The available short circuit current could affect the incident energy – especially if a reduced current results in a longer arc duration from a longer upstream device clearing time. This is about the SC current from the utility, not generators.

Here is the question:

Which utility short circuit current do you use for an arc flash study?
Select all that apply.

Normal – from utility
Minimum – from utility
Minimum – fixed percentage of normal
Minimum – trial and error for worst case
Worst case maximum – Infinite bus
N/A
Something else.

I usually run multiple scenarios.
Fault MVA supplied by utility
Fault MVA plus 20%
Fault MVA minus 20%

Infinite Bus on a utility transformer is actually impossible, there is always some limitation to supply. Plus Infinite Bus may give lower IE values as the higher fault current moves breakers into their instantaneous trip levels. You can run it as a scenario but need to interpret the results.

Problem is if the utility won't give you any value. Then I assume Infinite Bus as maximum and 40% less as minimum.

To a great extent, calculating IE is a best guess assumption anyways. My idea is to pick a scenario that is close but that errs to a conservative value.

I'll use the infinite bus for determining the short-circuit rating on equipment. For arc flash calculations, absence a defined value from the utility, I'll use a maximum of 10 kA on the primary.

I'm in the midwest. I've talked to several of the local larger utilities at IEEE events and in general, they say they try and keep the available short-circuit on the primary distribution feeders to 10 kA. That corresponds to roughly a 15 - 20 MVA substation transformer, which is fairly large for distribution systems. Also, the ampacity to carry that load is getting to be a really large single conductor 15 kV overhead conductor which the poles have to go to beefier construction.

I'm sure different utilities in different parts of the country have different philosophies on what they design their distribution systems to, especially large metropolitan areas versus rural. So, what is reasonable here, may not be valid in other parts of the country.

I normally start with infinite bus to get worst case maximum and consider alternatives from there, especially if the utility is not forthcoming with the needed data. On client sites, I normally work on the LV side of the utility transformer for battery energy storage systems. As a result, I must factor in any SCC contribution by the bi-directional DC/AC inverter as well.

~doug

This may vary depending on the data the utility provides. Typically, I will follow the idea of PCIC-2009-16 "Impact of Available Fault Current Variations on Arc-Flash Calculations" as well as your video detailing the approach. I start with an infinite bus source and determine the maximum fault current. I then use either this value or the maximum provided by the utility as my starting point. I analyze all buses and export the data to Excel. I repeat the process, reducing the fault current by 10%. I repeat this process until I reach the minimum fault current supplied by the utility. I compare these results to identify the point at which the utilities fault current causes breakers to trip outside of their instantaneous trip region. Given that this fault current is within the parameters identified by the utility I use this as my "worst case".

Additionally, I will check scenarios if there are multiple utility feed options, generator feeds, tie-switches, etc.

You read to consider the maximum fault current from the utility. You have no control of the actual fault impedance, and we have all seen faults that have sufficient impedance to not clear, so you need to consider lesser fault currents as well the maximum fault current. Quite often the longer duration fault, at a lower fault current, can generate higher incident energies than that quickly cleared maximum fault current level.

A good friend of mine use to always say "with an Easy-Bake Oven, even a 60 watt light bulb will cook a brownie."

Fortunately, the utility that I usually deal with provides both the maximum and minimum fault levels upon request.

I usually run multiples studies with a per unit voltage at 0.9, 1.0 and 1.1 for the both the max and min fault levels (and sometimes the average if required).

Once the minimum and maximum arcing fault currents are plotted against the protection curves, I can inspect visually to see if it falls uncomfortably close to a short-time/instantons/magnetic pickup. If it does a can adjust it a little more so I have a set of results that include an assessment below aforementioned pickup.

As with most engineering there is no one set process that will work well every time, sometimes you need to use your judgement to workout the best approach for each case.

