The local utilities vary a bit in how much info they provide on request. Sometimes they'll include everything upstream of the utility transformer all the way down to the AFC at the service meters, sometimes it's incoming AFC only, and I've seen everything in between. Sometimes they're also not interested in giving me more than they provide initially.

With that established, how do you go about modeling equipment upstream of the point of demarcation when utility info is limited? Is getting the correct AFC at the point of demarcation the most important factor to ensure accurate modeling of the downstream system?

I've previously treated incoming AFC as the crucial element. When the utility has been unable or unwilling to provide additional info, I've usually modeled a dummy utility and utility transformer and tweaked the impedance on the transformer to get the correct AFC at the demarcation point...but I want to make sure I'm not doing something that may introduce inaccuracies in the way the software (EasyPower) does its calculations. Thoughts?

What I've started doing recently with uncooperative power companies is sending them this video:

https://youtu.be/5TPIOAX6Ksk?feature=shared

I'm 2 for 2 now with changing their minds.

Wow, that is very helpful video....for my understanding as well. I'll keep it in my arsenal! Also subscribed to your channel.

I'm also interested in the question of "is AFC enough info to do accurate modeling?" to help me understand the modeling process better. Assuming the AFC provided is accurate, is that enough info to create an accurate model downstream?

Brent,
Nice video, and yes it does make sense.

As someone on the distribution utility side, I'd like to point out some of the issues involved in obtaining and providing "actual" fault current.
IEEE standards allow a tolerance on the transformer nameplate impedances while ANSI allows a range of voltage.
Through faults on all the transformers involved cause impedance to change.
In case of burnout, the service transformer you are basing your calculations on will be replaced quickly with one from stock.
Upstream line impedances are estimates based on a standard construction. Actual line construction varies from span to span and pole to pole.
The system is constantly being improved. Larger conductors, larger substation transformers, etc.
To allow for improvements as well as regular and emergency maintenance/restoration, the system is subject to reconfiguration.
The transmission company that serves us has all the same issues. For example a large portion of our territory is served by two parallel transformers. The company will commonly take one of them out of service without notice, greatly increasing our source impedance.

Especially where you are showing a few amps difference greatly increases clearing time, I would suggest using the longer clearing time.

This is something to consider. I also think the identity of the customer procuring the study should have an influence. If I were doing a study for a small building downtown, where the utility transformers are in a vault off-site, it would make more sense to go with the values that create the highest user-side fault current. My thinking is that the utility may switch gear out without the customer's knowledge.

OTOH, in a facility where the utility transformer is onsite and the customer would be informed of any such changes, using the most accurate info makes perfect sense.

I don't recall if I included this in that video or some other avideo or just shared it with a viewer via email, but transformer impedance has a larger effect on secondary available fault current than the primary AFC and is less likely to change. It's easy enough to prove this to yourself. Set up 9, 1000 kva transformers. Infinite primary, 10,000 amps primary and, say 2000 amps primary with 6%, 5.5% and 5%. I expect you'll find that impedance has the greater effect. By a lot.

Utility nominal, max and min primary amps that only differ by a couple thousand amos on at the primary is basically negligible at the secondary compared to say, a 0.5% difference in %Z.

For the reason Brent provides in his last post, is why I'll use the ANSI +/-7.5% spread in many cases to try to account for variance as stevenal also points out.

Mike

I found the example I was referring to earlier. Not exactly the parameters I described in the experiment, but the message is the same. The primary Nominal / Max / Min are straight from a utility company actual job as is the 5.89 %Z. You can see that only makes a few hundred amps difference on the secondary, even going to infinite bus on the primary.

The example of changing the %Z on the transformer makes 5,000 amps of difference. I think it is more important to get the actual %Z than the actual primary AFC, but both are still important. If you are doing a new design and want to make sure your equipment has adequate interrupting and/or withstand ratings I highly recommend using infinite bus primary and a low %Z. For arc flash purposes, I recommend being as accurate as possible for as-built conditions.

You will find few utilities that will provide actual fault current information on secondary metered customers.

At one time I created a spreadsheet of all of the faults currents provided in the Chicago area, everyone was based on an infinite bus on the primary and the standard %Z for transformers purchased by that utility.
In the Minneapolis are the utilities caved in and would provide a typical fault current, rather than just infinite bus, if they were given a letter that the information was only being used for an arc flash analysis and not for equipment selection or coordination.
In Wisconsin it was not uncommon for the utilitiy to ot even know the size of the transformer at the customer without sending someone out to look.

every utility was cooperative for customers purchasing primary/medium voltage power.

Must be a regional thing. I mostly haven't had many problems in Indiana and Ohio. Duke Energy in my area is the one that uses the typical 5.32 %Z and infinite buss but they will provide actual data if you explain why you need it. AEP/I&M and AES will provide actual data but AES charges you for it unless you're the customer so I try to get the customer to request the data. The small REMC's and municipals are usually the most helpful and easiest to deal with -if they know what you're asking for. There are some that just aren't cooperative at all though too.