Chiller SCC Calc, or Why Is This Even a Thing?

[kwired]

And there's the problem with the specifiers not paying attention to Article 409. All it takes is basically the same amount of effort it takes to determine the correct interrupt rating of a switchboard or panelboard. In fact, at some point you will ALREADY have that information upstream of the equipment. So by simply using that value as the AFC and requiring an SCCR at that level, you guarantee that the control panel will not need anything more when it arrives. Sure, the actual AFC is going to be slightly lower by the time you get to the control panel terminals, but that isn't a problem.

Where the issue comes in with the equipment provider being cheap (and often just lazy) is that in order to get a higher SCCR rating, the panel builder becomes more restricted on what they can use in terms of components. It's fairly easy now to attain reasonable SCCR values, even relatively high ones like 65kA, by simply choosing power circuit components that have those higher values, or are UL listed in series combinations that attain them. That however generally means that you can't buy each separate component from the lowest bidder and end up using an ABB starter behind a Square D breaker, because why would ABB test their starters in series combo with a competitor's breaker? So what they do is just "kick the can down the road" to the hapless EC who has to deal with it in the field, where it's MUCH more difficult. They are only allowed to get away with that because nobody told them they couldn't. That's where the job specs come to play here.

Like Bob Peterson, when I was a panel builder, I too would choose to take the high road and inform my customers what I was giving them. But he's right in that if nobody addresses it up front ahead of time, purchasing agents and buyers are only going to look at the bottom line costs, so the "can kicker" wins, which all but forces anyone trying to compete with them to do the same. That's why it HAS to be addressed up front.

But sometimes the equipment is selected by HVAC guys and they figured out all the heating or cooling demands and what capacity of unit they need but know nothing about electrical other then they can order the unit in 208-230 or 460 volts.

 

[petersonra]

Don was being generous when he said "unreasonable". I'm saying "impossible."

The chiller has a motor contribution of 2,800A, leaving 2,200A of the 5,000A SCCR for upstream impedance. But a 460V source needs an impedance of at least 460/2,200 or 0.21 ohms to be that low. However, with a source impedance of 0.21 ohms and an operating current of 440 amps, the voltage drop is 440 x 0.21 or 92 volts. 460-92= 368 volts, but the listed operating range is 414 to 506 volts. How is this even possible?

I don't think anyone here is disagreeing with you on the point you are making.

The problem is though that the people buying this stuff are basically specifying "make it as cheap as possible - I don't care what the SCCR ends up being".

It is not all that hard, or especially expensive these days to make a control panel 65 kA SCCR. But, I guarantee you that any HVAC manufacturer that chose to do so as a standard would lose sales over the little bit of added cost. The HVAC market is one that is essentially commodity items and price has become the driving force in that market.

 

[JoeStillman]

All it takes is these simple words added to the purchase specifications.

"Electrical equipment shall be listed and labeled for use with a Short Circuit Current Rating that corresponds to the Available Fault Current at the location where they are being installed." That puts the onus of providing a proper SCCR on the people supplying the control panels. If they go cheap and ignore it, getting the "courtesy" untested rating of 5kA (which is what that one has), they will be responsible for making it right. Even YOU, as the EC, can simply add this to your PO, or better yet, add something to your bid explaining that your price is BASED ON the fact that all equipment you will be connecting is suitable for the conditions of use.

This change took place in the 2005 NEC when Article 409 was introduced. There really is no excuse for people not paying attention to that any longer. 11 years is plenty of time to get current on the code...

These are all good suggestions and I will put them to use in the future. But in this case, we are looking at equipment where it's operating characteristics create an impossible contradiction. Under what conditions can this thing be suitable?

 

[Sahib]

Why not consider post#15 and possibly solve your problem?

 

[JoeStillman]

But sometimes the equipment is selected by HVAC guys and they figured out all the heating or cooling demands and what capacity of unit they need but know nothing about electrical other then they can order the unit in 208-230 or 460 volts.

This is the most common type of coordination-fail in all MEP offices. I wound up trying to change the electrical design to suit the mechanical equipment. That's how I discovered that it's impossible to meet the requirements of this thing.

 

[JoeStillman]

Why not consider post#15 and possibly solve your problem?

Thanks, this is actually the solution I am pursuing. There is a two-point connection version of this machine where I would use a 225 and a 300A CLF. Fuses rated this low have a CL threshold below 5,000A and it will work.

