AIC Calculations & It's Significance

[Ingenieur]

For example, in Washington, they adopt the NEC. There are state amendments/additions and they are either in the RCW (revised code of washington) or the WAC (Washington administrative code). There are some local codes too such as the Seattle electrical code which is made law by the city council. If an inspector wants something, he has to give me reference from one of the aformentioned documents.

So you are claiming this?

inspector/plan reviewer: the AIR of your equipment is too low
Sparky: No its fine, here are my calculations. See? Here is a picture of the transformer data plate.
inspector/plan reviewer: You need to use the figure provided by the utility.
Sparky: can you give me a code section requiring that?
inspector/plan reviewer: Its on that checklist on that paper discussing the plan review process.

IMO that is not legally acceptable. It needs to be formally stated in a legal document that has been adopted into law.

FWIW, I would predict they would be fine with an actual calculation. They probably just word it that way because they assume that is the route most people go.

yes

sure is 'legally acceptable' or legal
again, not law but code

the governmental body authorizes the code officer to do this
you can take it to court
good luck
he won't be around long if his decisions are overturned
and usually not court but a code board appointed by the muni
then court, but usually arbitration or a higher board/commission

wrong
they REQUIRE use of the pg&e number, explicitly, and in great detail
they want the letter on the dwg
they will stand by it

 

[electrofelon]

I admit I have seen very few figures that have taken into account primary impedance (the one that was half I mentioned was from a MH thread A while back). What are some reduction values you guys have seen?

....Hugely variable of course, I know that. Where I am, 8 miles from substation, 4.8 KV ( not sure if the whole way), Probably would be a substantial impact considering primary impedance. Other end of the spectrum, network distribution, etc. not so much.

 

[Ingenieur]

Some distribution circuits can significantly lower the fault at the spades- especially a 4160 volt system which has a good chance of being here considering the POCO. Even still, 12 and 16kv system scan still exhibit a good increase in impedance with distance.

This is a utility application.


The NEC- in so far they have yet to clear up the mud and fog on fault current.

:roll: It is well known that service will not reach 800amps.

not in my experience

utilities don't use generators?

300 kva at 208 = 800 A
but in generally a 300 may serve 500 load
so yes, it will see greater than 800 A

what is v drop
100'
23'

 

[mbrooke]

not in my experience

Please elaborate what you mean here. I have no idea what you are picturing or have experience with.

utilities don't use generators?

There are no wires, transformers and physical distance between load and generators?

300 kva at 208 = 800 A
but in generally a 300 may serve 500 load
so yes, it will see greater than 800 A

what is v drop
100'
23'

Again- come real world that transformer will probably me more like 100kva. If NEC load calcs are yielding a 800amp service for an actual 800amp load, then something is seriously wrong.

 

[Ingenieur]

Please elaborate what you mean here. I have no idea what you are picturing or have experience with.

There are no wires, transformers and physical distance between load and generators?

Again- come real world that transformer will probably me more like 100kva. If NEC load calcs are yielding a 800amp service for an actual 800amp load, then something is seriously wrong.

no
with you it becomes an endless cycle of questions answered with questions

what?
how is power tranferred without wires?

the example was 3x100 kva ~ 830 a at 120/208
are you saying a 100 kva would be used for an 800 a continuous load?

ok
500 kva
480 vac
600 a
5% pu
i fault 12 ka at xfmr lugs

continuous load 500 a
assume 1000 kcmil Cu
in steel conduit

how long a run to drop i fault to 6 ka? 1/2 of the xfmrs
hint conductor total loop Z = 24/600 = 0.04 Ohm
what is v drop at that length?

 

[mbrooke]

no
with you it becomes an endless cycle of questions answered with questions

Because you give nothing to back up your claims when its not substantiated outside of an equation. You say in your experience. What experience has lead you to believe that transformer will not be smaller? What experience has shown that the source (primary) impedance is infinite or near infinite?

