Short circuit current during parallel operation

[EXTMAJO]

In a project of me, we allow the short parallel operation of two transformers on one switchgear. However, that means that the maximum short circuit current crosses the design values of the switchgear in this time.
Is this allowed for short-term switch actions?
In which standard is this described?
Where can I find addition information to this item?

 

[Alex68]

I have never heard about a standard describing this procedure.
You can choice among some procedures of bus bar change over. Only you, as the designer, can decide what is better for your plant. Usually the risk of damages during a short parallel is largely compensated by the benefits, because the probability of a fault during the parallel is very low and can be accepted (the parallel usually lasts 100 / 200 ms).
Otherwise if your plant permits a voltage drop, you can choose a residual voltage transfer.
Only you know what is the best for you plant.

 

[davidbeach]

Allowable time for the connection is 0.0000s. It is not allowable, per standards and applicable codes, to ever exceed the withstand or interrupting ratings of equipment.

 

[rbulsara]

I agree with davidbeach.

 

[alehman]

I also agree with David. It seems like a switching operation (e.g. paralleling) is a more likely time for a fault to occur.

 

[Alex68]

Maybe in USA the "short time parallel" is not allowed by your codes. Here in Europe it is a typical practice in power plants during the start up.
Wherever the generator is not equipped with a circuit breaker, the auxiliaries are fed by a different source during the plant start up. After the generator is synchronised, the auxiliaries are transferred to the auxiliry transformer by means of a "short time parallel".
This is a make-before-break procedure commonly accepted.
The parallel lasts just the time of the logic to send the open command. The probability of a fault during this procedure is very low and the risk is accepted.
Everybody knows that during the parallel the withstand rating of the switchboard is exceeded. But the circuit breakers are always able to break the current contributions.

 

[ScottyUK]

In my bit of Europe this sort of thing is normally interlocked out using a Kirk or Castell key arrangement to prevent switchgear ratings being exceeded. If paralleling is required then design it fit for purpose. If a designer knowingly designs in a means of exceeding the switchgear rating then he is failing in his professional capacity and in my opinion is either reckless or incompetent. Either way it is a tough position to defend legally if the inevitable fault eventually causes a substation to blow up and kill someone. Alehman is right - switching operations are the most likely time for a fault or failure to occur.


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[Alex68]

ScottyUK
I don't understand how you can define many generations of engineers that for 40 years designed power plants as "either reckless or incompetent".

You don't accept the risk for 100ms every month. OK!
They do!

In 40 years no incident happened during short parallel but many injuries caused by bypassed interlocks.

Please can you explain me how to make a bus bar change over during the start up of a 750MVA without a voltage dip?
please do not suggest the fast transfer, because no device correctly works (nor ABB neither Siemens).

 

[rbulsara]

Alex:

Accpeting risk does not make it right. In fact there are cases that we have some across in the USA, which were found deficient on SCCR during parallel operation, but they were noted as defciecies and risk was made known to the Owner. But it still does not make it right.

Anyone who deliberatley designs a deficient system is negligent.

 

[Alex68]

You all are right if there is a possible technical solution.
But what should we (me and you) do when no technical solution is possible?

I think that in any situation of life, we try to avoid dangers but if no other chance is possible, we evaluate the risk and decide if it is acceptable or not.

 

[rbulsara]

Assuming you are an consultant or similar capacity engineer or owner's employee, You and I (consultants) should not be deciding what risk an owner or insurer should assume. Our duty is to notify them all known risks and advise minimizing them.

I would tell them, if you parallel the gear you are at great risk and my advise is either to upgrade the system or not perform closed transition transfers. There are always solutions, some are more afforable/feasible than others.

If you are an owner (not owner's emplyee), consult your insurance carrieers and lawyers.

 

[Skogsgurra]

I once lost a customer that wanted me to sign a paper saying that his old 10 kV switchgear would work (that is including clearing faults) when connected to his new 10 kV busbar with around ten times more Isc. I am glad I didn't sign that paper. His gear eventually blew up.

I know that your situation is different. But you have to include some means of current limit if you are going to have these things in parallel. There are always solutions. They may be expensive - but they do exist.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Alex68]

Maybe EXTMAJO is offended by this philosophical discussion because nobody said him which standard he should apply.
Please suggest him the standards he needs.

I agree with all of you.
I was an owner's employee, now I'm a consultant and in both cases sometimes I had to impose myself to avoid dangerous situations.
Anyway managing a so strong energy like electricity is always dangerous. We can minimise the risk until we agree that it is acceptable, but faults occurs anyway.
Actually we use protections that work only after a fault happened.
Actually we make grounding grids to avoid voltages that have been defined as not tolerable = not acceptable.

