Arc flash study questions/comments

[EC Dan]

I have some questions/comments about arc flash studies based on my recent experience with one.

• What are transformers considered as far as equipment type based on IEEE 1584 2018? Does it even matter what the type is as long as internal dimensions, conductor gap, and working distance are accurate? I can’t purchase the 2018 standard at the moment so I’m relying on online calculators. My thought is that it would be type MCC/Panel – Deep.
• For panelboards with MCBs, which OCPD do you use for clearing time? If the fault is upstream of the MCB, which I would think would occur most of the time, then the upstream OCPD should be used, however faults can occur on the bus in which case the MCB would be more relevant. My thought is that faults occurring upstream of the MCB will always result in longer or equivalent clearing times, and hence higher incident energy, so the upstream OCPD should always be used.
• If an arc fault occurs before the first feeder OCPD from the service transformer, what stops the arc? Surely it will not cause any utility-side fuse to blow, so does it just persist until enough of the infrastructure has blown apart and burned away?
• In almost every case here working with 480 VAC supply voltage and about 10 kA available fault current from the utility, the incident energy has been calculated at less than 1.2 cal/cm^2. The two exceptions are counter-intuitive: 208 VAC distribution panels downstream from 480/208 transformers. These resulted in 7-8 cal/cm^2 due to the maximum OCPD clearing time of 2 seconds. I understand the reasons why this is, but it’s amazing that these panels wouldn’t have even been considered in IEEE 1584 2002. It also tells me that as a design consideration, OCPD should be provided closer to the transformer secondary and outside of the distribution panel to prevent these types of panels from having ridiculous ratings.
• How does the POCO determine the available fault current at the service transformer secondary? Do they actually calculate line impedance all the way back to the power plant alternators? There has been significant development around here and I know for a fact utility lines have been upgraded, however the POCO did not change their estimate from three years ago to this year.

 

[wbdvt]

What are transformers considered as far as equipment type based on IEEE 1584 2018? Does it even matter what the type is as long as internal dimensions, conductor gap, and working distance are accurate? I can’t purchase the 2018 standard at the moment so I’m relying on online calculators. My thought is that it would be type MCC/Panel – Deep.

This questions seems to indicate that you are performing an incident energy analysis on a transformer. In my experience, this is not done as this is equipment that is not labeled as it is not required to be serviced, inspected, maintained, etc while energized so no analysis is done. That being said, you will have multiple IE values and shock hazard values on a transformer. So how to label? Worst case or multiple labels. I do not see much value in labeling a transformer.

For panelboards with MCBs, which OCPD do you use for clearing time? If the fault is upstream of the MCB, which I would think would occur most of the time, then the upstream OCPD should be used, however faults can occur on the bus in which case the MCB would be more relevant. My thought is that faults occurring upstream of the MCB will always result in longer or equivalent clearing times, and hence higher incident energy, so the upstream OCPD should always be used.

For equipment that has a main circuit breaker, the upstream OCPD is always used unless the equipment is certified as arc rated equipment. Since the MCB is located in the same space as the bus bars and breakers, it is assumed that any arc/plasma ball will engulf the MCB rendering it useless and continuing the fault with the line side terminals.

If an arc fault occurs before the first feeder OCPD from the service transformer, what stops the arc? Surely it will not cause any utility-side fuse to blow, so does it just persist until enough of the infrastructure has blown apart and burned away?

Most likely it will continue to burn and arc until it self extinguishes.

In almost every case here working with 480 VAC supply voltage and about 10 kA available fault current from the utility, the incident energy has been calculated at less than 1.2 cal/cm^2. The two exceptions are counter-intuitive: 208 VAC distribution panels downstream from 480/208 transformers. These resulted in 7-8 cal/cm^2 due to the maximum OCPD clearing time of 2 seconds. I understand the reasons why this is, but it’s amazing that these panels wouldn’t have even been considered in IEEE 1584 2002. It also tells me that as a design consideration, OCPD should be provided closer to the transformer secondary and outside of the distribution panel to prevent these types of panels from having ridiculous ratings.

It came down to how much money was available for testing when the initial 1584 was written. And that is why some places are making modifications to insert a OCPD between a txf secondary and the panel.

How does the POCO determine the available fault current at the service transformer secondary? Do they actually calculate line impedance all the way back to the power plant alternators? There has been significant development around here and I know for a fact utility lines have been upgraded, however the POCO did not change their estimate from three years ago to this year.

Be careful with this. The POCO's typically will provide, when first asked, the infinite bus fault current which is based on the txf kVA size, sec voltage and typical impedance so this will never change. What I do is ask the utility for the available fault current at the primary fusing, the primary fuse data, riser cable data and padmount txf data. I model this so I have a fairly accurate available fault current to work my model with and a reasonable incident energy value for the service entrance equipment.

If the fault current is not changing in spite of utility changes you know of then either you are getting the infinite bus current which should not be used for incident energy analysis or the utility is not timely in upgrading their modeling.

 

[EC Dan]

This is the text of the fault current letter provided:

Thank you for contacting FPL about the available fault current at *location*. Based on the plans you have provided dated July 29 2021, the maximum available fault current at the transformer secondary terminals is estimated to be 10048 symmetrical amperes at 277/480 volts. The protective device on the line side of the transformer currently in plane or to be installed and servicing your property located at the subject location is a 50 amp type K fuse. The primary service voltage is 23kV L-L. The calculated symmetrical fault current is not intended for use at the basis for motor starting calculations and does not include consideration for any motor contribution or fault current asymmetry.

