Due diligence

[jonesy29847]

I am preparing to perform arc flash calculations for several manufacturing plants, typical incoming voltage is 480 3 ph, typically from 2500KVA transformers feed from the utility.

What if anything must I include from the utility side in my calculations?

Thank you in advance.

 

[charlie b]

If they are willing and able to give you a fault current value at the primary of the transformers (or at the secondaries, if the utility owns the transformers), that is all you need. If they can't give you that information, I would treat the primary as an infinite bus.

 

[petersonra]

another scary question.

it seems to me that there ought to be some requirement that people making these kind of calculations have some kind of training in doing so. I am not suggesting PE type things, but maybe requiring some kind of exam based certification, and ongoing CEU.

I see people in some plants doing this based on a spreadsheet they got somewhere that they just don't understand. How do they even know the spreadsheet comes out with the right results, or that as best practices change that it is updated.

 

[jonesy29847]

Its interesting that you bring this up, fortunately the math is pretty simple, just college algebra. I have read a lot, and am amazed by what's going on during a short, and all the little bits and pieces that are part of the process.

I am in the very process of preparing such a spreadsheet. My incident energy calculations are based on NFPA 70E Annex D, IEEE 1584 method. To validate my calculations I have checked them by hand, found as many examples to check against as possible. For SCCR I am using a publication by Bussmann, just google "short circuit calculation bussmann", to have a look. I continue to find subject matter on the internet, ABB has a publication number 1SDC007191G0202, I have not yet read, but intend to.

To really answer your question, I think it would be a great idea if someone offered certification.

Thanks

 

[cornbread]

another scary question.

it seems to me that there ought to be some requirement that people making these kind of calculations have some kind of training in doing so. I am not suggesting PE type things, but maybe requiring some kind of exam based certification, and ongoing CEU.

I see people in some plants doing this based on a spreadsheet they got somewhere that they just don't understand. How do they even know the spreadsheet comes out with the right results, or that as best practices change that it is updated.

Even using programs like Etap we still debate, how to do an Arc Flash study... Do you base it on the normal configuration or in our case we have several back feeds where we parallel transformers. Some say worse case others use the normal configuration. We still debate the 208 three phase panels. Seems each new engineer has a different perspective on ARC flash.. some consistency via certification would be nice..just my humble opinion.

 

[jonesy29847]

cornbread.

My money says worst case.
What effect do parallel transformers have?
Sum up the KVA, What about %Z?
Are parallel transformers comparable to motor contribution?

 

[don_resqcapt19]

... If they can't give you that information, I would treat the primary as an infinite bus.

That is a problem with arc flash and incident energy calculations. Infinite bus is very often not the worst case, it may even result in the lowest incident energy. You need a reasonable idea of the range of available primary fault current to get an arc flash calculation that reflects the real world conditions.

 

[charlie b]

Parallel transformers are like parallel resistors: the resistance is cut in half. That increases the fault current, but it may or may not increase the arc flash energy. Sometimes the worst case is when the system is running on the standby generator. That is because it will have a smaller amount of fault current, which will allow the fault to continue for a longer period before a breaker trips to terminate the event. The total arc flash energy is a function of both the energy released and the time it takes before the event is terminated.

 

[jim dungar]

That is a problem with arc flash and incident energy calculations. Infinite bus is very often not the worst case, it may even result in the lowest incident energy. You need a reasonable idea of the range of available primary fault current to get an arc flash calculation that reflects the real world conditions.

I just did did a study where the normal operating conditions resulted in incident energy, at an MCC, of 6.2 cal/cm^2. However, when the facility was being shut down, such as during a maintenance outage, the incident energy climbed to >30 cal/cm^2 once the fault contribution from motors was no longer available.

 

[jonesy29847]

...Infinite bus is very often not the worst case, it may even result in the lowest incident energy.

We have pad mounted transformers fed from the utility, I am not sure what the secondary utility transformers supply. The nameplate gives rated volts at 7620/13200, but I don't understand where to read secondary, tap in use or, or is there a defacto standard?

 

[jonesy29847]

I just did did a study where the normal operating conditions resulted in incident energy, at an MCC, of 6.2 cal/cm^2. However, when the facility was being shut down, such as during a maintenance outage, the incident energy climbed to >30 cal/cm^2 once the fault contribution from motors was no longer available.

You bring an interesting perspective, I have 3 compressors with 300hp motors, and it sounds like you are saying that if a fault occurred while none of the compressors were running, it would be a whole new ball game.

Under the conditions you describe, can a fuse type change help and or hurt us in terms of incident energy?

 

[GoldDigger]

I just did did a study where the normal operating conditions resulted in incident energy, at an MCC, of 6.2 cal/cm^2. However, when the facility was being shut down, such as during a maintenance outage, the incident energy climbed to >30 cal/cm^2 once the fault contribution from motors was no longer available.

