Bus Swinging Capability

[HELENG001]

Hello

I am working in project where there is a 5kv Motor that ran from bus "A". The motor has to be transferred to bus "B" without interuption in case of failure in bus "A" to bus "B". The motor is protected by 269 Multilin Relay.

I am thinking of using synck ceck relays to accomplish the swinging between the two buses. However; my concern is weather the 25 relays are fast enough for such application.

I would appreciate your help.

Thank you,
HELENG001

 

[jbartos]

Suggestion: For 60Hz
1/60Hz=0.0166Second
If the 25 relay can synchronize within ±10°, which is 10°/360°=0.02777, then
the time will be 0.0166Second x 0.02777=0.00046277Second error.
Fast static transfer switches can transfer the load within a quarter of the cycle, i.e. 0.25 x 0.01666=0.004165Second, which is considered satisfactory for majority of loads including motors.

 

[dpc]

The synch check relay will not be the problem. It does not initiate the transfer, but simply verifies that the two buses are within the specified phase angle and voltage magnitude range to facilitate a safe transfer.

The problem will be how you sense a "failure" of either bus. If you sense undervoltage, then a fast transfer should not be a problem, provided you do not have any time delay in the undervoltage relay. But the synchronous motor will have a tendency to keep the bus voltage up for a while, even after you lose incoming power.

Also, how will you transfer the motor between the two buses?

You may want to consider intiating the transfer based on the opening of the main breaker on either bus, or based on inputs from your bus protection relays, whatever they are.

 

[redtrumpet]

See the papers Beckwith Electric has on their website, for example Application Note #19.

http://www.beckwithelectric.com/infoctr/iichome.htm

 

[jbartos]

Suggestion to the previous posting: The application note 19 focuses on electromagnetic switching. The fast electronic static switches can perform the load transfer within less than a quarter cycle.
Visit

http://www.e-d-i.com/pdf/b935.pdf

for sense and transfer time less than 1/4 cycle.

 

[bigamp]

I concur with redtrumpet re Beckwith, that is very good advice. We had a look at using one of their fast transfer systems a few years ago. They have realy good equipment. In the end we considered that their offer was too expensive and we managed to do without.

I also concur with dpc that rapidly detecting loss of bus voltage is the main issue. Perhaps an application of df/dt (rate of change of frequency) or voltage vector shift relays would work (you would need to check with the relay manufacturers if it was an OK application). These relays are normally used to detect loss of grid and are often used to rapidly uncouple from the grid installations that operate generators in parallel with the grid. A number of manufacturers, such as Alstom can provide such relays.

As dpc says, when you lose the bus that is powering the motor, you do not lose motor voltage immediately. The motor voltage and frequency decays as the motor runs down. It is desirable to re-apply voltage to the running down motor from the healthy supply at a point when the voltage to be re-applied is in momentary "synchronism" with the decaying motor voltage. This minimises the shock to the motor and also to the power supply system. I believe that suitable devices to do this sort of thing (and that can also allow for closing time of MV switchgear) do exist but I am not aware of any specifically, perhaps somebody else is.

The static switch device referenced has an excellent specification but I do not think it is applicable to your situation.

Regards

 

[jbartos]

Small encore about the frequency sleuth, (df/dt): Visit

http://ecpe.ee.iastate.edu/powerworkshop/papers/Chow.ppt

for more info

 

[HELENG001]

Hello;

Further to my original posting with regard to the 5kV motor transfer between bus A and bus B. I need to know the following:

1. For a 700HP motor, what is the time that is considerd to be safe for fast transfer between the buses to avoid damage to the motor.

2. The subject motor is feed via 400A Vacuum contactors, and the time for closing the contactor is 11-cycles and the opening time is 8 cycles. Are 11 cycles short enough for the residual voltage of the motor to decay to the safe value so imposing full bus voltage does not damage the motor.

Please comment.

Thank you,

HELENG001

 

[electricpete]

There is a lot of discussion in the following thread and another thread which is in turn linked from that thread:
thread237-11584

the method of using time delay before reclose without any sync check is called "Slow Transfer Reclosing" in NEMA MG-1-1998R1 - section 20.34.1 . In order to allow residaul flux/voltage to decay sufficiently, the power should be interrupted for >1.5 times the open circuit ac time constant for the motor. That time constant is defined in Section 1.60.1 as:
"Open Circuit AC Time constant"
= (Xm+X2)/(2pi*f*r2) where Xm, X2, r2 are equivalent circuit parameters (per-phase). f power frequency.

ANSIc50.41 approaches a similar approach.