I will ask whether this is infinite bus but response times are long. For people who may have seen many more of these than me or who deal specifically with FPL, does the text of this letter indicate it is an infinite bus calculated value?

 

[jim dungar]

This is the text of the fault current letter provided:



I will ask whether this is infinite bus but response times are long. For people who may have seen many more of these than me or who deal specifically with FPL, does the text of this letter indicate it is an infinite bus calculated value?

What is the transformer size and impedance?

The value they gave you looks like an infinite bus calculation of a 500kVA transformer.

My experience has been very few POCOs provide actual fault current data for customers purchasing power at less than medium voltage. Many engineers, who regularly perform arc flash studies, have developed methodologies for dealing with cases like this.

 

[EC Dan]

It's 300 kVA with impedance of 3.91 pu. With -10% adjustment to impedance for maximum fault current, I calculated 10.254 kA. With no adjustment, I calculated 9.22 kA. For our infinite bus numbers to match, their adjustment would have to be -8.15%, which seems like a strange modifier to use.

 

[jim dungar]

Many times utilities will give you fault currents based on the largest size transformer, that will fit into your mounting locations. This allows them to upsize your unit, if your usage turns out to be higher than their initial loading calculations, or as an emergency replacement.

 

[EC Dan]

Seems crazy that I've had to wait months to receive the results of a 30 second calculation that anyone can do with the information on the name plate. Anyway, I've asked the question so I'll just have to wait to see what they say.

 

[wbdvt]

Seems crazy that I've had to wait months to receive the results of a 30 second calculation that anyone can do with the information on the name plate. Anyway, I've asked the question so I'll just have to wait to see what they say.

Again that 30 sec calculation you mention is infinite bus fault current, not the available fault current which is needed for an arc flash study. Please understand the difference between the two and what the use of each is for.

Infinite bus fault current - determined by transformer secondary voltage, kVA and %Z. Useful in design phase to specify AIC and SCCR ratings of equipement.

Available fault current - this is the amount of fault current that is based on the fault current that the utility is capable of delivering. This is always going to be less than the infinite bus fault current. This is the value that needs to be used for any arc flash study.

My experience with POCO's is that typically the first contact will be a customer service person who has chart provided by engineering that is infinite bus fault current values. You have to be insistent on what you need.

I model from the utility primary line in. So I get the available fault current and X/R at the primary protective device (typically fuses), protective device data, riser cable data, transformer data. This way I have an accurate incident energy value at the service entrance equipment. I also get the X/R value so thru out the model, I am able to determine the equipment duty adequacy.

Remember that use of infinite bus fault current may result in lower incident energy values than what is really there. This is because an artificially high fault current will result in a protective device tripping in the instantaneous region whereas in reality it may be below the instantaneous trip point. This will increase the fault time and hence, the incident energy value. Workers may not be wearing adequate arc rated PPE in these circumstances.

 

[EC Dan]

I do understand the consequence of using artificially high fault current for these studies. I was just trying to argue with myself that they wouldn't make me wait months for an easily calculated value and were instead modeling out the fault current for me, however they have responded today and they say the fault current is based solely on transformer and fuse size so it appears I do not have an accurate fault current.

wbdvt: When POCO gives you available fault current at the primary fuse, isn't that value also just an estimate of maximum fault current at that point? I understand your model will definitely improve compared to if you just used the service transformer secondary fault current as reported by the POCO, but it seems like it's still all based on the wrong information.

By the way, the POCO stated I was only the customer to ever ask for this information, so either I'm not understanding something or almost everyone else is doing arc flash incorrectly (or just not using the incident energy method). Is it ever acceptable to just apply a conservative derating factor to the maximum fault current, like 80%? What are typical actual fault currents compared to infinite bus fault currents?

 

[jim dungar]

By the way, the POCO stated I was only the customer to ever ask for this information...

I have heard that from POCOs for the almost 20 years I have done studies.

For 480V and less customers, I would guess I have gotten meaningful fault current data probably less than 10% of the time. Many engineering firms have methodologies to deal with this issue.

Good luck.

 

[wbdvt]

Utilities now usually have a fairly accurate model of their system built using various software platforms geared towards utility systems so they can provide the fault current on the primary lines. They need to do this internally to determine such things as relay settings, coordination, line losses, loading restraints, etc. so it is not based on wrong information.

I have heard that from POCOs before and call BS on it. Usually what happens is it may be a new request to that specific customer service rep. I have had to get persistent with the POCO to get the information. I will usually cite OSHA and IEEE references on justifying the POCO to provide the information. Sometimes I will remind them that per OSHA 1910.269, they had to perform an arc flash assessment on their own system and did they use infinite bus fault current?

 

[EC Dan]

I appreciate all the feedback. I will try to push for better information from them.

 

[wbdvt]

EC Dan - not sure how the utilities commission in Florida is but in one state, I filed a complaint with the state utility commission and after educating them on the matter and not buying the utility's arguments, directed the utility to provide the information I was asking for.