You would have to have a good understanding of the network and how it is operated by the client and what supply interlocks are in place before you can answer this properly

1) How many utility connections are there
2) Can the utility connections ever be parallelled or are hardware interlocks in place to prevent this
3) Are there any other sources such as embedded generation that can energis the system (and can they operate in co-generation mode)

Once i know the answer to this i woiuld determine what scenarios i would need to run (different switching / supply arrangements) and different characteristics. I then complete and validate the model in SKM with a range of scenarios and use the data visualiser to compare the bus fault levels and identify the min and max and then configure and run the ArcFlash module to report the worst case.

When i quote for any system study i also quote for a maximum of 4 or underexceptional circumstances 6 scenarios to analyse and its my responsibilirty to work with the client to establish which operational scenarios give the most onerous conditions.

This has been my method but id be interested to hear if you feel there are flaws with this or better ways of approaching the issue

Funny, you risk confusing some of your audience on this one.

~Doug

Typically you won't get utility fault current if the transformers are utility owned. Curious how many of you try to find out the fuse in the transformer or on the pole and if you trust what the utility tells you, or do you only use it if it can be visually verified?

My personal experience has been that I get the information I ask for which is
>3ph and SLG available fault current and X/R at the primary fuse
>Primary Fuse data
>Riser cable data
>Transformer data

Sometimes you have to be persistent as the first answer will usually just be the infinite bus secondary side fault current a customer service rep is reading from a chart.

I have only had one utility that gave a hard time to providing the information, starting off by saying it is proprietary information. After numerous go arounds and excuses changing, I filed a complaint with the state regulatory board. Then I had to educate them on the need for the information. Eventually the asst. directory of the department told the utility they have to provide the information and develop a policy on providing this information going forward.

Hello "jturner".
Internally we are discussing 0.9pu, 1pu and 1.1pu voltage calculations. We can't find anything in any standards regarding this, but since equipment we are dealing with can operate in all those conditions, should we do it or not? The biggest issue we see with 0.9pu calculations where the incident energy increases a lot. While there is a chance that the equipment might be at 0.9pu especially with min SC, most of the time that would not be the case. It would be easy just select that worst case scenario 0.9pu with min SC, but the end user would be forced to utilize always heavy PPEs which is not always the best approach. In our scenarios we go from <1.2Cal/cm2 (1pu and 1.1pu) to over 20Cal/cm2 when 0.9pu is used. This is for LV equipment where we require energized access for troubleshooting - equipment is well protected to keep the incident energy below 1.2Cal/cm2, but now when we look at 0.9pu, that protection is not sufficient for that special scenario to keep incident energy below 1.2Cal/cm2 as desired in our original design. Note that we have thousands of this equipment with downstream cabinets installed globally and retrofitting it will be a challenge, but we definitely are not excluding that option.

How did you get to that decision to evaluate 0.9-1.1pu voltages? driven by some standards/regulations or was it the internal decision based on some other reasons?
Thank you for your help.

Hi ap

My comments were regarding a typical installation fed from the utility network. This network will determine the voltage tolerances and prospective fault contribution.

The network voltage will fluctuate, by how much it varies will depending on the network topology. Utilities will have there own requirement at which the network voltage must be maintained.

Provided in the fault study from my local utility are the conditions at which maximum and minimum fault level are calculated. In my case, for their LV distribution network it is 0.9-1.1 PU. The network will not typically operate at this voltage, but its plausible that it could. Other networks will likely have different parameters i.e. 0.96-1.08 PU. I would suggesting contacting the utility to identify what is applicable.

It sound like you may be trying to provide a blanket arc flash rating to some equipment that is installed in different networks across the world. It is hard for me to comment to your circumstance without knowing more than this.
If your equipment is something you manufacture and is installed different distribution network around the globe, then fault levels and network voltages can vary widely and prospective fault currents can be very low in situation, below the short circuit protection pickup of many protection devices.
Blanked assessments can be given, but usually they at least require fault study to ensure that the short circuit protection will operate for a arcing fault.

Every installation is different. Different prospective fault levels, difference impedances for consumer mains and different network conditions.

