Lowering the Available fault Current

Status
Not open for further replies.

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
I just thought I'd check in and see if anyone has more ideas on possible solutions to a particularly nasty problem.

A friend bought 4 big MCCs for a rock plant and had determined that the AFC was 35kAIC. So when he ordered the MCCs, he ordered them as 42kAIC. Then before he even got them installed, the owner kept adding more and more equipment, he had to keep increasing the service size and transformers, but forgot about the effects that would have on the already ordered MCCs. Now everything is installed and a service company is doing the Arc Flash study and guess what? They now have 58kA AFC. The individual MCC buckets and units are actually OK because they are 65kAIC listed, but the bus bracing is all only 42kA and I've been checking with the factory, it means ripping everything out and changing the bus bracing to increase it (let alone a UL site inspection to put new labels on it). I think changing the transformer(s) in the double ended sub station he put in to ones with higher impedance is out of the realm of possibility. So is changing the MCC structures, we are taling 70+ sections.

My suggestions to him so far:
  1. Hire a PE to do a study and see if he can either replace the Main CBs in the MCCs with Current Limiting versions, or add CL fuses up stream that would bring the AFC at the MCC line terminals down below 42kA. Problem is, 2 of the MCCs have 2000A mains, that ain't no simple thing.
  2. Start digging trenches and make the cables from the Switchgear to the MCCs artificially longer (might be the least costly solution).
  3. Add a Reactor ahead of each MCC. I've never done this, just heard about it from others.

Any other ideas?
 
I just thought I'd check in and see if anyone has more ideas on possible solutions to a particularly nasty problem.

A friend bought 4 big MCCs for a rock plant and had determined that the AFC was 35kAIC. So when he ordered the MCCs, he ordered them as 42kAIC. Then before he even got them installed, the owner kept adding more and more equipment, he had to keep increasing the service size and transformers, but forgot about the effects that would have on the already ordered MCCs. Now everything is installed and a service company is doing the Arc Flash study and guess what? They now have 58kA AFC. The individual MCC buckets and units are actually OK because they are 65kAIC listed, but the bus bracing is all only 42kA and I've been checking with the factory, it means ripping everything out and changing the bus bracing to increase it (let alone a UL site inspection to put new labels on it). I think changing the transformer(s) in the double ended sub station he put in to ones with higher impedance is out of the realm of possibility. So is changing the MCC structures, we are taling 70+ sections.

My suggestions to him so far:
  1. Hire a PE to do a study and see if he can either replace the Main CBs in the MCCs with Current Limiting versions, or add CL fuses up stream that would bring the AFC at the MCC line terminals down below 42kA. Problem is, 2 of the MCCs have 2000A mains, that ain't no simple thing.
  2. Start digging trenches and make the cables from the Switchgear to the MCCs artificially longer (might be the least costly solution).
  3. Add a Reactor ahead of each MCC. I've never done this, just heard about it from others.

Any other ideas?

I would not toss out the idea of changing transformers. If these are brand new units and standard sizes, it maybe worth either selling them or reusing them on another project internally and procuring high impedance transformers. It maybe a better economic/space sollution than adding the current limiting reactors.
Making the cables longer and increasing the impedance that way is probably the costliest solution.
Instead of adding CL Reactors on the secondary side of the transformer - 4 or even more as additional MCC's are added, two can be added to the primary side of the transformers. (They would be less costly than on the LV side, surprisingly.)
 

jim dungar

Moderator
Staff member
Location
Wisconsin
Occupation
PE (Retired) - Power Systems
Do you trust the values from the study? Did they use actual values from the utility, or a possible future worst case? How did they treat motor contribution?

What size and impedance are the transformers?

If you are worried about putting UL labels onto modified MCC's, I am surprised you would use unlisted 'series' SCCR rating.
 

templdl

Senior Member
Location
Wisconsin
I would not toss out the idea of changing transformers. If these are brand new units and standard sizes, it maybe worth either selling them or reusing them on another project internally and procuring high impedance transformers. It maybe a better economic/space sollution than adding the current limiting reactors.
Making the cables longer and increasing the impedance that way is probably the costliest solution.
Instead of adding CL Reactors on the secondary side of the transformer - 4 or even more as additional MCC's are added, two can be added to the primary side of the transformers. (They would be less costly than on the LV side, surprisingly.)