On the other hand I can confidently say that in nearly all applications POCOs drop a grossly undersized (in relation to the service) transformer and service drop (be it underground or over head). I can also confidently say a 4kv, 12kv, ect feeder is far from infinite. Even near the substation there can be significant impedance (ie, 12kv system fed from a 12 MVA unit off a 34.5kv system instead of a 230kv 60MVA transformer) Perhaps you deal with strong sources, but come real word 9 out of 10 its not.

what?
how is power tranferred without wires?

Exactly my point. Any wire will have both resistance and reactance when carrying power- including a 12kv (or which ever) feeder.

the example was 3x100 kva ~ 830 a at 120/208
are you saying a 100 kva would be used for an 800 a continuous load?

Yes

ok
500 kva
480 vac
600 a
5% pu
i fault 12 ka at xfmr lugs

continuous load 500 a
assume 1000 kcmil Cu
in steel conduit

how long a run to drop i fault to 6 ka? 1/2 of the xfmrs
hint conductor total loop Z = 24/600 = 0.04 Ohm
what is v drop at that length?

And your primary impedance?

 

[Ingenieur]

Because you give nothing to back up your claims when its not substantiated outside of an equation. You say in your experience. What experience has lead you to believe that transformer will not be smaller? What experience has shown that the source (primary) impedance is infinite or near infinite?

On the other hand I can confidently say that in nearly all applications POCOs drop a grossly undersized (in relation to the service) transformer and service drop (be it underground or over head). I can also confidently say a 4kv, 12kv, ect feeder is far from infinite. Even near the substation there can be significant impedance (ie, 12kv system fed from a 12 MVA unit off a 34.5kv system instead of a 230kv 60MVA transformer) Perhaps you deal with strong sources, but come real word 9 out of 10 its not.


Exactly my point. Any wire will have both resistance and reactance when carrying power- including a 12kv (or which ever) feeder.


Yes



And your primary impedance?

not accurate
you like to conflate and debate, it serves no purpose, you make up facts

??????
it is not whether is has Z, but the magnitude

wrong
they would not use a 100 kva for a continuous 300 kva load
silly, 300% continuous overload

assume infinite bus, ie, source >>>>> load
that should allow you to answer the question and convince you in a properly sized conductor (amapcity and v drop) fault current will not be halved
in the op's case it went from 56 ka to 53 iirc

 

[mbrooke]

not accurate
you like to conflate and debate, it serves no purpose, you make up facts

??????
it is not whether is has Z, but the magnitude

In so far, what do you feel I have made up? Perhaps we can re-solve that.

wrong
they would not use a 100 kva for a continuous 300 kva load
silly, 300% continuous overload

The load never reaches 300kva in reality- so whats the point in doubling or tripling a transformer?

assume infinite bus, ie, source >>>>> load
that should allow you to answer the question and convince you in a properly sized conductor (amapcity and v drop) fault current will not be halved
in the op's case it went from 56 ka to 53 iirc

A good (realistic) fault study never assumes infinite bus.

 

[Ingenieur]

In so far, what do you feel I have made up? Perhaps we can re-solve that.

The load never reaches 300kva in reality- so whats the point in doubling or tripling a transformer?

A good (realistic) fault study never assumes infinite bus.

pretty much everything
all non real world scenarios

continuous load = 800 a 300 kva
it always is 300 kva
still use a 100 kva?

a conservative one does if util data not available
have you ever done one?
how many feet in my first example? hint 500' v drop 5%
so in real world examples the conductor will not reduce i fault by 50%
since you can't answer that one try this

prim avail fault 10 mva 13.2 kv x/r ~ 12/1
(for ref: op's is on the order of 20 mva for a 0.3 mva load, 70 x's, est total pu Z 1.5%)
assume xfmr 500 kva 480 vac pu 5% use ph-ph fault
using i rated sec 601 a what is the pu i fault at the sec lugs using util data?
what is the pu i fault assuming infinite bus?
for a bonus what is the sec pu v (480 basis) under a bolteed fault?

