Hi friends,
When we perform arc flash hazard analysis, we obtain utility information to know the short circuit contributions. But if our grids are all interconnected as per NERC, then do the utility companies give us the largest possible short circuit contribution? This is because, lets say a 800A switchgear for MCC is fed from substation S. And substation S is fed by three different generators of 100MW, 50MW, and 20MW from three different locations. The substation busbar will be designed to withstand the maximum short circuit contributions from these generators, no doubts about that. But which value of short circuit will be given to us when we request for utility information? Will they give us the short circuit value based on 100MW generator? If so, that can have serious impact for the operators at 800A switchgear, if the fault occurs when the substation S is fed by 20MW generator because then the short circuit contribution will be much lower than we used in the calculations (using short circuit value from 100MW) and some of the protection device may not trip or may take longer to trip exposing the operator to higher incident energy. Please kindly share your opinion.

Hello,
When I ask for fault data from the utility for a study, I request the maximum and minimum three phase and single line to ground fault current along with the respective X/R ratios. Sometimes I get that, sometimes I get one value.

I will use the minimum short circuit to perform arc flash hazard analysis and maximum for equipment duty analysis.

You should always use the minimum short circuit for arc flash as that will provide a conservative, worst case value. That should provide the best protection.

Thanks Barry! Like you mentioned, I use both maximum and minimum SC values to compile arc flash analysis and equipment evaluations and flag the worst case scenario whenever I get both max & min SC values. But some utility companies give only one SC value (3-P & L-G) & X/R ratio with three extra pages of general disclaimer. In such case, I create a different scenarios with impedance of the transformer by varying it +/-10% to compare and validate my study. If all utility companies give two possible values of SC for (3-P & L-G) with respective X/R ratio, I think hazard analysis study would be more precise.

Max and min values will become increasingly important as more utilities move towards smart / self healing systems. I just talked with a person from a utility that has moved towards this type of a system and he said there are never quite sure what their short circuit current will be since the system has a "mind of it's own" and can make corrections.

Jim
I agree as the last utility I worked as was implementing a self healing system in one town. The system operators weren't too keen on it as they like to control the system.

For additional information about this subject, here are couple of articles that I have written about utility short circuit data. I just posted them both on the arc flash forum.

• [url='http://arcflashforum.brainfiller.com/threads/changes-to-electric-utility-short-circuit-data.3300/']Changes to Electric Utility Short Circuit Data[/url]
• [url='http://arcflashforum.brainfiller.com/threads/short-circuit-data-per-unit-amps-symmetrical-components-mva.3301/']Short Circuit Data - Per Unit, Amps, Symmetrical Components, MVA[/url]

Im fortunate to have worked for the Utility that supplies the fault current for the system I am studying. They normally have given me the interconnected fault current for the system with known system revisions tat will occur within the next twelve months. With this information I would note that it is probably at least as important, if not more so, to know what the long range plans are for the system you are analyzing, ad what switching option are available in the system. My particular system is fed by two pairs of transformer into seperatly operated buses 99.9% of the tim. Even with the .1% condition the fault current on the 480V systems throughout the plant have not increased appreciably even thought the 115kV source impedance has increase, or will by the end of the year, by 225%. The point I make is that the source can have an impact, but it is usually small - depending on your transformer sizes and impedance's.

That solves the high fault current end, but the low fault current is far more difficult to solve. The number of variables that can affect the lower fault current, and increase the arc flash risk are significant:
How many transformers are in service during a plant turn around.
How many motors are running during those times.
Motors across the Plant and Motors on the bus being analyzed.
Etc.
The low fault current situation is by far the most difficult to get a handle on.

[INDENT]The low fault current situation is by far the most difficult to get a handle on.[/INDENT]
- See more at: http://arcflashforum.brainfiller.com/threads/how-accurate-is-the-utility-information.3274/#post-15284

People make a big deal about the variability of the utility fault currents and their ability to increase, but it seems for most 480V customers, with <3000A services, low source fault currents offer a bigger threat due to long clearing time of protective devices (I find it is rare for arcing faults to reach the current limiting region of large fuses). It is an entirely different discussion for services in the MV range. Above >2400V everything is critical.

I believe we need to do a better job of teaching using examples and assumptions for <3MVA installations and <600V.
I just reviewed AF projects I did in the Chicago area for the past 3 years, only 1 had a 'provided' utility SCA higher than 35kA @480V (other than on the networks in the Loop).

