To Instantaneous or Not on LV MCC

[nollij]

This issue keeps coming up in my head and to be honest I can not settle on one side of the fence.

When implementing LSI (Long time, Short time, Instananeous) LVPCB (Low Voltage Power Circuit Breakers) for MCC main breakers, do you utilize the instantaneous function or not?

MCCs are only rated for their Short Circuit Withstand rating for 3 cycles. A short downstream of any of the MCC feeds will be cleared by the MCCBs or MCPs. So, the only issue would be when the fault occurs on the buswork of the MCC.

The argument from a Power Systems Engineer that we regularly hire for our Power Systems studies contends that if there is a fault on the buswork, the gear is toast anyway. So, don't utilize the instantaneous function to maintain coordination. This contradicts the settings he utilized for a Arc Flash reduction project that was undertaken before I started working here where the instantaneous functions are being used.

A short circuit downstream may cause an entire MCC and resulting unit outage if the instantaneous function is used which, is much more likely to occur than a fault on the MCC buswork.

However, I am not so sure that a fault on the MCC buswork would cause significant damage if it is cleared instantaneously. This would allow the equipment to be opened up and the situation potentially fixed. It could be a while for a replacement MCC to be installed, potentially keeping the unit down if there are no steam backups for that unit.


My question to all of you is, what do you do for the MCCs at your facilities that incorporate LVPCBs as MCC main breakers? Or, what do you typically default to if the customer does not specify to which they prefer? Basically, which side of the fence does everyone tend to fall on?

 

[ron]

For equipment rated for 3 cycle only, such as most MCC's, switchboards, transfer switches, panelboards, etc, I definitely use instantaneous settings. The code requires that we protect equipment at their ratings.

I realize you can buy specialty gear and transfer switches that have 30 cycle ratings, and I do that when I don't want to use instantaneous upstream of them, but when the downstream is 3 cycles, then instantaneous is the way to go.

 

[nollij]

The code requires that we protect equipment at their ratings.

At the potential risk of a process upset from an uncoordinated fault clearing though?

 

[dkidd]

You should look at 240.87 for the 2011 Code. There was a good article on NECPlus, but now it is only for subscribers. A web search on 240.87 will probably find more information.

 

[zog]

Using a maintenence switch gives you the best of both worlds. We do a ton of retrofits of LVPCB's for this specific purpose.

 

[kingpb]

IEEE 1015 (Blue Book)

General application considerations for the main circuit breaker, i.e. feeding MCC, is that the preferred trip functions for selective trip are long-time and short-time (and ground fault, if required). For coordination purposes, instantaneous should be provided only if necessitated by the circuit-breaker interrupting rating.

The estimated instantaneous clearing time for a LV PCB is 3 cycles.

IEEE Std C37.20.1-2002, ?The rated short-circuit withstand current of a LV ac switchgear assembly is the designated limit of available (prospective) current at rated maximum voltage that it shall be required to withstand for a period of no less than 4 cycles on a 60 Hz basis under the prescribed test conditions.

As a result, LVPCBs, ICCBs, and MCCBs must interrupt a fault within three cycles at maximum short-circuit current to protect the bus. It may necessitate using an instantaneous unit on the main circuit breaker to protect the bus at the sacrifice of obtaining selectivity.

If the concern is for arc flash, the maintenance switch, as already mentioned is a good option. But first you must look at protection of the bus.

 

[nollij]

Most LVPCB circuit breaker Short Time delay units operate around 0.1 sec (6 cycles). So, not applicable for protecting against short circuit current at the level of the rated equipment fault withstand.

I am not concerned about Arc Flash issues (the STD typically handles that quite well, not necessitating the instantaneous unit for that purpose). Merely, I want to know whether people prefer to put the MCC at risk to prevent miscoordination. The miscoordination is a far more likely outcome as bus faults are rare compared to feeder faults. A miscoordinated fault clearing by the main breaker would result in a process unit upset which could be far more dangerous than a broken MCC. But then, what is the point of the LVPCB being there if it is not protecting the bus besides lowering the Arc Flash and providing relatively pointless coordination?

What do other facilities consider to be the bigger risk?

I had not had a chance to go over the 2011 NEC yet. 240.87 is new news to me!

The ultimate solution to this whole problem is to standardize IEEE ratings to be for 6-8 cycles (typical STD times) instead of 4 cycles. I have thought this for a while now.

 

[jghrist]

As your Power Systems Engineer says, you are not "putting the MCC at risk" to prevent miscoordination. An external fault will be cleared by a feeder breaker instantaneous trip. The main breaker cannot protect the MCC from an internal fault. If an internal fault occurs, the MCC has already failed.

 

[kingpb]

Why not plot the curves and see what shakes out. It could be that with the Instantaneous set, you might still achieve selective coordination. If not, then you will have to decide if the process is critical enough to risk setting it or not.

