Parallel feed to Transformer Primary

[JohnHurst]

I work for a company that designs and manufactures motor controller equipment. Since our development lab is in a facility that only has 208V 3-phase available, we are running through a step-up transformer to get 480V to play with. Because things sometimes go wrong with new designs, and we occasionaly blow the fuses/breakers in the electrical room (to which we don't have private access), we installed two separately metered 200A service panels that feed the same lugs on the transformer. The thought being that, if we trip the breaker in the electrical room on one disco, we can - after correcting the problem in our device - switch over to the other disconnect and continue on while we wait for the landlord to reset the original breaker. This arrangement works fine.

My question is this: If I were to turn on both disconnects at the same time, would that essentially double the current available to me to 400 amps? The transformer is rated for it, so that's not a problem. The conductors are all properly phased, so my first guess is that I would simply have 400 amps available at the primary. I'd rather not have to install a 400 amp service if I can avoid it, as I will only have a very occasional need for that much power.

 

[iwire]

If they are both closed at the same time you will have a violation of 240.8 (Fuses or Circuit Breakers not permitted in parallel)

There are also likely issue with the over current protection for the transformer and one disconnect being back fed by another.

In short what you have is not the way to handle the problem and may kill an unsuspecting person working on the system.

To me it sounds like you work in a dangerous environment.

 

[petersonra]

we installed two separately metered 200A service panels that feed the same lugs on the transformer

The utility allowed this?

this would also appear to violate the rule on having only one service per structure.

 

[JohnHurst]

Setting aside the issue of code violations for a second, what problems do you envision with the overcurrent protection for the transformer? The only issue I can see would be if the system would tend to draw the current in an unbalanced fashion (i.e., significantly more from one feed than another), in which case I'd expect first one then the other breaker to trip.

As far as the environment, this is a development facility with extremely restricted access, even more restricted access to the disconnects, and only trained personnel permitted to come in contact with the system.

I know that nobody can recommend violating the NEC; however, in theory, is my original premise correct, and I would have 400A available at the primary of the transformer?

 

[winnie]

1) It sounds like you are using the term 'service' when you really mean 'feeder'. A 'service' is the connection that the utility provides, a 'feeder' is a circuit which feeds OCPD (circuit breakers) which then feed branch circuits or other feeders.

2) I am not an electrician. I work in a motor research lab. So I think that I can identify with where you are coming from. (And this also means that we should probably talk more My guess is that you had a single 200A feeder, and then to solve the breaker reset problem, you had another 200A feeder installed. Now you need more power, and want to 'get er done' with what you currently have.

3) At the time of the installation, my bet is that no attempt was made to match the characteristics (length, conductor size or type, conduit, etc) of these two feeders. You asked for 2 separate 200A feeders, and that is what you got.

In theory two feeders can be placed in parallel, but only if they conductor characteristics are matched. The general requirement for parallel conductors is that they be matched in terms of these characteristics, so that balanced current flows on the parallel conductors. If you can confirm that the two feeders are exactly the same, then they could be placed in parallel to form a single 400A feeder.

4) iwire mentions 240.8 which prohibits OCPD in parallel, unless they are part of a listed unit. So there is nothing wrong with the _theory_ have having 2 200A breakers in parallel to act as a single 400A breaker, but there is almost certainly a problem with the separate 200A disconnects that you currently have. I'm sure that you could purchase a bit of custom built switchgear that has 200A disconnects, and is designed _and listed_ to permit you to use one, the other, or both. I bet that such a bit of switchgear is more expensive than a normal 400A disconnect.

5) Going back to the design of the feeders, if they were not _intended_ to be used in parallel, then it is likely that they were never tested for proper phasing. It is entirely possible that if you throw both switches, you will effectively place a bolted fault across these feeders. In _theory_ proper phasing solves this, but the difference between theory and practise could involve rapid self disassembly of your switchgear.

-Jon

 

[JohnHurst]

1) It sounds like you are using the term 'service' when you really mean 'feeder'. A 'service' is the connection that the utility provides, a 'feeder' is a circuit which feeds OCPD (circuit breakers) which then feed branch circuits or other feeders.

You are correct. I did mean feeder. It's a multi-tenant building.