 

[texie]

I want to dare them to show me some kind of electrical system that can actually feed this monster.

Sent from my SM-G900V using Tapatalk

As I mentioned in my other post on this, this has been a real issue for us as well. I will be meeting with my UL and ETL reps next month and have a discussion about this. Your example is very intriguing as it looks mathematically impossible to be compliant.
While agree with jraef's comments in general, I'm not of the opinion that this is an Art. 409 issue since a typical chiller is listed as a complete package by the NRTL, not as a separate Industrial Control Panel. But I do agree the concepts are the same. In any event it would be impossible to comply with Art. 110.9 and 110.10.

 

[steve66]

Thanks, this is actually the solution I am pursuing. There is a two-point connection version of this machine where I would use a 225 and a 300A CLF. Fuses rated this low have a CL threshold below 5,000A and it will work.

I had a very similar situation not too long ago, but caught it before the equipment was ordered. When we asked the manufacturer, they provided a control panel rated for 35KA.

I'm not sure if there was a cost adder, and I believe the manufacturer may have just changed their standard non-fusible disconnect to a fusible one.

You are absolutely right - its crazy to have a piece of equipment that takes 200 amps or more to have a SCCR of only 5KA.

 

[ron]

Under what conditions can this thing be suitable?

It is possible in some scenarios if the available fault current is low already, whether because it is deep into the service entrance distribution path or the utility is provided to the building on a long radial line, etc.

An appropriate sized isolation or step down transformer with the "right" amount of fault current available on the primary (not infinite), will work ok for a 5kA controller.

 

[charlie b]

NOT SO FAST!

NOT SO FAST!

If you are counting on an item that bears the title, "Current-Limiting Fuse (CLF)" and that has a rating of, let us say, 5000 amps, to solve your problem, then you need to think again. You also need to bring either an engineer or the equipment manufacturer into the discussion. You can't take the "C" word or the "L" word at face value.

I should also advise everyone not to take the following information at face value either. Talk to someone who is an expert on fuses. Do we have such a person on this forum?

This lies outside my area of expertise. But I learned a few relevant facts from a colleague who is a long-standing manufacturer's representative. The issue comes from the presence of any kind of "active device" (e.g., a breaker or fuse, something that reacts to overcurrent conditions) internal to the controller associated with the chiller (or elevator or other equipment). As that "active device" reacts to a fault downstream of its location, meaning that the breaker starts to open or the fuse starts to melt, an arc is drawn across the contacts or across the melting element. That arc, however briefly it exists, will be in series with the fault, in series with the CLF, and in series with the ultimate power source. That arc will have a significant impedance that will (however briefly) reduce the overall fault current to a value that is outside the CLF's current-limiting range. In other words, a CLF in series with another fuse will not behave the same way as a CLF without a fuse downstream.

The bottom line is that you cannot take credit for a CLF as the method of reducing the fault current below the 5000 amps for which the chiller is rated, unless the CLF has been tested in combination with any active devices that are included in the chiller's controller. That, of course, puts us back into the hands of the chiller manufacturers.

 

[JoeStillman]

If you are counting on an item that bears the title, "Current-Limiting Fuse (CLF)" and that has a rating of, let us say, 5000 amps, to solve your problem, then you need to think again. You also need to bring either an engineer or the equipment manufacturer into the discussion. You can't take the "C" word or the "L" word at face value.

I should also advise everyone not to take the following information at face value either. Talk to someone who is an expert on fuses. Do we have such a person on this forum?

This lies outside my area of expertise. But I learned a few relevant facts from a colleague who is a long-standing manufacturer's representative. The issue comes from the presence of any kind of "active device" (e.g., a breaker or fuse, something that reacts to overcurrent conditions) internal to the controller associated with the chiller (or elevator or other equipment). As that "active device" reacts to a fault downstream of its location, meaning that the breaker starts to open or the fuse starts to melt, an arc is drawn across the contacts or across the melting element. That arc, however briefly it exists, will be in series with the fault, in series with the CLF, and in series with the ultimate power source. That arc will have a significant impedance that will (however briefly) reduce the overall fault current to a value that is outside the CLF's current-limiting range. In other words, a CLF in series with another fuse will not behave the same way as a CLF without a fuse downstream.

