fault current reduction

[don_resqcapt19]

There is a 2000kva transformer with 3-phase 480 volt solidly grounded secondary. It is connected to metal clad switchgear with 3000 amp rated bus duct. There are a number of 600 amp breakers that are feeding MCCs with parallel 350 kcmil copper. The available fault current at the MCCs is between 32 to 39kA. The MCCs are rated for 22kA. Would it be feasible to install a reactor between the transformer and the metal clad switchgear to reduce the available fault current at the MCCs? I looked around on the net a bit, but did not find much in the way of information on using a reactor for this purpose. This installation has been in place since the early 70's and the problem just came to light as a result of a short circuit study that was being done for the purpose of specifying arc-flash PPE.

Note, this is just for my information, the method of correcting this problem will be specified by a design engineer. There are a number of possible solutions being considered including MCC replacement and reactors for each MCC and maybe others.

 

[ron]

Don,
I've looked into using reactors for this purpose. There seemed to be solid core and air core types. The air core, had a lot of magnetic fields around it, so it had to be clear of ferrous metal for a good distance, so that was "out" in my application.
The solid core, had voltage drop problems during inrush for motors and such, so that one was "out" too.
The two main methods I've used is installing an upstream OCPD if it helps you get a series rating with the underrated MCC.
The second, is increasing conductor length.
I've also thought about an isolation trans, but then you deal with extra heat and losses.

 

[charlie b]

I have heard of reactors being used for this purpose, and for the purpose of facilitating selective coordination. But I have no experience in using them myself. So I can't offer any advice on that score.

Current limiting fuses is another option worth considering. But the first thing I would do, if I were assigned to the project, is to review the input data used to perform the short circuit study. If the model isn't right, the results won't be right. If, for example, the cable lengths, as modeled, is shorter than the actual length, then perhaps the available short circuit current is less than you think.

Another possibility is to use the model to artificially increase the cable lengths, until you get an answer that is less than your 22K limit. This will let you know if you have the option of simply replacing the feeders with longer conductors, in order to cut down the available fault current. Believe it or not, I have seen this one actually installed in the field. The feeders were run "around the room a couple times," in order to get the minimum required resistance.

 

[augie47]

Don, I'm short on details, but I inspected two older schools where additions generated the same problem. Engineers designed a system with reactors outdoor in the incoming service and reduced a sizable AIC to work with the 5k existing gear. To my knowledge there were no bad side effects.

 

[jim dungar]

I do not consider, reactors to be a common solutions, although they are often discussed. I have never spec'ed one and have never even seen one installed.

Back in the early 70's, series combination ratings did not exist (especially for MCC's) so you would probably need to fall into an "engineered" solution.

Electrical gear that is almost 40 yrs old should be considered for replacement, unless it is in 'well maintained' condition.

 

[weressl]

There is a 2000kva transformer with 3-phase 480 volt solidly grounded secondary. It is connected to metal clad switchgear with 3000 amp rated bus duct. There are a number of 600 amp breakers that are feeding MCCs with parallel 350 kcmil copper. The available fault current at the MCCs is between 32 to 39kA. The MCCs are rated for 22kA. Would it be feasible to install a reactor between the transformer and the metal clad switchgear to reduce the available fault current at the MCCs? I looked around on the net a bit, but did not find much in the way of information on using a reactor for this purpose. This installation has been in place since the early 70's and the problem just came to light as a result of a short circuit study that was being done for the purpose of specifying arc-flash PPE.

Note, this is just for my information, the method of correcting this problem will be specified by a design engineer. There are a number of possible solutions being considered including MCC replacement and reactors for each MCC and maybe others.

Installing reactors or specifying higher impedance transformers are engineered solution to high fault current availability mitigation. Done routinely on MV systems.
Adding it to existing systems should not be a problem either, but review large motor starts as your system may no longer be stiff enopugh to start them or cause unaccptable voltage dips on start.

 

[don_resqcapt19]

Installing reactors or specifying higher impedance transformers are engineered solution to high fault current availability mitigation. Done routinely on MV systems.
Adding it to existing systems should not be a problem either, but review large motor starts as your system may no longer be stiff enopugh to start them or cause unaccptable voltage dips on start.

Laszlo,
What would be a large motor for this application? I don't think there would be any over 100 hp on any of the MCCs in question.

 

[don_resqcapt19]

I do not consider, reactors to be a common solutions, although they are often discussed. I have never spec'ed one and have never even seen one installed.

Back in the early 70's, series combination ratings did not exist (especially for MCC's) so you would probably need to fall into an "engineered" solution.

Electrical gear that is almost 40 yrs old should be considered for replacement, unless it is in 'well maintained' condition.

As far as an engineered solution, I am not sure the engineer wants to stick his neck out that far

Many of the buckets have been replaced over the years and the plant is well maintained. The biggest issue with MCC replacement is the high labor costs. All of the loads are piped into the MCC with rigid, using one conduit per load. Often the first pull point is well away from the MCC and the amount of conduit in the room prevents you from removing the conduit from the MCC unless you first pull back the conductors to prevent them from being damaged. In some cases it will be impossible raise the conduit up enough to remove it from the MCC, even after the conductors have been pulled back.

 

[Pierre C Belarge]

Don
I know this is a particular manufacturer, but this may help you with some ideas.

Eaton Arc Flash Solutions

 

[don_resqcapt19]

Don,
I've looked into using reactors for this purpose. There seemed to be solid core and air core types. The air core, had a lot of magnetic fields around it, so it had to be clear of ferrous metal for a good distance, so that was "out" in my application.
The solid core, had voltage drop problems during inrush for motors and such, so that one was "out" too.
The two main methods I've used is installing an upstream OCPD if it helps you get a series rating with the underrated MCC.
The second, is increasing conductor length.
I've also thought about an isolation trans, but then you deal with extra heat and losses.