The theoretical model for above seems very simplistic. If you open circuit the primary, there will be a loop circulating through rotor and magnetizing branch. This is the L/R time constant. One time constant should get you down to about 37% voltage. 1.5 time constants should get well below the recommended 33% voltage.

There have been legitimate questions raised on this approach in the previous threads which I certainly don't have all the answers for.

Some references suggest that 10 cycles may be enough of a wait. I have data for one large 2-pole motor with an open-circuit time constant on the order of 2 seconds detailed in the linked threads.

 

[electricpete]

I guess that the main objection raised in the attached thread is that the mechanical parameters affecting coastdown are not considered. I reviewed my 10/7/01 response and I think it addresses the question by demonstrating that this is conservative (current decays according to the electrical time constant under the worst-case condition = no coastdown... any coastdown will only make it decay faster).

Probably the fact that NEMA and ANSI endorse this method should carry a lot more weight than anything that I say.

 

[bigamp]

Electricpete,

Excellent information and discussion in the above thread and also the threads you have referenced.

T'do of a synchronous generator can be measured very simply. You run the generator at rated speed and voltage, with no load and then remove the excitation. T'do is the time taken for the generator terminal voltage to decay to 1/e of (rated voltage minus residual voltage), residual voltage being the generator terminal voltage with zero excitation applied.

Is it possible to do a similar sort of thing with a cage induction motor, say remove supply to the motor bus (but leave the motor starter closed in) and time how long it takes for the bus voltage to decay to 1/e of its initial value (assumes you have bus voltage measurement)?

Regards

 

[electricpete]

thanks bigamp. It makes a lot of sense to measure that decay time constant directly, as you describe, when possible.

 

[jbartos]

Suggestion to HELENG001 (Electrical) Aug 17, 2002 marked ///\\1. For a 700HP motor, what is the time that is considerd to be safe for fast transfer between the buses to avoid damage to the motor.
///1/4 cycle static transfer switches are considered adequate fast transfering devices for motors. They are offered at 5kV level.\\
2. The subject motor is feed via 400A Vacuum contactors, and the time for closing the contactor is 11-cycles and the opening time is 8 cycles. Are 11 cycles short enough for the residual voltage of the motor to decay to the safe value so imposing full bus voltage does not damage the motor.
///The motor-load dynamics shall not be overlooked. Especially, the mechanical time constant Tm=J x nus/K
where
K=ms x |Ee|**2 / (nus x Rr')
This is where the mechanics, physical property and electrical property meet. Therefore, the electrical time constant is just at half-of-the problem. One is actually dealing with an electromechanical system and its property. If J is large then the speed will be decaying slowly thus keeping the back emf relatively high for longer period of time relatively to the external impedance as the load.\\

 

[electricpete]

Helen- you said: "Are 11 cycles short enough for the residual voltage of the motor to decay..."

On the surface it appears that you may be mixing the concepts of fast and slow transfer. If that was a simple typo, please forgive me. Otherwise, I'd like to review these two methods:

NEMA MG-1-1998R1, for large motors - section 20.34.1

"Slow Transfer or Reclosing" - power interrupted for >1.5 times the open circuit ac time constant. This is the NEMA-recommended approach since it allows residual flux/voltage to decay to acceptable levels before reclose.

&quot;Fast Transfer or Reclosing&quot; - power interrupted for <0.5 times the open circuit ac time constant. Voltage phase should not have drifted that far apart during this short time.

Returning again to the subject of whether the mechanical coastdown needs to be studied in applying these approaches:

MG-1 section 20.34.2 states &quot;...it is recommended that the electromechanical interactions of the motor, the driven equipment, and the power system be studied for any system where fast transfer or reclosure is used. There is no corresponding recommendation for study of mechanical characteristics in the case of slow reclosing.

For me the physical basis for these approaches (studying the coastdown due to mechanical loading for fast transfer, but not for slow transfer) seems clear:

- Any mechanical loading present will tend to make the phase shift and to make the frequency drop faster than is predicted by the electrical time constant.

- Not accounting for this mechanical coastdown would be nonconservative in the case of fast transfer where we want to reclose before the machine has had a chance to coast to an out-of-phase condition.

- Not accounting for this mechanical coastdown and relying only on the electrical time-constant of the flux decay is conservative for establishing a slow-transfer criterion. The stator-generated voltage is proportional to residual flux times rotor speed. Flux decay is in accordance with electrical time constant. Speed decay provides an additional voltage reduction. Neglecting speed decay provides a conservatively high estimate of residual voltage at the time of reclose (slow transfer).

Note MG-1 doesn't address sync check or other means which are also satisfactory.