The bottom line is that you cannot take credit for a CLF as the method of reducing the fault current below the 5000 amps for which the chiller is rated, unless the CLF has been tested in combination with any active devices that are included in the chiller's controller. That, of course, puts us back into the hands of the chiller manufacturers.

The mech dep't specified clfs for each motor starter. I feel like that should have done the job but it didn't.

Littlefuse and Bussmann both publish application charts showing how the clf will respond to various fault levels and quantifies the current limitation. Why can't I just go up, over and down on the chart and call it safe?

Sent from my SM-G900V using Tapatalk

 

[ron]

Littlefuse and Bussmann both publish application charts showing how the clf will respond to various fault levels and quantifies the current limitation. Why can't I just go up, over and down on the chart and call it safe?

Dynamic impedance and that it hasn't been tested as an assembly with the load it is supposedly protecting makes it a problem.

 

[dkidd]

Why can't I just go up, over and down on the chart and call it safe?

See the article at http://advanceelectricaltraining.com/iaei/nec-requirements-for-short-circuit-current-ratings/.

Look at the fourth paragraph from the end especially.

 

[JoeStillman]

See the article at http://advanceelectricaltraining.com/iaei/nec-requirements-for-short-circuit-current-ratings/.

Look at the fourth paragraph from the end especially.

Thanks, good article. I'm here to testify that the fourth paragraph is true, based on my personal experience.

Another question I have (besides "why are they allowed to build such large machines with low SCCR's") is this; The available fault current is affected by the equipment I am connecting. Does the SCCR rating take into account the contribution of the rated equipment? If I don't have to account for the motor contribution, the upstream impedance is just barely low enough to sustain operating voltage.

 

[ron]

If I don't have to account for the motor contribution, the upstream impedance is just barely low enough to sustain operating voltage.

If there was a fault at the input terminals of the equipment, current would flow from the utility and the motor to the point of fault. Depending on the portion of the equipment you would be analyzing, it could be both sources or one.

If it were the input disconnect, and the fault were on the load side terminals, so theoretically the fault current going through the input disconnect would only be from the utility. The load side terminals would however have to withstand fault current from both sources.

 

[JoeStillman]

If there was a fault at the input terminals of the equipment, current would flow from the utility and the motor to the point of fault. Depending on the portion of the equipment you would be analyzing, it could be both sources or one.

If it were the input disconnect, and the fault were on the load side terminals, so theoretically the fault current going through the input disconnect would only be from the utility. The load side terminals would however have to withstand fault current from both sources.

Because it's defined right at the point of common coupling, I would hope that the standard they follow stipulates whether the SCCR does or doesn't include the motor contribution. Is this part of the UL standard?

 

[Sahib]

Joe:
You need to worry only about the availability of sufficient short circuit current: if the short circuit current is not sufficient, CLF loses its ability to limit the current within its rating. Otherwise it works fine and you may go ahead per your post # 26.

 

[kwired]

Seems to me this equipment should specify what current limiting fuses should be used, especially if the withstand rating of equipment is under 10kA.

Seems to me they refuse to do proper testing and put a low rating on it for CYA purposes. Passing the burden on to some installer (in particular a design/build installer) that had no idea what they may be getting into with it.

 

[petersonra]

Seems to me this equipment should specify what current limiting fuses should be used, especially if the withstand rating of equipment is under 10kA.

Seems to me they refuse to do proper testing and put a low rating on it for CYA purposes. Passing the burden on to some installer (in particular a design/build installer) that had no idea what they may be getting into with it.

First off, you may or may not be able to use CLF with the components they used to get higher SCCR.

There are some very interesting things in the supplement that describes how the SCCR is calculated in UL508a.

You can add CLF or CLCB to the feeder circuit and you can take some credit for that on downstream devices, but the section that describes how this is done is so badly written I have been afraid to use it.

 

[kwired]

First off, you may or may not be able to use CLF with the components they used to get higher SCCR.

There are some very interesting things in the supplement that describes how the SCCR is calculated in UL508a.

You can add CLF or CLCB to the feeder circuit and you can take some credit for that on downstream devices, but the section that describes how this is done is so badly written I have been afraid to use it.

I guess what I am saying is if the unit were listed with instructions for a specific fuse and because of that a higher SCCR then everyone would be happy - except the manufacturer that don't want to get it listed that way.