Switch

[Engcan]

Hi
The discussion started between me and my sr. engineer on the selection of a transfer switch. It was for a emergency genertor and the some part of load is lighting and computer and rest is HVAC load. THe gen is around 1 MW diesel gen. He advsied that we should use automatic delayed transition and I advised him we can use open transition but closed tranistion is better. As in case of closed transition as for some part generator comes in paraller with utility, one has to bring this thing in to notice of local utlity so normally people advoid this. In case of open transition switch the time interval is around 45 m sec.and normally except computer no other device neither HVAC and nor lighting will notice this interuption.
But anyways he did not agree and stressed that we should go for a delayed transition. When I asked him that why he wanted to go for it he gave me an example of an instance which I could not quite understand.
The example was like this:
He said in one case , a utility breaker was triping as the power used to come on and it was happeneing as there were 4 transformers in paraller around 750 KVA total and the current was 103 kA. I was surprised he said as the transformers had no time to demagnetiz so they had more inrush current whic became 103 kA
Do you think this can be a possibility. Please advise and which loads are effected by open transition switches.
Thanks

 

[powerwagon75]

Won't vouch for your Sr. engineer's rationale one way or another, but have seen instances where bad things can happen with a highly inductive load and open-transition switches; ie, an approach radar motor sheared it's gear drive, chillers having nuisance circuit breaker trips. These were presumably caused by open transition, without any regard for phase angle (having an "in-phase" monitor installed in the ATS), and are a rarity. For the instance you and he were discussing, with a 1-Meg genset, how much load is on the circuit? How critical is the load to the customer? Can the customer accept open-transition? If so, many people are going with programmed transition (aka delayed transition) these days. Programmed transition was designed primarily for the reasons your Sr. was eluding to-- allowing magnetic fields to collapse in inductive equipment. If the local utility isn't a total pill to deal with, closed-transition would be a good way to go, if within budget constaints/customer value. If you are looking at a large amount of inductive load hitting the generator all at once (block-loaded), you could look at the possibility of a soft-load/unload closed-transition switch. However, that brings other things into play, such as bias control for the genset freq/volts that could make it cost prohibitive. I've seen several closed-transition systems pick up large block loads just fine; all depends on the characteristics/performance response of the genset.

 

[davidbeach]

Motor loads don't like interruptions in the range from about 2 cycles to about 2 seconds, different systems have different times. Out of phase transfer of any time less than about 2 seconds will cause severe trouble. Closed transition is great if acceptable to the utility, otherwise in phase monitor for very fast, open, in phase transfer or delayed transfer. You can often get away with less than 2 seconds if all of the motor controls drop out on loss of power, but then the control systems have to have a restart time of 2 seconds or more (larger motors with high inertia loads will need longer, sometimes much longer). Drives that can catch spinning motors are great for these types of applications, no need to wait for the motor voltage to collapse before reenergizing.

Typically a transfer switch that offers closed transition has the two systems tied together for only a few cycles (less than 100 ms) and non of the load sees anything, but you have to have permission of the utility.

In Phase Monitor provides a very fast transfer between the two source while nearly in phase. Computers will not see the outage, and motors won't object because the outage will be too short for phase shifting and the two sources are close enough together that they won't cause objectionable phase shift to the motors. BUT DON'T USE if you are feeding a UPS system, may also apply to VFDs. The phasing will be close, but not close enough for the front end filters and they may draw enough of a spike to trip the normal source breaker on return to normal.

Normal open transition transfer switch is BAD for motor loads. Transfer can happen at any phase relationship, but fast enough that starters may not drop out. Will make a mess of motors and loads. Computers may not ride through. Great for all other loads.

Delayed transfer is the standard solution for motor loads.

 

[swgrmfg]

Mr. Engcan:

On the micro-level...,

Spinning motors at full speed have built-up emf's that will not go away (decay) until a couple of hundred power cycles. Re-application of power within 3 seconds of removal will cause inrushes up to 20 times FLA, guaranteed.

Also, energized transformers have steel (iron plus carbon, silicone, and other stuff)laminations. These laminations are magnetized during operation. Magnetic di-poles in the laminations follow the sine wave (with some reluctance). When power is first interrupted, the magnetic di-poles are aligned in a specific direction. Normally, after a few seconds (sometimes much longer), the di-poles go atsray in random directions. If however, power is re-applied quickly, you will be forcing all the di-poles to re-align in a different direction, worst case, 180 degrees. Hence the higher inrush (from 10-15 times to double that).

I would try to talk the utility into an 'un-intentional' 3-5 cycle overlap, time after which the generator breaker would open. Using a good, expensive synchronizer with voltage matching and tight angles and a back up synch check sometimes attracts their fancy. Of course, blinking the lights doesn't.

Good Luck.

