Arc Flash Documentation

How should Arc Flash studies be documented?

After the labels have been applied, how should the all the equipment data and software options that went into defining the labels be documented?
Has anyone come-up with a policy for maintaining all of these records and up-dating them as required?

We have issued an electronic format copy of the model that may be maintained by the customer using the same software we did (actually we matched their coorporate standard). We also issue a full short circuit report and arc flash analysis report that is the standard printout from the software. That is the standard stuff I am aware of.

In addition to this, we issued a label schedule with a serial number for each location that may be referenced to a marked up single line diagram. The acompanying report has disclaimers about the calcs only being valid in normal configuration and at this point in time, given the understood distribution system as per the single line diagram.

Having issued all that as a benchmark, we have been employed to update the study with new construction projects. Theoretically we have given the customer enough information that others could revise the work but so far they have chosen to come back to us.

Gary B,
Thank you for your reply.
Our facility has about 50 buildings, 13.8/7.2kV, and 115kV. We are using SKM's PTW software to do our own calculations and labels.

I like the idea of matching label serial numbers to the single line diagrams.

The software files, you mentioned, define the exact values used for the study, but that my not be useful five years from now when the software may change. Did you also provide printed reports showing all relay settings, cable impedances, etc...?

How did you handle various operating modes? We have many large motors and some unusually large motors. Two facilities draw 180MW each at full load. PG&E, our utility, has provided us with both max and min MVA ratings for their various operating modes. The permutations are mind-boggling.

Are there other AF Forum members who are plant operators doing their own calculations and labels? How have you modified you documentation control to include arc flash?

ZeroSeq -

I did a arc flash analysis for a company in PG&E territory some years back and I must have been the first to request actual short circuit data at a specific location. I had to exchange numerous emails with their engineer as he kept replying to use an infinite bus methodology. After much education on my part, he finally provided a range. I got a range of 7 to 100 MVA. I ran several scenarios and since 7 MVA produced worst case, that was used. It was unfortunate as it led to many areas that were Category #3 and #4, which under a more reasonable scenario were #1 and #2.

I usually provide a hard copy printout of the data used (cables, protective devices and their settings) with the report, along with the AFH by bus location to back up label placement. Also in the report is a recommendation to contact the utility on a yearly basis to verify that the system has not changed. A sample letter is included for that.

Some executive decisions were made to simplify this:

1) it is assumed the distribution system was in "normal" configuration. If alternate feeders were in use etc. hot work was not to be performed. This facility had most motors on at one time, with an off line spare, so it was not difficult to estimate or consider 'normal'.

2) We printed out tables of relay protective settings. This was simplified to only include instantaneous, short term and long term settings etc. and not the multitude of other little logic function inherent in modern relays. THAT EXERCISE ALONE probably paid for our study in operating reliability, because so many anomolies were hidden in the previously too detailed files.

3) cable impedances were not documented but the modeled cables are in the software model and are supposed to match the single line diagrams. Most cable lengths were estimated.

4) At some point we had to make practical assumptions, and noted those, rather than getting stuck in perpetual academic thinking. For example if one calculates 12,000A max available and the instantaneous can be set at 4,000A, arc flash exposure will probably not suffer even if the utility is not up to maximum duty.

ZeroSeq,
I am performing the in-house study for my facility. It sounds like we are in similar situations. I have to use five separate SKM models to keep my facilities straight and stay within the 2000 bus limit. (granted, many of these are node buses between impedences.)

The utility contribution at my location is generally so large that most changes don't affect equipment downstream. We are being fed 161kV from the utility.

I've used the SKM scenario manager to keep up with operating modes, and have been able to isolate most transfer switches so that I have a minimum number of scenarios to go through. After running several case studies, I've been able to figure out what provides worst case fairly quickly.

So far, the project files are our largest documentation source. I have added many "user defined fields" to the components to be able to keep up with my notes on various buses and switches, etc. I keep the projects backed up, and GOD HELP who ever tries to take my SKM away! (I really should get some of the data exported, but have not found a good method that I like that I won't still have to go through and straighten to my liking afterward.)

As far as internal controls go, I am pretty well it. I try to keep everything documented so that the next poor sap won't have to figure out what I did and why, but there is no "official" methodology as of yet. We are still refining, and as SKM revizes, that opens up new methods for us. It's really a bit of a mess. There are others involved in the process, and there is a separate "design" department that does not use SKM as far as I know, so the whole process is a bit convoluted, as well as a few other "administrative" departments thrown in.
Bottom line, I try to keep good notes and files and stay informed of ongoing internal projects that modify my results...