2) I am not an electrician. I work in a motor research lab. So I think that I can identify with where you are coming from. (And this also means that we should probably talk more My guess is that you had a single 200A feeder, and then to solve the breaker reset problem, you had another 200A feeder installed. Now you need more power, and want to 'get er done' with what you currently have.

3) At the time of the installation, my bet is that no attempt was made to match the characteristics (length, conductor size or type, conduit, etc) of these two feeders. You asked for 2 separate 200A feeders, and that is what you got.

Sort of.... We just moved our development facility into this building, and I was having problems in the old place with breakers tripping and fuses blowing. I wanted a means of quickly getting back up and running while waiting for the landlord to reset the breakers at their panel. Unfortunately, after all of the planning was done, and the circuits installed, and the transformer wired up... the sales department got a serious request for a device that exceeds my 200 amps, and I need a way to test it.

In theory two feeders can be placed in parallel, but only if they conductor characteristics are matched. The general requirement for parallel conductors is that they be matched in terms of these characteristics, so that balanced current flows on the parallel conductors. If you can confirm that the two feeders are exactly the same, then they could be placed in parallel to form a single 400A feeder.

To the best of my knowledge and ability, the feeder circuits are matched. The meters and fuses in the landlord's electrical room are right next to each other. The electrical room is fortunately located right on the other side of the wall my disconnects are on, and the landlord's wiring just comes right through the wall separating us. I wired up the transformer, so the lengths of the cables are pretty near the same length (was easier to cut six conductors to the same length and just coil up the extra). I made absolutely certain to match the phases just in case both disconnects were on accidently. Not sure if I missed anything...

4) iwire mentions 240.8 which prohibits OCPD in parallel, unless they are part of a listed unit. So there is nothing wrong with the _theory_ have having 2 200A breakers in parallel to act as a single 400A breaker, but there is almost certainly a problem with the separate 200A disconnects that you currently have. I'm sure that you could purchase a bit of custom built switchgear that has 200A disconnects, and is designed _and listed_ to permit you to use one, the other, or both. I bet that such a bit of switchgear is more expensive than a normal 400A disconnect.

5) Going back to the design of the feeders, if they were not _intended_ to be used in parallel, then it is likely that they were never tested for proper phasing. It is entirely possible that if you throw both switches, you will effectively place a bolted fault across these feeders. In _theory_ proper phasing solves this, but the difference between theory and practise could involve rapid self disassembly of your switchgear.

-Jon

Again, I tried my best to ensure that the circuits were matched, if for no other reason than to eliminate differences when testing equipment. I know the phases are matched, the wiring on my end is the same. Because of the proximity of the electrical room to my facility, and the fact that both feeder panels were installed at the same time, I'd presume that the wiring on the other side of the wall is matched.

If you have any specific info that would be helpful, or want to take this offline, I'd welcome a PM with your experiences.

 

[winnie]

I do not have any personal experience with what you describe. I've simply never had to work at those power levels. If you can run your tests using a motor/generator type dyno setup, then perhaps you can run your tests with lower input power...but that is a whole 'nuther kettle of fish.

This being the internet, we could have a fantastic debate as to the legality of your current installation (presumably installed by a skilled non-electrician, in a well controlled environment, without inspection...) given that debates about the required circuit ampacity for a 15KW range can take many, many posts.......

Parallel conductors are permitted in specific situations. We could have a similar argument as to weather or not you actually have parallel conductors; clearly you want to have parallel current flows, but the conductors are not joined together at either end to act as a single conductor; each separate 'leg' of the system has its own OCPD.

Parallel OCPD are permitted in specific situations, but again we could have a long argument as to weather or not the OCPD would actually be in parallel.

If you consider your system as having parallel conductors or OCPD, then it violates the NEC. If you consider your system as having conductors which don't meet the strict definition of 'parallel', then you might argue that your questions are simply 'design issues'.

At this point I would say that basic theory says that your approach should work, but that you would be remiss if you don't get a second opinion from a suitably rated and _licensed_ professional who can look at the entire situation and catch possible errors. Speaking scientist to scientist, there are situations where what you really need is a skilled craftsman who can see your blind-spots.

