Fuse Modelling for Arc-Flash Analysis to IEEE 1584-2002

[EddyWirbelstrom]

For arc-flash analysis the fuse ‘total clearing time’should be used.
IEC Fuses
In the IEC world, fuse manufacturers publish fuse time-current characteristics which plot the ‘pre-arcing time’ as a function of rms prospective current. IEC 60269-1 states that for times longer than 0.1 sec, the difference between ‘pre-arcing time’ and ‘operating time’ is negligible.
The ‘operating time’ is the total clearing time which is the sum of the ‘pre-arcing time’ plus the arcing time.
American Fuses
In Amercian fuse catalogs I have seen fuse time-current characteristics plotted as :
- ‘minimum melting time’,
- ‘minimum melting time’ and ‘total clearing time’.
as a function of rms prospective current.
Questions
IEC Fuses
• What +/- tolerance should be used for the ‘pre-arcing time’ ?
• What additional time should be added for times below 0.1 sec for the ‘operating time’ ?
American Fuses
• What +/- tolerance should be used for the ‘minimum melting time’ ?
• Where only the ‘minimum melting time’ is published, what time should be added for the ‘total clearing time’?

 

[davidbeach]

Good luck with that. No idea about the IEC stuff. In ANSI practice, the all tolerances on minimum melt are positive. That sounds useful until you consider that the minimum melt time is for a "cold" fuse. The testing to determine the curves is done without any load current prior to the fault current. Load current (and/or ambient conditions) that partially heat the fuse prior to the application of the fault current will speed up the melting and clearing. If all you have is minimum melt and you want good clearing times you'll have to contact the manufacturer. You can probably find representative curves for general families of fuses, but those are better for estimating coordination margins than for detailed arc flash hazard calculations.

At least for arc flash all the uncertainty caused by preloading is on side of making things faster and therefore less hazardous. But for coordination the preloading makes the task more difficult. S&C has some good publication on preloading and other curve adjustments, but generally for power fuse (medium- and high-voltages) rather than for the 600V class fuses I'm guessing you're more interested in.

 

[dpc]

I agree with David, for arc-flash calculations, the main concern is the total clearing time. Pre-loading is only going to reduce the calculated incident energy. But the problem is worse than that for low voltage fuses. For many low voltage fuses the manufacturers in the US only provide a single line - "average melt" curve. We have to estimate a total clearing time tolerance from that. In addition, they no longer even provide this information for all fuse sizes. We have to interpolate for fuse sizes that they do not provide curves for.

Also, IEEE 1584 provide separate equations for determine arc energy downstream of certain low voltage current limiting fuses as an alternate method. We don't use these equations normally, but they are optionally available in most of the commercial software that is doing arc-flash calculations.

 

[EddyWirbelstrom]

Thank you davidbeach and dpc for your helpful response.

Where fuses other than those for which IEEE 1584-2002 provide empirical formula for arc-flash energies ( Class L and Class RK1 ) are installed , PTW arc-flash software recommends following fuse total clearing time for generic current limiting fuses :
a) The current-limiting range is assumed to start where fuse clearing curve drops below 0.01 sec.
b) Fuses operating in the current limiting range are assumed to clear in ½ cycle for currents 1 to 2 times the current where the current-limiting range begins, and ¼ cycle for currents higher than 2 times the current where the current-limiting range begins.

For Class L and Class RK1 fuses, PTW arc-flash software has the option of using the equipment empirical formulas.

My concern is for generic current limiting fuses 1000V and below where only the average melt time or average pre-arcing time is published and where the current is below the range defined in a) and b) above.

I am assuming the following fuse total clearing time:
For average melt times below 0.10 sec, add 15% to average melt ( pre-arc ) time
For average melt times above 0.10 sec, add 10% to average melt ( pre-arc ) time.

PTW documentation recommends :
For average melt times below 0.03 sec, add 15% to average melt ( pre-arc ) time
For average melt times above 0.03 sec, add 10% to average melt ( pre-arc ) time.

 

[dpc]

Is there a question in there somewhere? When you only have the average melt curve, you have to make assumptions.

 

[healyx]

Hi Eddy,

Did you ever find a definitive source to confirm the the +10% and +15% tolerances?

 

[dpc]

IEEE 1584 says add 10% plus 0.004 seconds to average melt time. If total clearing time is less than 10 ms, use 10 ms for the time. This is the latest guidance, contained in Amendment #2 in August 2011.