fault current calcs

Status
Not open for further replies.
S

steve brandt

Member
Location
lighthouse point, fl
Occupation
famous architect
Please help me understand what wire size and length to use in arc-fault calcs when the feeders vary between the transformer and the service equipment - this is typical in overhead feeders where the meter to masthead may be copper and the masthead to transformer may be aluminum - same thing happens with underground feeders where there is a hand hole at the bottom of the power pole and the meter to hand hole may be copper while the transformer to hand hole is aluminum
thanks
 
You have to go point by point. Start with the transformer to the meter. Then the meter to the disconnect, then the disconnect to the panel. For each piece, your starting fault current will be the resultant fault current of the previous point, and each time you use the respective length and the respective wire size feeding the equipment.

Example only (not calculated)

Transformer starting fault : 32 kaic
Conductor: # 600 copper, 50 feet to meter
Meter: 25 kaic
Conductor: #600 copper, 2 feet
Disconnect: 24 kaic
Conductor: #3/0 copper, 30 feet
Panel A: 9 kaic



So you basically just do it section by section until you reach the end. If you’re actually designing and not doing a test problem then most of this is theoretical based on inaccurate data provided by the power company using highly conservative worst case values and you’re best to error on the side of caution when it’s close because things change.




Sent from my iPhone using Tapatalk
 
Npstewart said:
You have to go point by point. Start with the transformer to the meter. Then the meter to the disconnect, then the disconnect to the panel. For each piece, your starting fault current will be the resultant fault current of the previous point, and each time you use the respective length and the respective wire size feeding the equipment.

Example only (not calculated)

Transformer starting fault : 32 kaic
Conductor: # 600 copper, 50 feet to meter
Meter: 25 kaic
Conductor: #600 copper, 2 feet
Disconnect: 24 kaic
Conductor: #3/0 copper, 30 feet
Panel A: 9 kaic



So you basically just do it section by section until you reach the end. If you’re actually designing and not doing a test problem then most of this is theoretical based on inaccurate data provided by the power company using highly conservative worst case values and you’re best to error on the side of caution when it’s close because things change.




Sent from my iPhone using Tapatalk
Click to expand...
This is completely equivalent to calculating the total series resistance of the path from transformer to target. Then if using a modelling program you
can input an equivalent single wire size path with that same total resistance. All of this disregards any AC impedance, such as that introduced by current limiting coils.
 
Thanks - that is a lot of help
a few questions though
your first calc 32Kaic..... i get = 21.6 kaic - not 25
i get the c value at 28033
do you have another value and if so what is your source
i believe i use bussman chart values
 
steve brandt said:
Thanks - that is a lot of help
a few questions though
your first calc 32Kaic..... i get = 21.6 kaic - not 25
i get the c value at 28033
do you have another value and if so what is your source
i believe i use bussman chart values
Click to expand...
You never ever calculate an AIC.
You can only calculate the amount of short circuit amps (SCA) available at a point in the system. You then select equipment with an AIC, or short circuit rating (SCCR), higher than the available fault current at its line side terminations.

For example: you just calculated an available SCA of 21.6kA so your equipment needs to be higher than this value, say something like 25kAIC.
But if you are doing arc flash calculations you use your calculated SCA.
 
steve brandt said:
Thanks - that is a lot of help
a few questions though
your first calc 32Kaic..... i get = 21.6 kaic - not 25
i get the c value at 28033
do you have another value and if so what is your source
i believe i use bussman chart values
Click to expand...
the poster you replied to didn't do a calculation, only was illustrating a point that the AFC will typically decrease as you go downstream. this might not be true if there are large electric motors (over 100 HP)
 
steve brandt said:
Please help me understand what wire size and length to use in arc-fault calcs when the feeders vary between the transformer and the service equipment - this is typical in overhead feeders where the meter to masthead may be copper and the masthead to transformer may be aluminum - same thing happens with underground feeders where there is a hand hole at the bottom of the power pole and the meter to hand hole may be copper while the transformer to hand hole is aluminum
thanks
Click to expand...
One problem with point to point is the omission of POCO impedances that could lower those values in the equation.
Point to point assumes your right beside the substation with infinite values. No line impedances at all. If your far from the station that will change your values, depending on system wire size and construction.
 
Hv&Lv said:
One problem with point to point is the omission of POCO impedances that could lower those values in the equation.
Point to point assumes your right beside the substation with infinite values. No line impedances at all. If your far from the station that will change your values, depending on system wire size and construction.
Click to expand...
The PoCo could help. A call to their offices is in order. But if you don't have the figures, you have to trace back the lines to the source and estimate the parameters of the transmission lines to arrive at a reasonable figure (if you are capable of doing the mathematical gymnastics).
 
Hv&Lv said:
One problem with point to point is the omission of POCO impedances that could lower those values in the equation.
Point to point assumes your right beside the substation with infinite values. No line impedances at all. If your far from the station that will change your values, depending on system wire size and construction.
Click to expand...

That’s infinite bus assumption. There is a subtle difference here. The infinite bus calculation assumes that there are no large inductances and only the transformer impedance matters.

Generally speaking utilities are modeled in SKM, ETap, Easypower, etc., as an ideal voltage source and a reactance, nothing more. This describes the infinite bus assumption.

Point to point starts there but then models each bus/cable leading to the load. It works as long as there are no significant load reactances.
 
Does anyone know of an on line course that will methodically take me through how these calculations are done and the theory behind them? I have looked but all I can find is training aimed at fulfilling CE requirements rather than actually teaching me something I need to know.
 
paulengr said:
That’s infinite bus assumption. There is a subtle difference here. The infinite bus calculation assumes that there are no large inductances and only the transformer impedance matters.

Generally speaking utilities are modeled in SKM, ETap, Easypower, etc., as an ideal voltage source and a reactance, nothing more. This describes the infinite bus assumption.

Point to point starts there but then models each bus/cable leading to the load. It works as long as there are no significant load reactances.
Click to expand...
Ours has reactance and impedances from the sub XF out.
If you were on our system and you call me I’ll put your stuff in our model and see what it spits out.
Not really sure why other POCOs wont do readily do that.
If the system is modeled correctly it’s extremely easy to get these numbers
 
Hv&Lv said:
Ours has reactance and impedances from the sub XF out.
If you were on our system and you call me I’ll put your stuff in our model and see what it spits out.
Not really sure why other POCOs wont do readily do that.
If the system is modeled correctly it’s extremely easy to get these numbers
Click to expand...
It would be simple if the system is just linear, with no embedded generations connected at various points along the way and no loops. That said, a simple way to get things done is to call the PoCo and ask for the available fault current from their system at your point of common coupling. I have experienced using the infinite source assumption and compared it with a calculation using the data from the PoCo. Found out that the fault at my load side was twice the value on the infinite source assumption compared to the calculation using the PoCo data.
 
Status
Not open for further replies.
Top