The engineer has taken a look at increasing the feeder conductor length using a computer program but it takes a lot of conductor to get where we want to go. One of the MCCs fed from the same switchgear has a feeder that is about 200' long and that one is within its rating. The others are in the same room as the switchgear and the feeders are under the slab and only about 30' long.

 

[Pierre C Belarge]

The engineer has taken a look at increasing the feeder conductor length using a computer program but it takes a lot of conductor to get where we want to go. One of the MCCs fed from the same switchgear has a feeder that is about 200' long and that one is within its rating. The others are in the same room as the switchgear and the feeders are under the slab and only about 30' long.

Coil the conductors in the floor and you can add a charge for radiant heat.

 

[steve066]

Since you indicated the study was completed as part of an arc flash analysis, I assume that real utility data was used on the primary side of the transformer, and the analysis wasn't done with an infinate primary. But I would suggest you verify that. No sense in changing gear due to artificially high numbers from an infinite primary calculation.


If there are only a few large motors that reactors might cause problems with, maybe a combination of reactors and some soft starts would be cheaper than replacing the MCC's.

 

[Besoeker]

There is a 2000kva transformer with 3-phase 480 volt solidly grounded secondary. It is connected to metal clad switchgear with 3000 amp rated bus duct. There are a number of 600 amp breakers that are feeding MCCs with parallel 350 kcmil copper. The available fault current at the MCCs is between 32 to 39kA. The MCCs are rated for 22kA. Would it be feasible to install a reactor between the transformer and the metal clad switchgear to reduce the available fault current at the MCCs? I looked around on the net a bit, but did not find much in the way of information on using a reactor for this purpose. This installation has been in place since the early 70's and the problem just came to light as a result of a short circuit study that was being done for the purpose of specifying arc-flash PPE

Note, this is just for my information, the method of correcting this problem will be specified by a design engineer. There are a number of possible solutions being considered including MCC replacement and reactors for each MCC and maybe others

I guess fault-limiting reactors are a solution. Replacing the MCCs might be a more elegant solution and, given that they are over thirty years old, there could be a case for doing that anyway.
Or perhaps it be feasible to modify the existing MCCs to withstand the 39kA fault level?

 

[rbalex]

I have successfully used reactors with both LV and MV systems. “Successfully” is not to be understood as a pleasant experience, although the MV systems were fairly straightforward.

In the early days of my career it was common practice to use transformers with non-standard impedances in situations such as you have described. You let the transformer manufacturer do the cable coiling for you. However 1500kVA was about the top that I remember and it would appear you would need something on the order of a 10%Z. They also have some of the voltage regulation problems Ron mentioned and are obviously energy wasteful. Since your MCCs are 600A, I suspect your motors may not be too large, so starting may not be too big a problem.

I’ll also echo Steve’s comment; make sure you’re using the utility’s actual available fault current and you’ve taken credit for every bit of impedance you can. Standard thermal-magnetic breakers may not be current limiting as we use the term today, but their thermal elements as well as motor overload heaters all add up.

 

[ptonsparky]

Sounds like replacing the transformer would be easier.

 

[don_resqcapt19]

A couple more questions.

Can you really use a "series" rated system to protect the MCC bus? I thought that a series rated system involved two OCPDs in series.

How big of an issue is the magnetic field around an air core reactor? Is it an issue with the normal loads or only when there is a fault? It is my understanding that an air core reactor is prefered for this application because a solid core one could go into saturation at high currents and not provide the required current reduction.

Some of the MCCs are still in production. Could we order a new one, strip out all of the buckets and busing from both the old and new MCCs and install the new buckets and busing in the old structure? I know the parts all fit as I have done that to replace bad vertical bus sections in the past. I expect that the issue with this solution would be the lack of a listing agency label.

i would love for the answer to be "replace the MCCs" as my quick guess is that would require some 3,000 man hours of work

 

[rbalex]

Laszlo and I appear to be on the same track here.

Your engineer will need to do the appropriate studies of course; but 100 hp against a 2000kVA transformer, even with a higher impedance, should be doable.

In response to your question about gutting a new MCC for parts; if it worked, you probably could have gotten by with rebracing the bus and getting higher kAic rated buckets. Contact the manufacturers. I've done it on MV gear, but never tried with LV.

 

[jim dungar]

Can you really use a "series" rated system to protect the MCC bus? I thought that a series rated system involved two OCPDs in series.

Technically, yes a "series combination rating" is for protective devices. However it is possible to have an upstream current limiting device protect lower rated 'non-active' components like bus bars (i.e. non protective devices). This is the 'up over and down' that Bussmann discusses in their SPD brochure.

Some of the MCCs are still in production. Could we order a new one, strip out all of the buckets and busing from both the old and new MCCs and install the new buckets and busing in the old structure? I know the parts all fit as I have done that to replace bad vertical bus sections in the past. I expect that the issue with this solution would be the lack of a listing agency label.

Check with the manufacturer to see if a field 'rebuild and upgrade' is one of their offerings. If a 'listing label' is important there are options for field certification.

 

[steve66]

What are the odds the only difference between your 22 KA rated MCC and a new 40 KA rated MCC is the label??

After all, if you can replace the buckets, wouldn't the only thing left be the bus bars? It seems like they would be braced against more than 22 K.

 

[jim dungar]

What are the odds the only difference between your 22 KA rated MCC and a new 40 KA rated MCC is the label??

After all, if you can replace the buckets, wouldn't the only thing left be the bus bars? It seems like they would be braced against more than 22 K.

Most likely the difference is simply the bracing of the vertical bus bars in each section. 22k SCA bracing was very common on MCCs until the early '80s, but most manufacturers had an option for higher ratings.