 

[mbrooke]

pretty much everything
all non real world scenarios

Then I would argue you know little about utility, residential and commercial power systems- or at least trying to come across that way. Now thats not to say I am discrediting what you have seen or what you do on the job- but outside of that things are very different.

continuous load = 800 a 300 kva
it always is 300
still use a 100

If the load is indeed a continuous 300kva- and no transformer loss of life is being factored. However, a typical 800amp service will not be pulling 800amps- let alone continuously. Not saying it can't happen- but in most cases it does not.

a conservative one does if util data not available
have you ever done one?

Sure- find the Z and Kva of said transformer. Assume infinite primary. Usually you can't go wrong. But chances are you might end up spending more $$$$ on high AIC equipment. So its best to actually factor in the source.

how many feet in my first example? hint 500' v drop 5%
so in real world examples the conductor will not reduce i fault by 50%
since you can't answer that one try this

prim avail fault 10 mva 13.2 kv x/r ~ 12/1
xfmr 500 kva 480 vac pu 5% use ph-ph fault
using i rated sec 601 a what is the pu i fault at the sec lugs?
what is the pu how many feet in my first example? hint 500' v drop 5%
so in real world examples the conductor will not reduce i fault by 50%
since you can't answer that one try this

prim avail fault 10 mva 13.2 kv x/r ~ 12/1
xfmr 500 kva 480 vac pu 5% use ph-ph fault
using i rated sec 601 a what is the pu i fault at the sec lugs?
what is the pu
for a bonus what is the sec pu v (480 basis) under a bolteed fault?
for a bonus what is the sec pu v (480 basis) under a bolteed fault?

Honestly- I don't want to spend the time to crunch the numbers- especially Positive, negative and zero sequence components since most faults are L-G an not L-L-L. Yes L-L-L might be enough- but often you need to do all 4 (L-L, L-L-L, L-L-G, L-G) to know where you stand.

But- going back to what Sylvester the cat said, the NEC really is not specific on how those numbers are obtained so pretty much anything might go.

 

[Ingenieur]

Then I would argue you know little about utility, residential and commercial power systems- or at least trying to come across that way. Now thats not to say I am discrediting what you have seen or what you do on the job- but outside of that things are very different.

If the load is indeed a continuous 300kva- and no transformer loss of life is being factored. However, a typical 800amp service will not be pulling 800amps- let alone continuously. Not saying it can't happen- but in most cases it does not.

Sure- find the Z and Kva of said transformer. Assume infinite primary. Usually you can't go wrong. But chances are you might end up spending more $$$$ on high AIC equipment. So its best to actually factor in the source.

Honestly- I don't want to spend the time to crunch the numbers- especially Positive, negative and zero sequence components since most faults are L-G an not L-L-L. Yes L-L-L might be enough- but often you need to do all 4 (L-L, L-L-L, L-L-G, L-G) to know where you stand.

But- going back to what Sylvester the cat said, the NEC really is not specific on how those numbers are obtained so pretty much anything might go.

I know enough to be dangerous
regardless, what I postulate is true
if you worked my first example excercise you would see

perhaps not clear enough
I only have 30 years experience 50/50 field/office, not sure they can exist independently
800 a 208 v continuous load, factory motors
is a 100 kva xfmr sufficient?

you don't need sym components
it would take 5-10 minutes max of algebra
give it a shot, it would be very enlightening for you imho
you spent more time on this post

 

[mbrooke]

I know enough to be dangerous

Oh- I know :happyyes::happyyes:

regardless, what I postulate is true
if you worked my first example excercise you would see

perhaps not clear enough
I only have 30 years experience 50/50 field/office, not sure they can exist independently
800 a 208 v continuous load, factory motors
is a 100 kva xfmr sufficient?

you don't need sym components
it would take 5-10 minutes max of algebra
give it a shot, it would be very enlightening for you imho
you spent more time on this post

Everything equation wise is 100% true- and it is 100% for continuous loads- and yes your fault current equations can technically be sufficient. But like many posts, you refuse to see the most obvious. The obvious is this: that service will most likely never draw 800amps continuously.