-Jon

 

[dbuckley]

The OP is quite clear, there are two separate meters, so there are two separate services, and paralleling the feeders by closing both switches is paralleling the services, and I'm not sure the utility would be too happy with that.

In my opinion the whole arrangement is quite scary, and the two switches should have been mechanically interlocked at birth to prevent simultaneous closure.

 

[winnie]

If you read further in, these are in fact feeders. I figure that the building owner is sub-metering.

I agree: my first thought was that the feeds should be interlocked in the fashion of some sort of transfer switch. The more that I think about it, however, the less scary this seems. Strange design, but no more scary than transformer secondary ties (explicitly permitted by the NEC). With separate OCPD at both ends of each feeder, I don't see a safety issue. But I believe that this could really only be evaluated by an unbiased professional on site.

If this were my facility (and I had infinite funds), then what I would want is a single feeder, with standard OCPD on the building owner side, and some form of electronic OCPD on the tenant side, such that both the inverse time and instantaneous pickup could be coordinated to make the tenant side OCPD trip in all but pathological situations. However funds are always finite, and what we get with infinite funds is usually different than what we get with what is in the wallet.

-Jon

 

[zog]

You also need to consider the available fault current when both breakers are closed, you may exceed the AIC ratings of your OCPD's. Even if they are phased corectly other loads on the system can cause a phase shift, when you close the 2nd breaker you may get high fault currents from the phase shift.

 

[Les Fischer]

Dual feed

Dual feed

What is the size of your transformer.

Les:-?

 

[JohnHurst]

What is the size of your transformer.

Les:-?

Transformer is 112.5 KVA. My requirements are for ~115A on the 480 side which is about 280A on the primary side. The transformer should be more than capable of this. My initial choices for transformer (based on cost and availability) were either this 112.5 kVA or a 75 kVA (which would have been barely adequate for the 200A feed envisioned.

 

[JohnHurst]

You also need to consider the available fault current when both breakers are closed, you may exceed the AIC ratings of your OCPD's. Even if they are phased corectly other loads on the system can cause a phase shift, when you close the 2nd breaker you may get high fault currents from the phase shift.

Are you referring to other loads on the feeder panel that feeds my two disconnects? Or other loads on my end of the system? The reason I ask is that I only have one load on my system at any given time (not including the parasitic load from the transformer) and that is a single motor.

If it's other loads on the feeder, they should be expected to affect both my feeder panels pretty equally, given that they are right next to each other, installed at the same time, and there is only about 10 feet of 3/0 cable between the main panel and my feeds.

 

[Les Fischer]

Transformer

Transformer

I calculated the 480v and 208v sides.

480v = 135a

208v = 312a

For worker safety the 2 feeds are dangerous because which ever switch you have on,the load side of the other switch is hot . I would rethink the single source issue.


Les

 

[iwire]

The easiest, least expensive, safe solution to this is simply install a 200 amp manual transfer switch right at the transformer.

Essentially it is a three pole, double throw, center off disconnect switch.

With this switch you can only connect one supply at a time to the transformer.

 

[JohnHurst]

The easiest, least expensive, safe solution to this is simply install a 200 amp manual transfer switch right at the transformer.

Essentially it is a three pole, double throw, center off disconnect switch.

With this switch you can only connect one supply at a time to the transformer.

Of course, this defeats my original question which had to do with connecting both feeds at once. If I may summarize what I've gotten from this thread:

1. IF the conductors are matched, and
2. IF the phases are matched, and
3. IF there are no other factors that will unbalance the supply side or shift the phases, then

My plan will work. However,

1. It violates the NEC (probably, almost certainly), and
2. It is unsafe, hazardous, and dangerous, and
3. Nobody in their right mind would attempt to do so.

I don't plan on performing this feat more than once or twice for brief periods of time. I will take every precaution I can to ensure my safety and that of others. I also plan to take Bob's suggestion and place a switch at the transformer to prevent the load side of the "unused" disconnect from being live during normal operation. Is there anything else I've missed?

 

[don_resqcapt19]

John,

Is there anything else I've missed?

Only the fact that you need to make the correct code compliant installation. If their is ever an accident, you and your company will lose the civil suit, and it is very possible that criminal charges would be brought against you personally.
Don