I have a question regarding arcing gap discontinuity in CSA Z462 D.8 derived from CSA Z462 tables D.6 and D.7. For visualization purposes, check a plot displaying arc gap as a function of system voltage - all according to tables D.6 and D.7 from CSA Z462, or tables 410-1 and 410-2 from the NESC standard.



The jump at 46kV point is most obvious and alarming. The problem with the 10kV/in number and arc gap calculated based on it is that it is used in calculation the distance from the arc to the worker, namely CSA Z462 Table D.7 Note 2 states the distance from the arc to the worker is calculated using the minimum approach distance from Table 441-2 of IEEE C2 and subtracting two times the assumed arc gap.

Any comments?

Sincerely, M. Furtak, C.E.T.
www.arcadvisor.com

The minimum approach distance from Table 441-2 of IEEE C2 for 46.1 kV (3'-2'') is a lot longer than the 15 inches used for Table D.6 of CSA Z462-08.
So a few inches more or less (from the substraction of 2 times the arc length taken as a flashover at 10 kV/in) won't change much in the end.

Compare the jump from 78.7 cal/cm^2/s for 20 kA at 46 kV in Table D.6 to 12.4 cal/cm^2/s for 20 kA at 46.1 kV in Table D.7. That's because the distance from arc to worker is much longer, so less energy radiates to him.

Also, this discussion should not be targeted at CSA Z462, but at IEEE C2 (where the values in Tables D.6 and D.7 come from, specifically Tables 410-1 and 410-2 in IEEE C2).

"So a few inches more or less (from the substraction of 2 times the arc length taken as a flashover at 10 kV/in) won't change much in the end." [ quoted ]

I've done some analysis ( see below ) and I think it does change a lot in the end.

46kV working distance = 15in.. Gap is equal to 9in. ( from Table D.6 Note 2 )

46.1kV working distance = min approach distance - 2 x gap ( from Table D.7 Note 2 ).

Now, if gap was assumed to be equal to 2.7in. at 46.1kV ( as per Table D.7 Note 2 ) than 46.1kV working distance would be equal to approx 33in. = 38in - 2 x 2.7in. I agree the jump from 78.7 cal/cm^2/s for 20 kA at 46 kV in Table D.6 to 12.4 cal/cm^2/s for 20 kA at 46.1 kV in Table D.7 could probably be justified by larger ( twice as much ) working distance. Even in such case more than 6 times ( 78.7 / 12.4 ) heat flux reduction is quite profound and hard to believe.

Now the logic dictates that the 46.1kV + gap should not be less than 46kV gap ( 9in.). Assuming at least 9 inch gap in 46.1kV system, than according to Table D.7 Note 2 the working distance would be equal to 20in. = 38in - 2 x 9in. that implies 78.7 / 12.4 = 6+ times heat flux reduction by moving just 5 in. = 20in. - 15in. further away from the arc ?

I've put the numbers above together to make easier to visualize the problem I envision with the CSA Z462 section D.8



Comments?

Sincerely, M. Furtak, C.E.T.
www.arcadvisor.com

FYI Annex D in CSA Z462 came from NFPA 70E. The CSA Z462 Technical Committee did not create any of this information.

Regards;
Terrry Becker, P.Eng.
Owner
ESPS Electrical Safety Program Solutions INC.

do you mean you agree there is a problem with arc gap discontinuity in CSA Z462 and NFPA 70E?

- Michael Furtak
www.arcadvisor.com

Arc Calculation Issues including arc gap

There is definitely a problem.

All the data in Z462 and 70E comes from two early papers.

1. Ralph Lee's paper
2. Dick Doughty's paper. Tom Neal, Dan Doan and Dick Doughty did some studies of arc flash in boxes. They did this paper a few times with refinements and we all did it together with some of my PPE research. This was the beginnings of IEEE 1584 but IEEE only spent about $60K on research on the early 1584 and at 15kV there is NO data. It all reverts back to the Ralph Lee equation which is TOTALLY theoretical. If you want something in medium voltage, use ArcPro. Most of those on the 1584 committee are industrially focused. I started in utilities and only ArcPro gives realistic energies with gaps in medium voltage. ArcPro was designed for T&D for Ontario Hydro. I have NO commercial interest in ArcPro. I do most of the arc testing at Kinectrics.

IEEE 1584 has issues in medium voltage which are unlikely to be fixed because the focus of the committee is indistrial. EPRI has done a lot of research and starting more on longer gap arcs. This is being done at Kinectrics. The EPRI committee is focused on the medium voltage systems since the arc flash calculations are overkill when using the Ralph Lee equation. If you look at that equation, arc flash is DIRECTLY proportional to voltage. This is not true. If you don't work medium voltage hot, these calcs don't matter much but they matter to utilities. We use ArcPro with multipliers on medium voltage IF a customer must work it hot.