My most recent assignment was with a dry type transformer manufacturer where I supplied numerous air core CL reactors to an MCC manufacturer but supplied more MV AC CL reactors for reducing the fault current to MV dist systems. The MV ones were in (3) 10'x10'x10' enclosures that had to be shipped via special carrier permit to a University in Massachusetts and I also have one up in Boise Idaho.
Fun stuff to sell. Quite often air core reactors are overlooked as a solution to high fault current availablitiy.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
Do you trust the values from the study? Did they use actual values from the utility, or a possible future worst case? How did they treat motor contribution?

What size and impedance are the transformers?

If you are worried about putting UL labels onto modified MCC's, I am surprised you would use unlisted 'series' SCCR rating.
Unlisted, yes, but not without a PE stamp (not from me, I don't have one).

Eaton field services has done the Arc Flash study and they did visit the site, but I don't know how they determined everything, I have not seen the study yet. I know people who work for them and they are pretty good, but they did not do this particular project so I can't comment on the accuracy or thoroughness of the ones who did. But that's a good point, I'll bring it up. Sometimes it can be "challenging" to get data from PG&E out here so they may have made assumptions.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
I would not toss out the idea of changing transformers. If these are brand new units and standard sizes, it maybe worth either selling them or reusing them on another project internally and procuring high impedance transformers. It maybe a better economic/space sollution than adding the current limiting reactors.
Making the cables longer and increasing the impedance that way is probably the costliest solution.
Instead of adding CL Reactors on the secondary side of the transformer - 4 or even more as additional MCC's are added, two can be added to the primary side of the transformers. (They would be less costly than on the LV side, surprisingly.)

Ugh, he had to wait over 8 months to get those transformers...

On the bright side though, if he CAN find some that would work, he could likely sell the ones he has now to someone on the east cost and make money. I've heard stories of contractors in NY and NJ contacting people out here and offering cash to get their place in a production allocation from some of the mfrs.
 

Phil Corso

Senior Member
Jraef?following are my thoughts:

The source impedance resulting in a SC-Duty of 65kAIC equals 4.26x10?? W [480/(3x65,000)]. To reduce it to 42kAIC requires a total impedance of 6.60x10?? W. Thus, an additional impedance of 2.34x60x10?? W is required.

Several solutions have been mentioned, i.e., primary CL-fuses, secondary CL-fuses, primary reactor; and a seconary reactor!

While reactors are the more clinical approach, I suggest you investigate installing cable, to replace or augment the secondary connection between the xfmr and CB terminals.

Note: voltage-regulation may be an issue if there are very large DOL-started motors!

Now the caveat? obviously the xfmr is quite large*, so without specifics I can?t give a meaningful example! Personally, I lean toward the CL?s! Or the explosive-type fuses used in Europe!

Regards, Phil;

* An aside: I still regret having once used a 3,000 kVA, 400V, xfmr!
 
Last edited:

tesi1

Member
Location
florida
fault current

fault current

in your fault current calc's has the motor contribution been to the mcc's this may make it even higher
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
in your fault current calc's has the motor contribution been to the mcc's this may make it even higher
That's another thing I don't know yet because I haven't seen the report. But one of my initial tasks will be to see how that was done. Almost all of the biggest motors, even some of the crushers, are on VFDs. I have yet to hear definitively if they count towards motor contribution in fault calcs or not (some say yes, some say no). I say no because they are isolated by the converter. This was (obviuosly) not engineered so there is no one-line for them to work from yet, only a motor list. If all they did was to look at the motor list and assume they can all contribute, that may be an issue. And even if you are of the mind of VFDs being able to contribute, these all have line and load reactors, so were they factored into that contribution as well? I just don't know yet.
 

jim dungar

Moderator
Staff member
Location
Wisconsin
Occupation
PE (Retired) - Power Systems
...are on VFDs. I have yet to hear definitively if they count towards motor contribution in fault calcs or not (some say yes, some say no). I say no because they are isolated by the converter.

Motors fed through 'normal' inverters do not contribute short circuit current back to the grid. Motor circuits that bypass the VFD may contribute. Many people doing studies do not take the time to determine the operation of the VFD, and make the mistaken assumption that bypasses are regualarly in place.