As far as Arc Flash, the extended time of the short time setting will directly increase the energy released, so I would not discount that the instantaneous is not required in order to achieve the desired AF incident energy at the boundary.

 

[templdl]

This is the way I have always looked at this application:
The interesting thing about the instantaneous element of breakers is that should there be a bolted fault on a feeder or branch circuit the fault current will also be seen by the main breaker. Everone on use would like to assume that the breaker closest to fault would trip and clear the fault first before the main trips. But, should that fault current be high enough in magnetude where it would fall within the trip pickup of the main and feeder or branch breaker it would be pot luck on which breaker would trip. Remember that both the short time and instantaneous ajustment of a breaker are calibrated such that they are within the withstand of the breaker. By adjusting the short time delay you may get some coordination with low level faults with down stream breakers, that is the main breaker would delay slightly in order give the down stream breaker a chance to clear the fault.
Usually the instantaneous setting of breakers come set in there lowest setting and most installers that I'm aware of reset them in there highest setting to prevent nuisance tripping.

Coordinating for faults is real ambiguous as there is no sure solution to assuming what they will be. The magnetude of a fault could be anything. Arcing faults are a real problem as they may not be high enough to trip a breaker instantaneously and not enough to trip a breaker on overcurrent. Quite often arcing faults are line to ground where ground fault is a real advantage.

 

[mull982]

Very interesting discussion.

My take on selective coordination vs MCC/Switchboard withstand ratings is how far down you need to achieve selective coordination to. If selection coordination is only required down to .1 s then using an instantaneous setting on a main breaker can still be considered coordinated in this case as well as protect the 3-cycle rating of the gear. If coordination is required down to .01 s then this is a case where 30-cycle gear may be considered in order to consider not using an instantaneous setting on the main in order to achieve selective coordination. Thats my understanding aanyway.

What is the withstand fault current for the 3-cycle rating of most gear? 65kA or something other?

I believe UL489 requires all breakers located in a switchboard to have an instantaneous setting in order to clear faults within 3 cycles. Is it possible on an LVCB to turn the instantaneous setting off? I'm not sure if this is a possible setting option? If so is this a violation of the the testing standards, or does the breaker still have some sort of overide that will trip the breaker in 3 cycles?

 

[jdsmith]

I'm a project engineer in a refinery as well. You are correct that in the past petrochemical plants generally could claim that the process risk of tripping 15-20 motors (a whole MCC) due to a single motor fault was a greater hazard than using LS protection (no instantaneous) on equipment that is only tested for 3 cycle fault duty. I don't claim to have a comprehensive understanding of equipment design or UL 845 (the dominant MCC testing standard), but I am aware of two parameters that we are concerned with in the equipment: the magnetic forces that will shake the conductors in the equipment, and the total fault energy that the equipment must withstand.

As an example let's take an MCC rated 65kA for 3 cycles:
1. The maximum magnetic forces this equipment must withstand will occur in the first half cycle of the fault, regardless of whether the fault persists for 1 cycle or 60 cycles or any period of time in between. The only way to limit this initial high "pulse" of current and corresponding magnetic forces is to use current-limiting devices for protection, but that is a whole different animal regarding coordination. The essential takeaway here is that the bracing in the equipment is designed for a peak current based on the 65 kA rating and the bracing is not closely related to the time duration portion of the rating.
2. The other part of the equipment rating is the total fault energy that the equipment can withstand, represented by I^2*t. For any given equipment withstand rating you can extrapolate the withstand energy to other points on the same I^2*t curve. For example, 3 cycles * (65 kA)^2 falls on the same curve as ~7 cycles * (42 kA)^2 and ~30 cycles * (20 kA)^2.

Whether you're comfortable with a physics-based understanding to set equipment protection is your choice. The salient point here is that while manufacturers only provide the withstand at a single rating for their equipment, in reality the equipment has a withstand curve and will withstand lower magnitude faults for longer durations. For application-specific guidance talk to your MCC manufacturer. They will likely be able to provide a much better theoretical answer than I did.

I agree in principle with your suggestion that the standards bodies look at ratings at alternate points, such as 6-8 cycles. What we in the user community need to do is demand this information from our manufacturers. For example, specify a UL 845 listed 3 cycle 65kA MCC also meeting an 8 cycle, 40 kA withstand, or some similar setting depending on your ST pickup and delay and coordination concerns. For a manufacturer to sell something like this requires actual testing not just analysis, but if a few of us start requesting this type of equipment they will do the testing and provide a second rated withstand point. This would allow us to be fully 'covered' on the ratings and protection side without having to use the exception allowing us to do things because the alternative presents a greater risk to the facility due to process plant hazards associated with transient operating conditions. I talked to a couple manufacturers about this exact issue last week at the PCIC conference and they indicated that they are aware of the issues, the changing safety landscape, and are open to doing this sort of testing.