So in your case the motor contribution somehow subtracted from fault current? It is hard for me to see a way that motor contribution could increase the current through the OCPD itself and shorten the trip time unless you are looking at an arc fault location where the motors are upstream from the protective OCPD.

 

[petersonra]

Its interesting that you bring this up, fortunately the math is pretty simple, just college algebra. I have read a lot, and am amazed by what's going on during a short, and all the little bits and pieces that are part of the process.

I am in the very process of preparing such a spreadsheet. My incident energy calculations are based on NFPA 70E Annex D, IEEE 1584 method. To validate my calculations I have checked them by hand, found as many examples to check against as possible. For SCCR I am using a publication by Bussmann, just google "short circuit calculation bussmann", to have a look. I continue to find subject matter on the internet, ABB has a publication number 1SDC007191G0202, I have not yet read, but intend to.

To really answer your question, I think it would be a great idea if someone offered certification.

Thanks

there is a significant difference between SCCR and IE calculations.

part of the issue is that assuming the maximum scc is available often leads to the OCPD tripping faster. if the real SCC is lower, it will take longer for the OCPD to trip. the math may well end up with the worst IE being some weird combination of events that requires considerable iteration to come up with that is not real easy to do in a spreadsheet or by hand.

 

[jim dungar]

So in your case the motor contribution somehow subtracted from fault current? It is hard for me to see a way that motor contribution could increase the current through the OCPD itself and shorten the trip time unless you are looking at an arc fault location where the motors are upstream from the protective OCPD.

No.
The motors provided additional fault current. The higher level of fault current allowed the protective device to enter its current limiting range. The lower fault current had the protective device operating in its time delay region.

 

[jim dungar]

You bring an interesting perspective, I have 3 compressors with 300hp motors, and it sounds like you are saying that if a fault occurred while none of the compressors were running, it would be a whole new ball game.

Under the conditions you describe, can a fuse type change help and or hurt us in terms of incident energy?

Yes. One of the mitigation solutions we are looking at is to get rid of the 'colored' fuses and go back to the 'old' fast acting ones.

 

[GoldDigger]

No.
The motors provided additional fault current. The higher level of fault current allowed the protective device to enter its current limiting range. The lower fault current had the protective device operating in its time delay region.

So, as I said, the motors were upstream of the OCPD in question?
If not, how could the motor contribution increase the breaker current and not just the arc current?

 

[jim dungar]

So, as I said, the motors were upstream of the OCPD in question?

Some motors were upstream while others were downstream. The biggest effect came from the upstream MV motors.

My point is that determining the worse case arc fault incident (AF) energy, needs to consider all of the typical conditions that are considered normal.
During maintenance it is quite common to turn off motor loads prior to opening protective devices.
In the case of my customer, it turns out their highest AF energy occurs while they are doing the riskiest work of racking out breakers when their process is not running.

 

[mayanees]

Its interesting that you bring this up, fortunately the math is pretty simple, just college algebra. I have read a lot, and am amazed by what's going on during a short, and all the little bits and pieces that are part of the process.

I am in the very process of preparing such a spreadsheet. My incident energy calculations are based on NFPA 70E Annex D, IEEE 1584 method. To validate my calculations I have checked them by hand, found as many examples to check against as possible. For SCCR I am using a publication by Bussmann, just google "short circuit calculation bussmann", to have a look. I continue to find subject matter on the internet, ABB has a publication number 1SDC007191G0202, I have not yet read, but intend to.

To really answer your question, I think it would be a great idea if someone offered certification.

Thanks

The only real way to be legitimate is to use a recognized software package like SKM, ETAP, EasyPower, ..., and develop multiple operating scenarios from which to determine the worst-case incident energy levels. Then learn how to interpret and defend the results.
You should get minimum and maximum Utility contribution info, and you'll need their overcurrent protective device information if you're labeling the service-entrance bus.

 

[jonesy29847]

Can you elaborate on multiple operating scenarios?

I will get more information, in all fairness when we buy equipment the build source calculates incident energy, however some take the lazy way out (minimum SCCR of lowest rated component) and when this method is used PPE levels typically increase.

Typically most of our systems will be PPE level 1, we are removing 480V from the 'working' side of the enclosure, and control voltage is 24VDC. The objective is to reduce the need to be on the 480V side.

PPE is the least preferred method to safety.

 

[mayanees]

Can you elaborate on multiple operating scenarios?

Evaluating multiple scenarios involves performing the incident energy calculations for different operating conditions. The base scenario is typically when the system is tied to the normal utility source. Alternate scenarios would be when powered from different sources, like backup generation. Or perhaps when powered from a transfer switch fed from another area of electrical service.
IEEE 1584.1-2013, IEEE Guide for Performing Arc-Flash Hazard Calculations, section 6.2 advises that "Number of calculations (scenarios) needed" should look at two Utility contributions: one with large motors running, and the other without.
The label should then display the worst-case incident energy level and the scenario in which it was developed.