Arc Flash Forum
https://brainfiller.com/arcflashforum/

Do Switches Need To Be Considered In Arc Flash?
https://brainfiller.com/arcflashforum/viewtopic.php?f=4&t=2057		 Page 1 of 1
Author:  Superman Engineer [ Tue Jan 31, 2012 9:05 am ]
Post subject:  Do Switches Need To Be Considered In Arc Flash?
I am performing an Arc Flash and I'm not sure weather I can include the switch in my studies.. Single line shows that switch is connected to 13.8 kV at one end and a fuse on the other end. Thank You in Advance
Author:  PaulEngr [ Tue Jan 31, 2012 8:37 pm ]
Post subject: 
Can it arc there? Most definitely yes. Do you need pictures to show you what can happen with these things?

That being said, there are a couple considerations. Is it a load break or non load break switch? If it's not under load it might be silly to consider whether or not it is a problem. But when you are making electrical things change states, that is the time when arcing faults are most likely to happen. This is particularly true of load break switches.

There is a problem with current limiting fuses. IEEE 1584 doesn't model them very well because it assumes the fault current is symmetrical and constant at the bolted fault condition, not true with current limiing fuses. There is one model that accounts for this (Wilkins) but it is not in common use and quite complicated to calculate (and missing some data in the paper...what value for VsubE?).

The second consideration is something of a value/judgement call. What's the probability of an arc occurring? Under conditions of proper maintenance and normal operation, 70E-2012 says that the Technical Committee believes that 600 V gear is safe so PPE may be unnecessary (which is what they put in the table version). Above 600 V, they don't say anything. Perhaps you can use the notes in Table 410-1 in IEEE/ANSI C2 to decide since that is the Code for transmission/distribution gear. NFPA 70E really isn't the right Code, and the IEEE 1584 model is pushed right to it's curve fitted limits at 13.8 kV.

I've been kicking around the idea of using IEEE Gold Book data supplemented with some calculations to account for frequency of maintenance to come up with a guideline on whether or not PPE should be required based on equipment reliability but haven't had time to do it.
Author:  engrick [ Mon Feb 13, 2012 12:12 pm ]
Post subject: 
I agree with PaulEngr but I would add this:
Does it require a label? Are there servicable parts - requiring manitenance or operations? If it is only remotely operated and cannot expose anyone then a label MAY not be rquired
Author:  Superman Engineer [ Mon Feb 13, 2012 2:24 pm ]
Post subject: 
I have discussed with my Sr. Engineer about this issue and he said it need not to be consider for the given single line. Thanks for your suggestions
Author:  stevenal [ Tue Feb 21, 2012 4:21 pm ]
Post subject: 
And his reasoning?
Author:  jdsmith [ Sun Mar 18, 2012 6:33 pm ]
Post subject: 
PaulEngr wrote:
I've been kicking around the idea of using IEEE Gold Book data supplemented with some calculations to account for frequency of maintenance to come up with a guideline on whether or not PPE should be required based on equipment reliability but haven't had time to do it.


This has already been done - there was an article in a recent NETA World magazine where someone used Gold Book reliability data to calculate likelihood of an incident and then used the likelihood of an incident to alter the level of PPE required. That being said, it's a really bad idea and a sizeable group of electrical safety professionals disagree with the approach of reducing PPE levels as the likelihood of an incident goes down.

Risk management best practices segregate the likelihood (or probability) of an incident and the severity of the effects of the incident. Once the likelihood of an incident is higher than some given threshold then PPE should be applied based on the severity determined from an arc flash hazard study. If the likelihood of an incident is lower than the threshold then PPE does not need to be applied.

Clearly equipment maintenance history/practices and Gold Book historical data have a quantifiable impact on the likelihood of an incident - I encourage you to work on this area of better defining probability, then come to the Electrical Safety Workshop and share it with others in the industry!

Some good resources for occupational risk management practices applied to electrical safety are the papers that Daniel Roberts has presented at the Electrical Safety Workshop the last few years. Mr. Roberts has a background as a field engineer with a major manufacturer and after being moved into a safety and training role for his field service organization he began studying occupational safety and health quite seriously, I believe earning a degree in this area in the process.

I am working with some collaborators on some of these topics as well. A friend of mine is presenting a paper on these topics at the Petroleum and Chemical Industry Conference (PCIC) in September this year, and we are performing some other research on related areas and expanding his initial work as well.
Author:  Superman Engineer [ Thu Mar 22, 2012 10:13 am ]
Post subject: 
@stevenal!! I don't have asked why???But the reason could be it's not a protection device.
Author:  PaulEngr [ Thu Mar 22, 2012 3:28 pm ]
Post subject: 
I agree 100% with disagreement with the article in NETA World. I had several E-mails with the author. The intent is well meaning but as you rightly pointed out, mixing consequence and likelihood are a really bad idea. I do fault trees and similar types of analysis all the time as part of IEC 61511 baesd risk analysis methods. On an ad hoc basis I've taken a look at the arc flash scenario. There are lots of potential variables but it looks like it can be reasonably quantified. As far as I'm concerned, we can at least quantify the consequence side to roughly 2 cases:
1. Incident energy assuming the upstream protective devices operates correctly.
2. Incident energy assuming that it doesn't. We can either model using the NEXT upstream device or just go for a worst case, possibly assuming 2 seconds using IEEE 1584.

We can also potentially consider some lower risk cases such as the influence of phase angle which reduces or eliminates the assymetrical current at the beginning of the arcing fault.

Depending on the likelihood of either or both scenarios and a given risk aversion (usually 10^-5, 10^-6, or 10^-7 and assuming fatality if the incident energy exceeds 1.2 cal/cm^2 without PPE), it shouldn't be that bad to perform the assessment.
Author:  stevenal [ Mon Apr 02, 2012 9:50 am ]
Post subject: 
Moeen ud Din wrote:
@stevenal!! I don't have asked why???But the reason could be it's not a protection device.

If your name is on the study, I suggest you find out why. Certainly a switch is not a protective device, but it is a probable arc location.
Page 1 of 1 All times are UTC - 7 hours
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/