 

[ron]

The code requires that we protect equipment at their ratings.

At the potential risk of a process upset from an uncoordinated fault clearing though?

Absolutely. That is almost always irrelevant to the code.

 

[nollij]

I believe UL489 requires all breakers located in a switchboard to have an instantaneous setting in order to clear faults within 3 cycles. Is it possible on an LVCB to turn the instantaneous setting off? I'm not sure if this is a possible setting option? If so is this a violation of the the testing standards, or does the breaker still have some sort of overide that will trip the breaker in 3 cycles?

You can order the breaker as a seperate piece of switchgear and specify an MCC with no main breaker.

You can order LS units without the instantaneous or just turn the instantaneous setting up higher than the maximum fault current on the bus.

I talked to a couple manufacturers about this exact issue last week at the PCIC conference and they indicated that they are aware of the issues, the changing safety landscape, and are open to doing this sort of testing.

Interesting. Unfortunately I was not able to attend the PCIC due to it being outside of the country but I will be in New Orleans for it next year!

For any given equipment withstand rating you can extrapolate the withstand energy to other points on the same I^2*t curve. For example, 3 cycles * (65 kA)^2 falls on the same curve as ~7 cycles * (42 kA)^2 and ~30 cycles * (20 kA)^2

I had not thought of this approach before.

If an internal fault occurs, the MCC has already failed.

That is pretty much what the Power Systems Engineer says.

It could be that with the Instantaneous set, you might still achieve selective coordination.

As far as I am aware, the only coordination that is achieved below 0.01s in terms of coordination curves is achieved with fuses by the same manufacturer. They make tables on what the next size that will coordinate before "pickup" of the upstream fuse.

 

[jim dungar]

As far as I am aware, the only coordination that is achieved below 0.01s in terms of coordination curves is achieved with fuses by the same manufacturer. They make tables on what the next size that will coordinate before "pickup" of the upstream fuse.

Most breaker manufacturers now have similar tables, based on actual testing.

 

[tish53]

Thanks for the great discussion, I have learned alot just reading this thread.
We spec all our new MCC main breakers with an inst. for two reasons. Many of our arc flash caculations show the inst. is needed to minimize the hazard level and as a back-up for a malfunctioning feeder breaker ( or MCP on a motor circuit). We live with the coordination issues but so far have not seen any gross miscoordination situations.

 

[jdsmith]

Most breaker manufacturers now have similar tables, based on actual testing.

The widely-published tables are generally fairly conservative so they can apply to a number of situations and can be applied fairly simply by many folks. For any large customer that wants to push coordination to higher fault current levels than the tables indicate will coordinate, talk to the manufacturer's Application Engineer. If you furnish your single line and short circuit study, they can often provide selectivity guidance that is more specific than the tables that are published for all the world to see, but their guidance may have more caveats and be for more specific situations. This higher-level guidance is not widely distributed because the average user wouldn't understand it completely and may apply it improperly, but the manufacturers will put it in writing and stand behind it once they are sure it applies to your situation, and once they are sure they are dealing with people that understand the limitations of the more advanced selectivity guidance.

We live with the coordination issues but so far have not seen any gross miscoordination situations.

A small amount of experience in only a few plants is not statistically significant compared to rigorous testing conducted by manufacturers and years of tabulated experience across hundreds of installations. We are fortunate in the electrical industry to have resources like the IEEE Gold Book (IEEE standard 493) that contains reliability data that has been tabulated across many industries for forty years. Other IEEE papers are also available that dive more deeply into certain areas like motors and safety statistics. You could obtain a more rigorous Total Cost of Ownership analysis by reading the IEEE Gold Book and examining the nature of your operations to determine how much money it makes sense for your plant to spend on breaker coordination. There is always a trade-off between initial cost, reliability, service life, and maintainability and the job of a plant engineer is to optimize that balance given particular operating practices and cost of different types of downtime.

I'm not saying everyone needs 100% selectivity between all of their layers of overcurrent devices - far from it. Each facility has different costs of downtime and different abilities to schedule outages, so each facility should analytically determine the solution that provides the lowest total cost of ownership. Some of the manufacturer's application engineers I know are quite good at doing this, but since their conclusion is that the lowest TCO is obtained by spending more $$ on equipment upfront many people don't believe them even if they are correct.

 

[jim dungar]

This higher-level guidance is not widely distributed because the average user wouldn't understand it completely and may apply it improperly, but the manufacturers will put it in writing and stand behind it once they are sure it applies to your situation, and once they are sure they are dealing with people that understand the limitations of the more advanced selectivity guidance.

Schneider Electric, Eaton, General Electric, and Siemens all publish their '100%' breaker selectivity tables for all to see on the world wide web.