If we need a plant, we must do it, so if we can't avoid that faults occur (we are not gods), then we make their probability or their consequences acceptable applying the correct procedures.

 

[rcw retired EE]

Momentary paralleling circuits in Main-Tie-Main switchgear are a common application in some industries. The Mains and Tie breakers and the switchgear bus are rated for the short circuits that could flow during paralleling. Only the feeder breakers are under rated. The paralled time is minimized by redundant control circuits sensing when all three breakers are closed and immediately tripping one of the three based on position of a selector switch.

During that paralleled time, a close-in fault on a feeder leaving the gear may exceed the feeder breaker's rating. The Main and Tie breakers ratings can't be exceeded because fault current can only flow in one direction through individual breakers. The two fault sources are not combined until they meet at the feeder breaker.

Given the above design, some owners accept the low risk of a feeder fault occurring during the 100 mSec that the Mains and Tie are closed. Note that the fault has to be on a circuit breaker that is not being switched.

Each case requires study to verify and evaluate the above assumptions. Look carefully at any automatic transfer schemes that could try to force a transfer during a fault in response to the fault events. Evaluate just how much "over" the feeder breaker rating the calculated current may be. In most cases that I have seen momentary paralleling applied, the worst case fault level with all sources tied together was 15% or less above the feeder breaker's rating. This worst case calculated fault seldom occurs and for it to occur on a feeder during unrelated switching is a low probability.

That's our reasoning behind some momentary paralleling systems. Identify the risks to the owner and let them decide.

I would never use momentary paralleling on molded case circuit breakers or switchboard construction or if the combined calculated fault level was reaching 120% of the breaker rating, or if the gear layout had feeder breakers in the same vertical section as a main or tie. Again, evaluate very carefully, or just don’t do it.

 

[rbulsara]

Recommended standard: Red Book_ IEEE Std 142-1991
Read Chapter 1 in there carefully.

Alex:

You say faults do occur, that is exactly the reason not to ignore paralleling time, even if it is momentary.

There is no standard that accepts ignoring duration of paralleling. In fact all fault calculation methods takes in to account the paralleling, regardless of its duration.

There are no standards as to what risk you should accept. So called accepted practices keep changing with time. As the new technology become available to minimize risk they become more accepted.

In this particular case availability of technology is not the hindrance, just willingness to spend money and that is no job of a standard to specify.

 

[Alex68]

rcwilson
very good answer! You describe the procedure better than me. I would add that the rating of the feeder circuit breaker is exceeded only in case of 3ph faults.


rbulsara
standards are made by men like you and me. Men who decide what is acceptable or not.
Respecting standards reduces risk but don't remove it.
That's why engineers are always needed.
Doing things as described by standards (if they are up to date!) is the best known way to do them but it isn't the perfect way.
In my country, saying "I respected the standards" is not sufficient to defend yourself in a court, but you must demonstrate that you did all the possible to remove the danger.

During the refurbishment of old power plants, I found the "short parallel" applied and I had to maintain it. The plant owner was conscious of the risk: a 3ph fault (extremely rare) in a feeder was cleared only by the main and tie breakers during the 100ms of the parallel.

 

[rbulsara]

alex:

I have no dispute with your last paragraph in your last post. I said that earlier.

 

[ScottyUK]

Hi alex68,

Where a situation has arisen where fault levels have risen due to network reinforcement beyond what the original designer could have anticipated then the scenario you describe is probably acceptable under carefully controlled conditions. It's entirely different to designing equipment knowing from the outset that it will be operated beyond its rating - my earlier point was that a designer should not knowingly design in such a risk; if a design proves to be inadequate many years later due to changes elsewhere in the system then it is unfair to blame the designer unless the change was reasonably foreseeable.


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[waross]

I worked on a system where the available short circuit current avaiable at the main 13,000 volt switchgear exceeded the rating of all the unit subs in the plant. Someone did some heads up calculations and specified that no unit sub may be fed with less than 100 feet of cable. In the few unit subs that were within 100 feet, the excess cable was doubled in the cable tray.
On another project iron core reactors were applied to the generators to limit the total available current to the switchboard rating when all generators were on-line.
In another plant, air core reactors were used to limit the available current.
There is almost always a way to keep within the ratings. Sometimes it is as simple as calculating the added impedance of the cabling. I have bailled myself out a couple of times by including the impedance of the supply conductors in the short circuit calculations.
Sometimes economics preclude an accepted solution and all you can do is warn the owner and cover your own ASSets.
respectfully