 

[waross]

I'm in agreeament with all of the above.
Closed transition, or open transition with 2 to 5 seconds transfer time. The occupant reaction to open transition may be based more on the voltage loss rather than the time of the loss. That is, you will get just as many comments/complaints for 2 seconds as you will for 5 seconds, so you may as well go for the safety of the few extra seconds.
General Interest
That said, A few weeks ago I was checking the brochure for the transfer switch on a 40 KW standby generator install that I will be doing.
This mechanism has been out for some years, but I had not previously seen any information.
This transfer switch has the fastest mechanism that I have seen. The control incorporates a sync. monitor. It waits until the incoming power is in phase with the generator power and then does an extremely fast open-transition transfer. There is no time spec. but the brochure states that the operation is "Beneficial to motors".
GE ZTX Series 40-400 Amps. The identical fast transfer mechanism is also used by other transfer switch makers.
The rational may be that if an open transition is fast enough, it can be completed before the detrimental conditions develop.

 

[Engcan]

Hi,
Thanks for your comments, suppose some day, I want to use this generator in parallel with utility. Then here what I think I would do.
At present in case of emergency generators, neutrals are not grounded at generator location, they are connected to
utility ground by connecting to transfer switch, as electrical codes dont allow the ground at two locations.
But what you guys would advise, in case of a small generator, may be 500 kw, if I am using it in parallel with utility. and it is at 600 V no transformer, it is connected to the switchgear which is also being fed from the local .utility. I know utility has to give permission and they
need an interconnection protection but my concern is regard to gounding. Should I ground this generator seperately or also connect this to utility neutral like emergency generators. I read in IEEE grounding book about big generators used by insuatries, they have groundings.
As I will be coming across something like this shortly,
so I am looking for a suggestion. DO you think this violates the code, of this generator's neutral should be separately grounded. But if we ground it separately then what advantage we will have in terms of ground fault protection and if we connect this neutral to main switchgear ground and eventually which is connected to utility neutral, what advantage we have. A manufacturer advised me that in
United States, if we have some thing like a generator operating in parallel
then we will connect the neutral to the ground stud of generator FRAME and then run a properly sized conductor to
main switchgear neutra. But what raised an alarm to me, we can still supply aingle phase loads, but what about ground fault protection. Actually code, does not wants us to have a ground fault protection, then I can avoid this but anyways, I am getting a little bit offtracked of my question.
So here is the question, should I ground this generator separately or connect it to utility entrance neutral and the doubts about ground fault.

Thanks a lot for earlier responses.

 

[waross]

There are two grounding concerns;
1> System grounding,- This is the ground connection on the neutral. I prefer the one connection at the panel as allowed by the codes. It avoids switching the neutral with a 4 pole transfer switch and possible open neutral damage to single phase equipment. This may also make ground fault detection less complicated because you do not have parallel ground paths.
2> Equipment grounding,- This is the ground connection to the big stud, but not to the generator neutral.

For co-gen, it may be prudent to check with the utility as to their requirments which will be the final answer to that question.
If you do in the future, go Co-Gen, you will most likely have quite a bit of protective equipment to install, and at that time you would also make any changes to your grounding that the utility may require.

 

[mc5w]

Cahier Techniques paper ect181.pdf covers relaying to protect 2 4-wire wye power sources that are paralleled together. Just put ect181.pdf into a search engine and several download addresses should come up. National Electrical Code allows the bond between neutral and equipment ground to be relocated to the transfer switch with separate neutral and equipment grounding conductors running to both the supply transformer and generator. This would mean that the electrical service would need to be rewired with 5 wires instead of 4 and each service would need a dedicated transformer. This is so that ground fault protection can be done using a current transformer in the single bond between neutral and ground, what is known as source return ground fault protection. This is the simplest method.

If the supply system has multiple grounds for the neutrals, typical for U.S. electrical utilites, you can still do directional bround fault protection using 3 current transformer and 3 voltage transformers connected to give a residual current and residual voltage. This produces the net ground fault current AND the neutral displacement voltage so that the directional relay can figure out on which side the ground fault is.

If I were building an electrical service with cogeneration I would make the system resistance grounded rather than solidly grounded for several reasons:

1. Ground faults are limited to a low and usually nonincendive value. This actually increases reliability. Having 2 sources of zero sequence current on a resistance grounded system is also not quite so pernicious as on a solidly grounded wye system.

2 Paralleling 2 wye power sources with the neutral included can have circulating current in the neutral because wye connected generators tend to have relatively low zero sequence impedance.

3. Loads that do need the neutral such a lighting can be supplied with a transformer or must be supplied with a transformer anyways.

4. Resistance grounded systems are just as resistant to excessive static electricity buildup as solidly grounded system. The rate of motor failure because of silent and invisible lightning on ungrounded systems is so much higher that I would never ever consider using an ungrounded system that is supplied by a utility or otherwise not confined to a single room.

Metering of a resistance grounded wye transformer secondary should be done with a form 9s meter with a wye-wye voltage transformer interposed so that the meter neutral can be solidly grounded. The metering should be based on the voltage source so that the power drawn by the grounding resistor during a ground fault will register on the meter.