Seems that I am running into another utility not wanting to provide the available fault current only infinite bus on the secondary side of a 2500kVA transformer. Where this will get interesting is that the AHJ will not provide an energizing permit without the service entrance being labeled in accordance with NEC 70 2017 Article 110.16. One item that this article requires is the available fault current at the service overcurrent protective device.

Just sent my first reply back to the utility explaining in detail why the available fault current is needed and not infinite bus on the secondary side of the transformer, so waiting a reply.

The client, a major international company, built a new facility with one riser cable to utility but there are three 2500kVA transformers off this riser. They want to energize on Feb. 19 so I don't think they are going to be too happy with the utility, if the utility in spite of my efforts, still refuses to provide the available fault current and prevents energization. I imagine there will be some big legal guns calling the utility.

Update: The utility has said that Wisconsin and Minnesota Public Service Commissions have upheld Xcel's policy on arc flash requests with regard to providing only infinite bus fault current based on transformer size. We all know that infinite bus fault currents will result in erroneous incident energy levels.

Has anyone else on this forum had dealings with Xcel regarding arc flash requests and available fault current?

I have been dealing with Xcel for over 10 years. I remember when their policy was more severe than it is now.
Their policy is clearly spelled out in their 'blue book' Standard for Electrical Installation and Use. They provide tables of infinite bus Short Circuit values based on the supplied transformers. Xcel has a note suggesting the study use multiple scenarios with larger and smaller transformers, to address possible transformer changes. Over the years they continue to clarify that they will not provide any other fault current value to a 'secondary metered customer', Primary metered customers can obtain realistic values, which of course are subject to change.

Their position is that everything be de-energized before it is worked on.

Well, that policy of de-energizing certainly makes it difficult to troubleshoot, perform IR, rack out breakers, etc.

I even sent to the Xcel engineer one of the documents I had used with another PUC that showed a system that was modeled using the actual available fault current and then used the infinite bus fault current and was able to show in that one model examples where available fault current had a higher IE than the infinite bus and where available fault current had lower IE value. Fell on deaf ears.

I am waiting to hear from my client if they want to make a formal complaint with the PSC there. The AHJ won't issue the energizing permit until the service entrances are labeled per NEC 70 with the available fault current and clearing time. This has the potential of holding up energization of three 2500kVA transformers.

I am not sure the NEC knows what it wants.
110.24 specifically requires the use of the Maximum available fault current. This could easily be interpreted as a design or infinite source value.
110.16(B) Asks for the actual fault value and device clearing time at that value. However, there is an exception saying that a fault current value is not required when an 'arc flash label, such as per NFPA70E, has been installed.