I are the utility, so I said something else. I generally supply what is asked for, with the disclaimer that the system may be reconfigured and equipment changed without notice.

Regarding the +/- 10% voltage, I would point out that ANSI C84.1 range A and B service voltages are tighter than that.

Hi ap,

Regarding the +-10%, This is derived from the fault study provided by my local utility. The condition at which the minimum fault condition occur, show the network voltage at 0.9x nominal for the distribution network, likewise maximum 1.1x nominal. This is going to be different depending the utility and where it resides. As several pointed out, the ANSI tolerances are tighter than this (c=0.94-1.06 I think). There are also client specific arc flash standards I work to that specify some arc flash study parameters, assessment at the edge of the network voltage tolerances being on of them.

If your intention is to provide a blanket arc flash assessment for you equipment's, this can tricky to do, typically a generic assessment will require at minimum a fault study for each installation to confirm that minimum fault levels are above a certain threshold to ensure that the arcing current will be high enough for the short circuit protection element to clear the fault.

As you mentioned, your equipment is installed globally so you are likely going to encounter large variety in network tolerances.

I don't know the nature of the equipment you mentioned so it is hard for me to comment. Assessing at the edge of the voltage tolerance may or may not be relevant to you.

Thank you.

Thank you.

Depends on what we can get from the utility. If we receive a normal calculated primary fault current, we use that. If we receive a calculated min and max, then we would run two scenarios and label to worst case. For the difficult utilities that will only provide secondary infinite bus, we will reduce infinite bus fault current by 30-50% and run multiple scenarios and label to worst case.

Some suggestions I use when the utility fault data is not available. I will use a 10 MVA customer transformer (pipeline booster site) in my example calcs below.

1.) Absolute minimum utility source (no large motor inrush): Assume the utility can supply the 10 MVA transformer and maintain a 3% loadflow voltage drop to the site. For a 3 % voltage drop , Utility MVA would need to be approximately 333 MVA. 33 times the transformer nameplate, calculated from "(10/0.3)".

2.) Practical minimum utility source: Assume same 3% utility drop (flicker allowance) for the largest inrush (motor start). In the case of a typical pipeline booster site, the largest motor is 35% of the transformer. 600% inrush on the motor would be 21 MVA, however, the bus voltage will typically sag during the start to 80%. The actual inrush (@80%) is 16.8 MVA (constant locked-rotor impedance). Add 5MVA (other motors running) and you are around 22 MVA.
Performing the same math as item 1 above: 22 MVA x 33.3 = 732 MVA.

3.) Practical maximum utility source: 100 x transformer nameplate. The difference between 100x and infinite bus is only a few percentage point on the secondary current.

*****************

Final thought. I use the three scenario method mentioned above when an expected value is provided from the utility. Nominal and +/- 20%. It is important to run all three scenarios since there is not one single value that will give the worst case values across the study.

a) points being cleared by current limiting fuses or instantaneous device will give their worst case with the highest utility MVA.

b) points being cleared by devices along an inverse time curve will give their worst case with the lowest utility MVA. (ie. current
may go down by 80%, but time increases 140%).

c) and before you toss out the nominal case considering the above, some points do give worst case at nominal MVA. The reader
can ponder this (non-linear TC curves).

I hope I have given a few ideas. Cheers!!

Wm Pommeranz, PE

Hi all,

I would like to have your view regarding external grid short circuit scenarios in Arc Flash calculations. I have some doubts about how this issue should be addressed.

What should I do if there is no info about the available short circuit power in the grid?

As far as I understand, considering only hypothetical maximum and minimum short circuit scenarios for calculation I may not be achieving the “Worst Case� incident energy scenario. The reason is that there may be some intermediate Scc scenarios in which the combination of FCT + Icc may lead to a higher Incident Energy. So the correct approach may be doing a sweep with different values. There is also some scientific literature with conclusions pointing in the same direction.

Similar happens when the utility gives you the data: what about short circuit intermediate scenarios?

How do you see it?

Thanks for your support.

Regards,