And yet this post took even more time- but go look at any service disconnect rating and then look at the pad/pole pig. 98% of the time that pig or pad will be 1/2 to 1/3 the service. And when its not, good chance it will be serving other customers.

 

[electrofelon]

800 a 208 v continuous load, factory motors
is a 100 kva xfmr sufficient?

It rather annoying how you twist things around and ignore what people are saying. Like in the beginning regarding the 54K fault current I said possible but not likely. Then you jumped all over how it is perfectly reasonable. For that to be reasonable, the utility would have to have transformers closely matching the NEC service size which rarely happens. Now above in what I quoted you are talking "continuous factory motors". Large motor in industry probably running near the high end of its rating, of course the utility wouldn't cheat that down as much as they would for a machine shop with a few 25 HP VMCs using a quarter inch end mill on aluminum, or NEC calculated stuff that has multiple fudge factors. The former is not typical of most installations for most of us. Of course no one is saying it is ok to power an actual 800 amp load with a 100 KVA transformer (800 amp NEC calculated sure). No one is saying fault current values provided by the utility are NEVER reasonable. No one is saying there is NEVER a service where utility KVA matches service KVA. .

 

[Ingenieur]

Oh- I know :happyyes::happyyes:

Everything equation wise is 100% true- and it is 100% for continuous loads- and yes your fault current equations can technically be sufficient. But like many posts, you refuse to see the most obvious. The obvious is this: that service will most likely never draw 800amps continuously.

And yet this post took even more time- but go look at any service disconnect rating and then look at the pad/pole pig. 98% of the time that pig or pad will be 1/2 to 1/3 the service. And when its not, good chance it will be serving other customers.

since you thought you needed sym comp and can't solve either of my examples I now understand your basic knowledge level
It shines light on why you ask the questions you do
no disrespect meant but we are at 2 different levels

 

[Ingenieur]

It rather annoying how you twist things around and ignore what people are saying. Like in the beginning regarding the 54K fault current I said possible but not likely. Then you jumped all over how it is perfectly reasonable. For that to be reasonable, the utility would have to have transformers closely matching the NEC service size which rarely happens. Now above in what I quoted you are talking "continuous factory motors". Large motor in industry probably running near the high end of its rating, of course the utility wouldn't cheat that down as much as they would for a machine shop with a few 25 HP VMCs using a quarter inch end mill on aluminum, or NEC calculated stuff that has multiple fudge factors. The former is not typical of most installations for most of us. Of course no one is saying it is ok to power an actual 800 amp load with a 100 KVA transformer (800 amp NEC calculated sure). No one is saying fault current values provided by the utility are NEVER reasonable. No one is saying there is NEVER a service where utility KVA matches service KVA. .

isn't it 56.3 ka at the xfmr?
nothing was twisted, the convo evolved well beyond the op

a 225 kva at 1.2% would do it

surprised niether of you 2 solved my example
1 length to reduce fault by 1/2
2 inf bus fault
3 util paramters given fault

where's the curiosity?
you guys are convinced I know nothing lol
it would show me your level of understanding of the concepts and WHY

I like solving/discussing systems with numbers and solutions
not hypothetical "what could be's"
that way we can follow the steps to see where we stumble

 

[electrofelon]

a 225 kva at 1.2% would do it

Yes it would. But the transformer size is higher than typical, and the impedance is very very low. Maybe the OP has a confluence of those two things. Fact is utilities give unrealistic fault values most of the time - in my experience.

surprised niether of you 2 solved my example
1 length to reduce fault by 1/2
2 inf bus fault
3 util paramters given fault

where's the curiosity?
you guys are convinced I know nothing lol
it would show me your level of understanding of the concepts and WHY

I like solving/discussing systems with numbers and solutions
not hypothetical "what could be's"
that way we can follow the steps to see where we stumble

I am not an engineer. I do have a math degree so I can follow your math no problem, but some electrical theory is either above me and/or I dont find it interesting enough. I was in a PhD math program but I dropped out because I am to picky about what I find interesting LOL. Calculus, fun; abstract algebra, boring; analysis, tedious..... I am curious about values obtained by taking primary impedance into account, just not real interested in the process to get there so Ill leave that to you.