VFDs that 'regenerate back onto the grid', like those with active front ends, may be sources of fault current but it is rarely at the level of a bypassed motor as the current flows trough some type of 'semiconductor'.
 

charlie b

Moderator
Staff member
Location
Lockport, IL
Occupation
Retired Electrical Engineer
Sometimes all that is needed is a sharper pencil. Calculations of this nature are often done without complete knowledge of the as-built conditions. For example, if I did not know how long a particular conductor really was, I would look at the floor plans, and make an estimate. If I were doing a voltage drop calculation, I might make a conservatively high estimate of the length. If I were doing a fault current calculation, I might make a conservatively low estimate of the length. So I suggest getting the installer (or someone from the plant) to figure out the actual, as-installed cable lengths. Then give those numbers to the people who did the calculation, and ask them to run the calculation again. This, of course, might give you an answer that was worse than the first, but that would not change the situation for the worse. As there is a possibility that a more precisely performed calculation might give you a fault current value below the 42KA, it is worth a try.

By the way, I agree with those who say that a motor driven by a VFD does not add significantly to the total fault current. But if the VFD has a bypass mode, then I would include motor contribution in the analysis.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
Motors fed through 'normal' inverters do not contribute short circuit current back to the grid. Motor circuits that bypass the VFD may contribute. Many people doing studies do not take the time to determine the operation of the VFD, and make the mistaken assumption that bypasses are regualarly in place.

VFDs that 'regenerate back onto the grid', like those with active front ends, may be sources of fault current but it is rarely at the level of a bypassed motor as the current flows trough some type of 'semiconductor'.
That's my take on it too and these are neither bypassed or regenerative. I've seen people using SKM to do the analysis and although it does have a sort of wizard for VFDs, twice I've seen people just plug the boxes that say bypass without checking first. The comment when I asked was "better safe than sorry". But in this case if they did that, I think it could make a significant difference. From my recollection, there is probably 1200-1400HP worth of the 3000+HP connected load that is on VFDs of various sizes.
 

jim dungar

Moderator
Staff member
Location
Wisconsin
Occupation
PE (Retired) - Power Systems
But if the VFD has a bypass mode, then I would include motor contribution in the analysis.
It seems consulting engineers like to put bypass on every VFD, even when it is impossible for the process to operate, properly, in that manner. About 20yrs ago a local hospital went to the bypass mode (full speed) and blew open several access doors on their ducts. In a different facility during full speed operation caused problems with the air flow in the OR. Neither facility uses bypass mode (they disable it), but somehow it remains in every HVAC spec package.

But, when in doubt I 'bypass' VFDs when doing Short Circuit studies, and run two arc flash scenarios one with and one without.
 

JoeStillman

Senior Member
Location
West Chester, PA
I agree with Charlieb. A sharper pencil could help.

I like the reactor idea better than current limiting fuses or breakers. The reactors will cost less than a transformer replacement.

In my world, improvements to drive reliability are making bypasses obsolete. We've come a long way since the first IGBT/PWM drives showed up. I've seen guys spec drives with bypass just because its always done that way, even for redundant fans/pumps that never both run at the same time.
 

kingpb

Senior Member
Location
SE USA as far as you can go
Occupation
Engineer, Registered
Reactors, no doubt. Quite common in this application, especially when MCC's at one time had max rating of 42kAIC.

You said you conferred with factory. Double check. I ran into this same situation in the past, and found that even though the label on the MCC said 42KAIC, the manufacturer only built 65KAIC and that was the case even if you special ordered a 42KAIC line-up. We paid them to come out and slap on new nameplates giving them the full rating of 65KAIC.
 

templdl

Senior Member
Location
Wisconsin
Reactors, no doubt. Quite common in this application, especially when MCC's at one time had max rating of 42kAIC.

You said you conferred with factory. Double check. I ran into this same situation in the past, and found that even though the label on the MCC said 42KAIC, the manufacturer only built 65KAIC and that was the case even if you special ordered a 42KAIC line-up. We paid them to come out and slap on new nameplates giving them the full rating of 65KAIC.

Back to the point that I made in my post #6 regarding a reactor.
 
Status
Not open for further replies.
Top