Main Tie Main Closed Transition Re-Transfer

[elmatador]

Hi Everyone,

We have a main-tie-main configuration for our system. When reverting back to normal service we want to use a closed transition re-transfer so that we don't have stop all the motors in the facility and cause a process upset. Both A and B sources on the M-T-M are fed from a utility source. When carrying out the short circuit study, I noticed that the fault levels are quite substantial when paralleling up the two transformers (which makes sense). It is around 78kA and the gear is only rated for 65kA. Since this re-transfer would only take a couple of cycles, is it recommended to beef up all the switchgear and mcc ratings to Sym. 85kA? The parallel operation will never be used in a Continuous Operation. Please let me know your comments.

Thanks in advance.

 

[living2learn]

I would combine both sources and rate the gear as required like you described. If a fault were to happen during this transition it would be most unfortunate if the gear could not handle the current appropriately. Are both sources fed from the same utility substation? I would add a sync check 25 in between the mains or at least some sort of protective relay with that or similar functions.

 

[ScottyUK]

I agree with living2learn.

I am aware of organisations which do allow short term paralleling in excess of the capability of the equipment, but under British legislation there is no provision for breaking the rules for a short period and the regulations require that equipment shall be rated for the duty. I'm not going to try to save money and by doing so potentially put myself and my employer in a very awkward position in a courtroom. Perhaps more importantly I don't ever want to have to explain to the family of a dead workmate why I decided to cut corners and put cost ahead of people. Bad enough if an accident ever did happen, but if it ever did happen I'd like to know that it happened in spite of me doing everything I could to make things safe. I'm not old enough to have been to the funerals of three workmates - it changes your perspective on taking risks at work.

 

[vandal06]

Scotty, what kind of effort would be involved in increasing the fault bracing ratings of the switchgear and MCCs?

 

[dpc]

This type of situation is fairly common, but is still an NEC violation in the US, at least IMO. You also have to consider the magnitude of the arc-flash incident energy when the transformers are paralleled especially if there will be workers standing in front of the gear doing the switching. You might check with the switchgear vendor regarding the bus bracing of the gear (regardless of the nameplate). It may be possible to simply upgrade the circuit breakers. As a minimum, you really want to be able to all of this switching without anyone standing anywhere near the gear during this transition.

 

[ScottyUK]

Vandal,

Somewhere in the difficulty range between a nameplate change through replacement of breakers to virtually rebuilding the busbar structure. Depends on the design.

 

[waross]

How about adding some impedance to the feeders. I have some very old text books that show a device used to add impedance to a transformer to allow parallel operation at full rating with a transformer of higher impedance voltage.
It was a stack of transformer laminations about 5 inches square and about 10 inches high with a hole in the center. One secondary conductor passed through the hole and a few percent impedance was added.
Another possibility is to consider the impedance of the feeders. Perhaps the feeders have enough impedance to bring the actual fault level at the switchgear below the rating of the gear.
Maybe you have already done this.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[elmatador]

Thanks everyone for your comments. Waross thanks for pointing out the impedance factor. This is going to be a brand new installation but we still need to look at the economics. We have to buy brand new transformers for the incomers. I changed the impedance settings of the transformer and we are able to bring the fault contribution from both sources to well below 65kA. Each incoming transformer would require atleast 9-10% impedance to bring the fault levels down. Would it not be cheaper to have a higher impedance transformer than to raise the interrupt rating for the switchgears and mcc's to 85kA? Each transformer is a 2.5MVA. I understand that The impedance I am looking for is considered quite high at this MVA rating and would also affect losses. Would there be a significant cost impact to pursue this option instead? Are there any negative implications of doing this? We would request the transformer vendor to ensure that the tested impedance is a minimum of 9-10%.

Please let me know your thoughts. Thank you once again for taking the time to respond. Much appreciated.

 

[waross]

You would be best to address cost questions to your vendor.
I remember seeing an installation where the primary feeder was used to increase the impedance. Calculations showed that 100 feet of primary feeder cable would add enough impedance to bring the fault current below the rating of the switchgear. The spec was that no feeder shall be LESS than 100 feet.
Depending on the layout, cable impedance may be used on the primary, the secondary or both.
I have also seen an air core reactor installed to limit the fault current. If you consider an air core reactor you may want to consider the stray magnetic field. An air core reactor installed close to a concrete wall may cause unacceptable heating of the rebar under fault conditions.
I would ask the transformer vendor for advice on availability,price and shielding.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[rcw retired EE]

Many installations do a momentary paralleling where the short circuit levels during the few cycles of closed transition may exceed the rating of the downstream equipment and the bus. The Main and Tie Breaker's rating cannot be exceeded because fault current through those breakers can only be from one source. It is the feeders that end up with combined currents.

Some designs are as simple as a timer that trips the selected breaker x cycles after all three breakers are closed. More sophisticated designs use breaker failure relaying plus bus differential and fiber optic arc sensing to trip the Mains if a fault occurs during the switching. Or the controls are designed to allow all breaker operations from a remote location so no personnel need to be near the switchgear during the transition.

Another design is to use a bus transfer relay like the one made by Beckwith. It can do an open transfer (Trip Tie, Close Main or Trip Main, Close Tie) without dropping loads if there are enough large motors on the bus. It can also do the other schemes mentioned above plus breaker failure, plus delay manually initiated breaker operation long enough for the operator to leave the room.

Bottom line, is closed transition against Codes? Yes. Do many industrials do it and provide other protections to minimize risk? Yes. Your call.

Adding impedance to reduce fault levels does work, but it may affect your motor starting voltage drops.

 

[ScottyUK]

To be clear - closed transition is against code, or exceeding the capability of equipment is against code? In my patch closed transition, or indeed operation in parallel, within equipment capability is completely acceptable.

 

[Mila15]

We have installed few systems using the M-T-M configuration. When an Auto transfer occurs, I don't think we have a problem because this happens if one source is lost. It's the manual transfer we are worried about. In the last design, it worked out fine because One feed could not handle the entire load, so we had to implement a load shedding scheme, where interlocks and control circuits prevents the tie breaker from closing unless certain loads are removed. This might be one way worth considering. An other alternative is a reactor, as waross suggested, but you need to make sure you don't end up with voltage drop and/or motor start issues. There are many other solutions, but whatever you end up doing, you need to consider all issues: voltage drops, motor start, short circuit, and arc flash.

 

[rcw retired EE]

Scotty - According to most interpretations of codes, exceeding the short circuit rating of the connected equipment during a momentary paralleling event, no matter how short, is against the code. Paralleling two sources which do not cause the short circuit rating to be exceeded is permissible.

 

[ScottyUK]

Agreed.

 

[waross]

I have installed a couple of very fast open transfer switches on small standby generator sets. It may be worth investigating the possibility of installing a very fast open transfer scheme as suggested by rcwilson.

Bill
--------------------
"Why not the best?"
Jimmy Carter