 

[mbrooke]

since you thought you needed sym comp and can't solve either of my examples I now understand your basic knowledge level

For a real fault study with accurate numbers, you do.

It shines light on why you ask the questions you do
no disrespect meant but we are at 2 different levels

Its the other way around- I am asking questions because you have not shown a solid understanding of utility practices hence why I am trying to gauge your knowledge level while simultaneously trying to evoke critical insight. Much like me asking a person "then why does code only require #6 to (a ground rod) but a much larger gauge for else-where in article 250" who claims a ground rod's purpose is to open an OCPD.


Again- you took 800 amps and automatically derived 300kva. The math is 100% correct in terms amps x volts= VA, but its far away from how utilities actually size transformers and that is the point I am trying to get across.

Service size ≠ to transformer size. Period.

 

[topgone]

Service size ≠ to transformer size. Period.

Agree.
Besides, utilities have their own way of sizing transformers based on their huge experience in the business. I was told they base their computations on transformer loss of life and temperature rise calculations in part.

 

[kwired]

Agree.
Besides, utilities have their own way of sizing transformers based on their huge experience in the business. I was told they base their computations on transformer loss of life and temperature rise calculations in part.

Yes. You may find situations where something like the 800 amp service in question here is heavily loaded for say 6-8 hours a day , but is lightly loaded the rest of the day - this gives transformer time to cool down again - they factor that in when selecting a transformer. Now if that same service is heavily loaded 24 hours five days at a time they also consider that and will likely size it much closer to actual KVA being used.

 

[mbrooke]

Agree.
Besides, utilities have their own way of sizing transformers based on their huge experience in the business. I was told they base their computations on transformer loss of life and temperature rise calculations in part.

Yup- thats how its done- which can further reduce size on top of everything else. Here are a few factors:


1. utilities know the real (measured) load rarely approaches the actual service rating. This is based on what is being served and how its used. NEC load calcs are grossly conservative and actual rating stickers on equipment is often not far behind. Not to mention you have equipment thermostatically cycling relative to on another (load diversity).


2. How long is the peak load, and is the loading asystole (flat line- term I use lol), rolling hill or a saw tooth as plotted over a 24 hour period? The greater the discrepancy between base load, average load and the peak load, the more overloading is done. Ie a sewage plant or 24 hour store might get a pad mount with only a 115% over load for peak, while a home might do a 300% overload during peak. This is due to the fact the transformer oil gradually cools when a dwelling is only drawing a few amps during the day or at night (everyone at work or sleeping) and absorbs heat (preventing core damage) while occupants are cooking, showering, taking the thermostat out of set back, doing laundry and so forth when waking up or coming home from work- a peak that usually takes place 1 to 3 hours in duration and then rapidly falls off. A load that is more steady does not allow for this so less overload capacity is available.


3. Is the customer a summer or winter peaker? A winter peaker, especially when from electric heat, allows a smaller unit since temperature dissipation at 5*F with 2ft/sec wind is greater than at 95*F with sun beating down.

4. The equipment does not have to last 60+ years. 30 years is often money well spent for pole pigs. A unit can purposefully be over loaded such that it will last X years but odds go down after Y. As someone else on here who knew a POCO guy put it "if you don't burn up a few occasionally, you are over sizing them" In fact fused cutouts are often not intended to protect transformers from overload (they could be though if desired) but rather to remove a failed unit from service without tripping out the entire feeder.