Arc Flash - Extent of Study

[EP007]

I'm wondering how far all the experts are presently taking the arc flash analysis to. I have a 480 V MCC which supplies loads such as VFD's, motors, unit heaters. I've been on sites where tests were made at each of these, with covers off. The hazard could therefore exist, which implies analysis made at terminals of all those items.

 

[alehman]

You need to calculate the arc flash at the point(s) where the work is to be performed.

 

[Zogzog]

Most studies go down to the point mentioned in IEEE 1584, <240V, <125kVA source.

 

[EP007]

That means that we don't stop at the 480 V MCC, but perform the analysis up to the terminals of all 480 V equipment, which means at every VFD, motor, and unit heater terminal. Note: 208 V is not part of my question.

 

[dpc]

If maintenance is ever done while the equipment is energized, then you need to consider it. Some end use equipment may always be de-energized before any covers are removed - such as motors.

We do not generally put arc-flash labels on motors.

 

[Eleceng01]

dpc, do you evaluate a local disconnect to the motor if it is present?

 

[dpc]

It depends on the site. Often it is decided to not label the disconnect switch since there is never a reason to open the cover when the circuit is energized. And in general, the disconnect is not operated when the motor is running. In an emergency, no one is going to stop to put on PPE before operating.

I know a case can be made for labeling the disconnect switches, but we generally try to avoid it.

 

[EP007]

The scope would be limited to the Arc Flash Analysis. That would also include short circuit and protection coordination. Decisions on implementing the safety program and on labels, would be by the plant owner. As I cannot anticipate nor control what they will do or how, I think it is best if I generate the numbers for all locations. From my previous experience with engineering measurements, I can imagine a need to measure at any terminal in a non-hazardous location, setup in a de-energized state, then, with covers off, the circuit is energized.

 

[jghrist]

The following paper presents a method to limit the scope of studies through generalizations:

Arc flash hazard calculations

Tinsley, H.W. Hodder, M. Graham, A.M.
Eaton Electr., Warrendale, PA;

This paper appears in: Industry Applications Magazine, IEEE
Publication Date: Jan.-Feb. 2007
Volume: 13, Issue: 1
On page(s): 58-64
ISSN: 1077-2618
INSPEC Accession Number: 9243927
Digital Object Identifier: 10.1109/MIA.2007.265802
Date Published in Issue: 2006-12-26 08:38:30.0

Abstract
This article discusses a method for generalizing the arc flash calculations for select areas of a power distribution system. The method presented here helps to reduce the number of equipment locations and also calculations which are analyzed using a commercial software package. In an industrial example, this method reduced the number of small feeder locations that required a detailed analysis by 94%. This resulted in a large reduction of onsite data collection, detailed modeling, and arc flash warning labels

 

[dpc]

You can do whatever you want, obviously. But if you are planning on modeling and analyzing every 480 V feeder down to each motor, you'd better put a lot of hours in your data collection and modeling estimate.

As a practical matter, 480 V feeders downstream of a molded-case circuit breaker or current-limiting fuse will have arc-flash levels less than 8 cal/cm2 and probably less than 4 cal/cm2 except for situations with very low available fault current.

 

[EP007]

DPC, the max fault on 480 V is about 9.3 KA sym. It is a big problem to present the data in a concise manner, while not over-looking hazards. I am not doing "at the motors" for this one, but there is only two DOL anyways, at 480 V. For a large system, this could really get out of hand.

jghrist, I found that article very useful.

 

[burnt2x]

EP007,

"That would also include short circuit and protection coordination."

I see two issues that you may have considered synonymous, when they are in fact opposite! Protection coordination is the opposite of arc flash protection. When we try to limit the arc energy, we do this by making the tripping time as short as possible, but when we try to coordinate trip times, we try to "delay" some protection setting to let the breaker nearer the fault to clear first.

 

[unclebob]

burnt,

Very true. A consultant asked me, not long ago, a coordination study to reduce the energy at some switchgear...

 

[davidbeach]

If you are willing to spend enough it is possible to have a fully coordinated protection system that will trip instantaneously (or nearly so) for all faults, but only trip the minimum amount necessary; thus achieving both coordination and minimum arc energy. But to do this you will have to use more than just independent overcurrent devices. Differential zones and blocking schemes will be critical components.