Arc flash at 13.2kV

[Volta]

So I will replace 2 switchgear heaters in the HV service disconnects.

Skipping the long story, I didn't trust my first engineer's AF study, so asked another to run some quick numbers to see if my worries were justified. This mini-study shows 220 cal/cm² after (1 of 2) pad mounted transformers (13.2kV:480wye, 2.5MVA). Extreme danger, about 550" for an Arc-Flash Boundary. It is believable to me.

Of course we will do some true calc's this week, but if the secondary is as shown above, what kind of numbers might we be looking at on the HV side, typically, at the disconnects?

The voltage is 13.2Y/7.62KV.
POCO fuses are Kearney T, 100 amps.
The available fault current is 3285 Amps (LLL), 2320 Amps (LG).
I do not know the conductor type/size, but seems to be 1/3 neutral cables. Length is about 80' from pole cutouts to dual switches.

Obviously much more info is needed, these are complex calculations. I'm not looking for official information here, just experienced guesses of incident energy at that point.

 

[jim dungar]

Do not be too surprised if the incident energy on the 13.8kV side is found to below 8 cal/cm?.

 

[Volta]

Do not be too surprised if the incident energy on the 13.8kV side is found to below 8 cal/cm?.

I would be elated. And I thought it would be low on the primary, but didn't know what to hope for.

 

[jdsmith]

I ran a quick study and here are my results along with some assumptions:

13200 L-L transformer primary voltage
480V L-L transformer swcondary voltage
2500 kVA
Padmount, standard ANSI impedance
Kearney 100A type T primary fuse
utilty fault contribution that you listed
no cable lengths or impedances included

36" working distance on primary of transformer
bus gap on primary of transformer typical of switchgear clearances (reasonable guess for primary bushings on a padmount)

24" working distance on secondary of transformer (assuming switchgear, not switchboard)
bus gap on secondary set for switchgear

primary clearing time 0.109 sec, resulting in 0.42 cal/cm^2
no secondary protective device, secondary clearing time assumed as 2 sec, resulting in 75 cal/cm^2

If I change the secondary bus gap to panelboard type clearances instead of switchgear, and 18" working distance instead of 24", I get 127 cal/cm^2 and a 309" arc flash boundary.

From your post I'm guessing you may have two of these transformers paralleled from the same utility source? If so that would explain the 220 cal/cm^2 value.

From the info above you can see why it is critical to get the bus gaps and working distances correct - these have a substantial impact on the resultant incedent energy. If you are having studies run by a consultant be sure to review the bus gaps and working clearances in printed report form before printing and applying labels. Refer to IEEE 1584 tables 2 and 3 for more information on typical bus gaps for different classes of equipment and typical working distances for different equipment classes and voltage levels.

 

[Volta]

Thanks for the help, jdsmith!

The work I am currently planning will be in the primary disconnects proper, not actually at the transformer itself.

The verbal/text riser would be:

POCO 13.2kV cutouts/100a > underground > POCO tap box/primary metering > underground > customer dual disconnects (GE 100S OEM, circa 1977 (pdf is too large to attach)) >>
> T1= 2500kVA::480 volt secondary > 3000 amp GE PowerBreak circa 1987?
> T2= 1000kVA::208 volt secondary > 4000 amp SQD idk circa 1969.

The calc I posted was originally created to show whether or not the first engineer's numbers were trustworthy. The energy values at the 13k switches at this time are not intended to be posted on the equipment, but rather a one-time figure for me to work safely once. A larger study is needed for the whole facility, but I am having trouble accepting the figures given to me from the previous study.

I feel like it is more and more likely that I can use a low HRC suit to deenergize and test this switch. No work is contemplated on the secondary of the transformer.

In your quick unofficial study, was the 0.42 cal/cm^2 at a distance of 36", and was it assumed to be at the transformer?

The calcs for that secondary section would have been based on switchgear, but his clearances may have just been guesses.

 

[jdsmith]

In your quick unofficial study, was the 0.42 cal/cm^2 at a distance of 36", and was it assumed to be at the transformer?

The calcs for that secondary section would have been based on switchgear, but his clearances may have just been guesses.

0.42 cal/cm^2 assumes a 6" bus gap and 36" working distance.
A 6" bus gap and 18" working distance results in 0.82 cal/cm^2

The variable I have not accounted for is short circuit contribution from downstream motors - this could increase the incident energy to some extent depending on what the loads are.

I looked at the NFPA 70E tables to see what HRC catagory the tables would require for this work, and it's much higher because the tables are valid for fault currents up to 35 kA, and assume a clearing time of 0.2 sec. You have a tenth of the fault current and half of the clearing time, so the tables will be extremely conservative.

Here is the guidance from IEEE 1584 on bus gaps and working distances:

Table 2 - Classes of Equipment and Typical Bus Gaps
15 kV switchgear - 152 mm
5 kV switchgear - 104 mm
Low voltage switchgear - 32 mm
Low voltage MCCs and panelbaords - 25 mm
Cable - 13 mm

Table 3 - Classes of Equipment and typical working distances
15 kV switchgear - 910 mm
5 kV switchgear - 910 mm
Low voltage switchgear - 610 mm
Low voltage MCCs and panelboards - 455 mm
Cable - 455 mm
Other - to be determined in field

 

[Volta]

Thanks. There is still a bunch I don't know about this, but I am slowly getting more comfortable with it.

This facility does have some motor loads, maybe the biggest I know of is 100 hp +/-, not more than two of that size, through VFDs, and a handfull of smaller ones for other equip. I guess there is a 50 ton chiller too, with other RTU's, but the equipment use is intermittent, and it's not too hot right now, so I guess the contribution is low. Maybe not too scientific, but . . . :roll: