When you have a Wye/Wye bank of 3 transformers each with a different impedance which impedance do you use for the purpose of the arc flash model? In my case I have impedances of 6.49, 6.61, and 6.53. Do you use an average which would be 6.54 or do I use the lowest value of 6.49?

Are these single phase cans wired for 3 phase? The impedance of each can seems kind of high for that but it looks like you must have nameplate.

If you're using the info to get secondary side fault currents just for an arc flash study I'm not sure it will make much difference but I have no idea what kva and voltage and available fault current and your system X/R etc. is. Run all 3 values in your model and see.

It may be possible to model each single phase can in your software if that is what you have but I don't have that ability with mine so I can't try it.

It is an older transformer...not far away from being replaced. It is set to it's lowest possible tap setting of 12.54 kV for the primary but is being supplied with 12.47 kV. I do have nameplates for each can identifying the impedance's as I listed them. There is a couple calorie difference on my low voltage switchgear when I model using the lowest imp versus modeling the average imp. Not a major deal since both are around 200 cal/cm2. I'm just really curious what is the correct method.

Sorry I can't help you but 200?

Are you using the "2 second rule" and still that high?

Yes, 200.3 cal/cm2 to be exact and with the 2 second rule. These values are seen at the low-voltage switchgear which is directly connected to the secondary of the transformer. Luckily, in this case we just replaced the switchgear and all buckets have ARMS. With ARMS activated I can get the IE down to 4 cal/cm2 for all feeder buckets.

IEEE 1584 is only accurate to about +/-15% so realistically you are arguing about the 3rd decimal place in a model that is only accurate to 1-2 decimal places. The impedance that you use doesn't matter.

We use an average of the three impedances.

Something isn't right here. To get 0.75 ohms you have to have two really good Earth electrodes. That's substation level ground grids, not a typical ground rod or a pole butt ground. I would typically drive 30 or more feet of ground rods or circle the building with a ring or tap into the buildinv concrete rebar to get anything that low even with just the electrode, not including Earth. That's on par with readings I would expect for #14 THHN, 30 feet long. Earth grounding is somewhat counter intuitive in that it is inversely proportional to distance but typically I don't see readings under 1 ohm (calculated or measured) until you get at least a mile away. That's also why utilities don't bother running grounds except the static line in pole grounding systems...The Earth is lower resistance with that inverse resistance property than a wire which is linearly increasing with distance. 0.75 ohms is an excellent reading for a motor feeder ground with a separate wire, not bonded through conduit.

Did you use a low ohm (Kelvin) bridge? Did you check the resistance of your jumper? There are a lot of tricks to making a reliable measurement under 1 ohm and 0.75 ohms sounds like your jumper resistance. In comparing even high end multimeter like a Fluke 87 to a micro ohm meter, the multimeter accuracy is poor compared to consistent results among micro and millliohm meters when I've checked accuracy.

Either way there's almost no way in common Earth systems to get to 0.75 ohms on a pole ground electrode, let along through 30 feet of dirt to another ground electrode. I'll bet the pole is tied in, it's just buried with the customer ground grid. Not common practice but we do it all the time industrially. Even if it's not, NEC kind of suggests 25 ohms or less for grounding electrodes, and IEEE standard 8p for substations recommends under 1 ohm. If your ground is really that goid it is effectively grounded really well. You should be treating it as solidly grounded in terms of arc flash modelling because that's what it is. Low resistance grounds at 480 V ads around 0.7 ohms and high resistance systems are usually at least 10 ohms, often closer to over 100 ohms. IEEE 1584-2002 would treat high resistance as ungrounded but low resistance and solidly grounded the same.

Are you including the utility fusing on the primary side of the pole mount cluster? Or are you just letting it run to 2 secs? utililizing the utility fuses may make a big difference.

We are letting it run for the full 2 seconds. Another firm completed the arc flash and did not investigate to determine what fusing is being used on the primary side. I now have plans in place with the customer to investigate the primary fusing when they are being accessed for other reasons so that we can get them added to the model.