[INDENT=1]People make a big deal about the variability of the utility fault currents and their ability to increase, but it seems for most 480V customers, with <3000A services, low source fault currents offer a bigger threat due to long clearing time of protective devices (I find it is rare for arcing faults to reach the current limiting region of large fuses). It is an entirely different discussion for services in the MV range. Above >2400V everything is critical.[/INDENT]
[INDENT=1]I believe we need to do a better job of teaching using examples and assumptions for <3MVA installations and <600V.[/INDENT]
[INDENT=1]I just reviewed AF projects I did in the Chicago area for the past 3 years, only 1 had a 'provided' utility SCA higher than 35kA @480V (other than on the networks in the Loop).[/INDENT]

Good discussion! If the power distribution system is "electrically remote" from the source, it is true an increase in utility short circuit current may not have as significant of an effect on the distribution system. "Electrically remote" simply means a lot of impedance between the utility and system under study.

As an example, a 1500 kVA transformer with a 5.75% impedance at 480 volts has a maximum (infinite bus) secondary short circuit current of a bit of 31,000 Amps. A weaker utility may reduce this a few thousand amps which may or may not make a difference in how a protective device operates - probably no real difference. As I believe most people are aware, lower short circuit current could cause a protective device to take longer to respond and clear a fault but this isn't always the case. If a protective device operates at the same speed with lower short circuit current as it does at a higher short circuit current, then the higher short circuit current will result in the greater incident energy.

Some utilities (my former employer in a past life) can provide minimum values of short circuit but not all of them can. In that case, you can "fish" for the bottom by reducing the available short circuit current in increments of perhaps 10%. Rerun the study and check the PPE level after each reduction. As the short circuit current is lowered, the incident energy will likely also be lower. However, keep reducing the short circuit current and when the incident energy suddenly becomes large requiring a jump to a higher level of PPE, that means the current became lower than the protective device instantaneous. This is not a perfect method but it gives you an idea of the range of currents where the PPE level you wish to use is valid.

Of course if this is medium voltage and the system is not "electrically remote" from the utility, then the utility short circuit current becomes more important.

First let me say: if you guess at the input data you end up guessing at the output.
But some amount of assumptions and hypothesis testing are valid parts of engineering analysis.

How long does it take to run multiple scenarios versus waiting for utility data?
The problem is most people do not have the tools (e.g. are not trained) for determining the validity of ranges and results.

Good method. But does it seem a bit odd in this day of computing power that such a manual incremental approach would be needed? Where is the program that will plot IE versus current on a fuse TCC and and show clearly what and where the maximum is?

No guessing is involved, since the protective device characteristics are known for each increment. Every incremental input is possible in this world, even those below the "minimum."

As a utility guy, I find the concept of "minimum" to be a somewhat shaky concept. Our mission is continuity of service, and I may find I need to reroute power in a way never planned before in order to accomplish that mission. I like Jim's incremental approach since it can be used to find a maximum IE that might exist far below any minimum current I may have provided based on an assumed abnormal system configuration.

Because the available software packages don't allow you to just input a "search" function and with a fairly large simulation, it could turn out to be very time consuming to actually do it in practice. What I've found is that at some point incident energy is driven more by available fault current than trip times while at higher fault currents it is generally driven more by the trip times, forming a maximum somewhere in between. I typically see this effect with incresing distances on overhead power lines (miles vs. feet). Sometimes the effect is minor and sometimes it is the driving factor.

The same thing can happen if you start looking at a large number of very large motors, any combination of which may be offline at a given time. It quickly "blows up" the scenarios. This is what typically happens in a large mining environment.

The only thing that makes any of this truly tractable is that the fact of the matter is that most of the time, conditions on the "secondary" or "other side" of the transformer are generally fixed and not tremendously influenced by variable impedances. So although we could have 10-60 cal/cm^2 of fault current on the "utility side", the range might be much more narrow on the fixed/customer side of things.

It used to be that constructing a huge number of scenarios was pretty simple, too. SKM and its ilk back in the DOS days frequently was little mor than a text-oriented program where the end user had to spend a huge amount of time constructing the input files and then running the software to generate outputs. It was a simple matter to construct another program to automate the process of running hundreds of conditions through the power analysis software. As the system became graphical in nature and easier to "access", we lost the ability to automate the whole system as easily.

My point exactly. The available software is lacking.

As far as time, My PC has lots of it. Idle most of the time anyway, and needs little sleep.

Thank you everybody, especially Jim for brain storming on this topic. It was very helpful. I feel that down the road, there should be a guideline from NFPA 70E/OSHA to Utility companies clearly stating that they should provide us both maximum and minimum short circuit contributions when they receive utility information request for arc flash study. There are some utilities which provide both (example SCE), but there are many others who insist with just one value and mentions that is all they can give. And we spend a lot of time creating different scenarios to get the closest result.

It doesn't matter what 70E puts in there, it will get ignored since Article 90 descopes utilities, even if some are following it.

OSHA is not going to do it either. This is an engineering/design decision which is way outside of anything they want to get involved in.

Best bet would be to somehow make an argument for those same issues and get it inserted into NESC which definitely carries weight with utilities.