap the feeder around the wall, and forget about all this current-limiting fuse stuff.
RE: Reactor Design and Fault Current Levels for a Chiller
I factored in motor contribution. This bumped up the original fault current from 36kA to 38KA, ish. So we were still a bit over.
Wrapping feeder is certainly not the ELEGANT solution to this. And I'm not even convinced its the cheapest. At approx 60$/ft, it wouldnt be cheap. A fuse would be cheapest if possible.
RE: Reactor Design and Fault Current Levels for a Chiller
I am trying to model putting a particular fuse in my system to see if SKM dappor/captor short circuit simulation is somehow adjusted if you put a particular fuse in the system. I'm just putting busses on the same side. I still don't quite understand though I don't expect SKM to pop out lower numbers for the short circuit, but yet someone's telling me here and from Littlefuse that the fault will be cleared immediately for anything below 200kA of a fault .
I don't intuitively understand the relationship of a fault, the breaker coordination to make sure tripping order works (whats tripping order) , and this ultimate relationship with your fault current values etc. I can see a relationship forming, so that I will understand this stuff REALLY WELL, but the synapsises just haven't started firing quite yet....
jimbo
RE: Reactor Design and Fault Current Levels for a Chiller
Identical.
I don't get it, I think I need to be thinking at this more from a time of trip/break the circuit perspective...
please shed light for the love of all that is good
RE: Reactor Design and Fault Current Levels for a Chiller
Unless the fuse and the breaker are a listed series rated combination the calculation of peak current is just an exercise, you can't do anything with it, it does not change the fact that the breaker is over its rating. If the fuse and the breaker are a listed combination, then the calculation of peak current is also irrelevant and you can use the combination as long as your fault current does not exceed the combination rating.
While looking at the interrupting rating of components in the motor circuit, the motor contribution is just a red herring, forget about it. For any fault devices on that circuit your device will only have to interrupt either the fault current from the supply or the motor contribution, but never both. Contribution from your motor becomes important when evaluating any other circuit as it then combines with the supply fault current.
RE: Reactor Design and Fault Current Levels for a Chiller
Click on Run/Balanced System Studies
Click on SC
Click on Setup
Under report specifications for Branch Currents, select "Second Branch from fault" or "All Branch Currents"
Under Faulted Bus, select the bullet "Selected Buses" and then click "Select" and select the bus on one side of the fuse or the other.
Now run the fault study.
SKM does not make it all that easy to see the individual contributions. To do this, you either
a) study the SC.rpt, or whatever you named it. It can be a tad confusing to follow what the report is saying, but it will show the currents in all the reported branches for the specified fault.
b) You can create a custom Data Block Report that shows currents in all cables right on the one line.
c) Create a Probe Data Block where the probe displays branch currents (enable the probe via View/probe, and then click on the cable of interest to see the current in the cable).
An annoying issue I have with SKM does not have a "Close in fault" function. If you want to see the current going to a close in fault on a feeder that has a motor at the other end, and want to see the current from the bus and from the motor for a fault on either side of your fuse or breaker, you have to create a fake bus a few feet outside the main bus, and attach the motor cable to the fake bus, and then create a short (3ft?) cable between the main bus and the fake bus. Then you fault the fake bus. Clumsy as heck and the pits in terms of user friendliness.