When I was doing those kinds of projects, everything was in SKM. That way I had fault values, panel schedules, load flows, etc. Arc-flash was just emerging when my work focus changed.
I'm not suggesting that it's not necessary. Of course it is often necessary. I'm just considering the practicality of spending time and a client's money to do a short circuit study for a building supplied by a 100 kVA transformer. Let's say the transformer has 8% impedance. That transformer cannot produce 25kA, which is a typical rating for panel boards. Is a study still necessary?
Fault current calculations need not take a tremendous amount of time for simple systems. Who knows, maybe you can save your client some money and use 14kA rated panels. Don't forget to include motor contributions!
Alan
“The engineer's first problem in any design situation is to discover what the problem really is.” Unk.
I understand and agree. I think that an engineers goals should be, in order, safety, reliability, and economy. It is terribly easy to over-engineer something.
But you have to do panel schedules any way, right? So use the same program for panel schedules and fault calcs and once you have everything in for the panel schedules the fault calcs are essentially free.
I'm not trying to be difficult (although often this tends to be where I end up). I have little experience in commercial electrical, and I have a potential job opportunity doing commercial electrical and industrial controls. The controls part are easy, this is my background. And so I'm trying to feel my way through the requirements of commercial power distribution. I have been reading through IEEE STD 241 to get some ideas. The buildings I'd be working on are mostly office space and retail stores. Mostly lighting loads and HVAC. So, what I am asking those of you who have done a lot of commercial, what are the baselines for fault studies, ground studies, protection coordination, and arc flash?
Specifically - Fault studies...
In a typical industrial application, I would map a fault onto every bus in the 480V system. Using ETAP, it takes a few hours at least to properly lay out the breakers, cabling, motors, and transformers. But then I could use the one line to do over current protection coordination, fault currents, and arc flash protection requirements.
In a commercial application, such as a small building fed by a 250KVA kva transformer, one 3 phase 208V distribution panel, and let's say 3 lighting panels. To do a full blown short circuit study would take some time. But to just look at the transformer available short circuit, a quick calculation tells you it really cannot produce very much, even with an infinite bus. If the available fault current does not exceed the ratings of the breakers in the distribution panel board, then why would you spend the time to model a one line of the building?
Other - Does anyone do arc flash, grounding studies, and protection coordination studies for such a building as described above?
Like I said, when I was doing those kinds of projects every one went into SKM for panel schedules, load flow, and fault study. The simple systems like what you describe would take 4 or 5 minutes to model. Not much "wasted" time there. The fault study itself would take a few seconds and the every panel schedule would include the available fault current at that location. Any decent analysis package should let you get maximum fault current at every bus with minimal effort, certainly far less effort than doing so one bus at a time.
A short circuit calculation should be performed on all electrical systems no matter how simple they are. The simpler your system, the simpler the calculations would get. The complex the system, the complex the calculation get since you tend to model more information. I assume you already know how to apply the results of short circuit calculations.
Whether you do it manually or by software, it is required for you to verify equipment ratings and perhaps, overcurrent protective device settings for proper coordination. You may use per-unit R and jX calculations to determine the symmetrical and asymmetrical faults or use the KVA method (as long as you tend to go to the conservative side). You may lump some motors loads (preferrably lower than 50HP); anyway, your LV non-motor loads will not contribute to the short circuit fault. If you neglect cable impedances, your symmetrical fault calculations may become conservative. You see, you have options on how to get it done as long as you have decent information or equipment estimation to get at least a conservative calculated fault. You need this as a basis for specifying equipment.
For arc flash on the other hand, if you intend not to verify it, then you should design your electrical distribution equipment to ensure shutdown during maintenance/inspection or locate them in such a way that they are readily accessible to public while energized. At lower voltages (220V, 208V, 120V etc), it is possible to get issues on them due to the use of MCCBs and possible levels of fault current that may generate enough arc to hurt somebody nearby.
I have encountered several electric distribution companies or cooperatives in the past wherein they also specified the kAIC ratings of the MCCBs. Motor contribution (if there are motors like window type ACUs, small water pumps) in residential areas are normally not significant and generally not considered. Hence the only contributing factor is the utility source which is limited by the impedance of the step down transformer. If you assume infinite impedance on these transformers (which are normally are similar and at standard sizes, they may consider a minimum standard KAIC ratings for your residential equipment and ratings. Some engineers looking into a house electrical design may apply prudence and take the nearest distibution transformer, assume infinite source and take the short circuit level and use it or compare with the power company's specified minimum kAIC.
For a pretty small commercial building, the same may apply. For a large commercial building, the power company might design a dedicated power supply for your building loads and they may require you to provide them more information of your electrical loads (check loading, estimate harmonics, etc.). In parallel, you may have to obtain the MVAsc and X/R from them and calculate the fault current
The whole system is designed not to require fault studies on residential. The standard home panel has a AIC rating of 10k. Utilities design their systems to never provide more than 10kA at a residence. To the point that I've seen diagrams showing the process of running the service lateral away from the service location before doubling back to gain the required conductor length such that the available fault current at the service will be less than 10kA even though it wasn't at the point the lateral connected to the transformer.
Email and/or forums can be tremendously inaccurate means of communication.
I have to admit I have been quite confused at the replies in this thread, especially the 4 or 5 minutes required to do a fault study. I don't know anyone who can do anything of value in 4 to 5 minutes. But I get it now. The assumption here is that everyone already owns a copy of SKM and that the one line for the system has already been done in SKM, and the system impedances and available fault currents are already on the one line. This is an inaccurate assumption, but I do get your points. Thank you for your advice.
David,
I don't think you can make it a blanket statement that all dwelling units would have an AIC of 10,000 amps or less. There are utilities that require the service equipment to have a rating of 22,000 for services greater than 150 amps and less than 401 amps. Maybe that is just CYA on their part, but they will not turn the power on if you install 10,000 amp rated equipment.
![[profile]](/data/assets/smilies/profile.gif "[profile] [profile]")