Unusual EM Generator problem 1

[kyphil]

As contractors, we have installed an emergency generator at a prison facility here in Lexington that is not functioning correctly. The configuration is a wye 12,47Kv primary feed to a 15Kv xfer switch. The xfer switch has two 15Kv C-H breakers in it with Lakefront electronics logic and control. The line side of the xfer sw. connects to the incoming primary overhead circuit and the emergency side connects to a 12.47Kv pad mount xfmr which steps down to 480v 3p3w. A 480 volt 600Kw genset feeds the 480v side of the pad mound and steps it up to 12.47Kv for emergency feed. The load side of the xfer sw. feeds the primary circuit if the utility is out. The grounding is good, the genset neutral is bonded to ground and all connections are good. We have installed surge suppressors on the pad mount primary and secondary sides. The transfer switch will work fine when the utility goes out-genset starts, pad mount is energized and xfer occurs. However, when the xfer sw. goes to retransfer to utility, it doesn't work. The xfer sw. supplier has replaced most of the control boards in the switch, but it still fails to retransfer. The problem occurs rarely when the manual retransfer is done within 5 seconds during testing, but if we wait 1-4 minutes, it fails to retransfer 2 out of 3 times. They are blaming the problem on a back EMF surge from the padmount windings and core saturation, but I don't think that is the problem. To me, this is rather unconventional, handling xfer at 15Kv-I have always preferred to do it at the 480v level. Has anyone run into a similar problem? The engineer of record doesn't have any idea what could be wrong.

Any advice or help with similar successful designs or problem installations would be most appreciated.

Thanks / Phil Pulliam / Lexington, KY

 

[davidbeach]

What do you mean by it fails to transfer? What fails? What are the settings? What are the voltages and currents? No where near enough information there to even begin to do more than just guess.

 

[waross]

Hello kyphil
You should have several discrete control systems.
1> A circuit which monotors the incoming voltage. This should only be connected to the normal line connections and nothing else. If you have abnormal voltages on the line when the load is removed this may malfunction. The reclose voltage may be set too high.
Testing;
This circuit must give a permissive output when suitable line voltage is present. If not, check the actual line voltages for abnormalities, and check the voltae settings.
2> You should have a similar circuit to detect the voltage from the standby set. From your description this curcuit apears to be working.
3> You shopuld have a control logic circuit that accepts inputs from the above mentioned circuits and includes the timing functions. This circuit controls the opening and closing of the breakers and the starting and stopping of the generator. It may also provide "cool-down" timing for the generator.
There is often a time delay associated with returning mains power. When the mains monitoring circuit reports that mains power is available the retransfer timer will wait for a predetermined time to be sure the mains power is stable. If the returning mains power is fluctuating, this timer may be reset by the fluctuations. This timer may be failing.
Some transfer switches have a fast retransfer feature that will bypass the retransfer timer and transfer immediately if the power outage is of very short duration.
Testing;
1> Check the setting of the retransfer timer. Some switches may be set for several minutes delay. You may not be waiting long enough.
2> Check the output of the mains power monitor. If you have an output with mains power present then concentrate on the control logic circuits.
If you don't have an output from the mains monitoring circuit when the mains power is prsent, check your voltage settings.
If your transfer switch is switching the neutral be sure that the connections of the mains monitor including the neutral are connected properly including the neutral connection to the mains neutral. Similarly, the standby monitor must have all connections including the neutral connected to the generator side.
If your transfer switch has a solid, unswitched neutral, then the voltage monitor neutrals may be connected to a common point.
What make is the transfer switch?
respectfully.

 

[ScottyUK]

Is this an open or closed transfer switch? If the latter, is the generator synchronising correctly? Sometimes the governor on a large lightly-loaded set will struggle to maintain fine enough control to allow the two supplies to maintain synch, for example.

Does the ATS attempt the transfer and abort back to generator amid the rapid opening and closing of breakers?

As David says, more info would help.

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

I don't suffer from insanity. I enjoy it...

 

[dpc]

If this is using some type of PLC for its logic, there may be a bug in the program that is blocking re-transfer. A timer race or some other logic error maybe. You need to know what signals and inputs the control is looking at to determine when to re-retransfer. When you say it "fails to re-transfer" - what does this mean? Does it try to re-transfer then trip out, or does it just sit there and refuse to try.

Do you have any large motors running on this generator when you re-transfer?

 

[kyphil]

kyphil to the group:

I have some more information based on your good posts that might help shed some light on the situation.

When the utility failure is simulated, the transfer switch signals the genset to come on and energize the pad mount to generate 480 volts wye that in turn generates 12,470 volts to feed the emergency side of this twin 15Kv breaker transfer switch. The load on the primary circuit is approximately 50 amps, 3 phase wye. The neutral of the genset is not used for 277 volts and is bonded to ground at the genset. The ground grid is good and bonded throughout the system. teh problem comes wehn the utility power is restored. The transfer switch then senses restoration of utility. I set the delay for 3 minutes to make sure utility will stay on, then start the retransfer to utility source by opening the emergency side breaker. I set a delay in this cycle of 5 seconds to make sure all back EMF has died down and then the transfer switch closes to the utility source. Two out of three times out of 5 the transfer switch will not do it automatically. The control board on the transfer switch freezes up, along with the LCD display and must be manually reset to restore automatic operation. Another bit of information-the transfer switch receives it's power (120VAC) from a 12,470v to 120 v 3 Kva step-down transformer which supplies power to the breaker recharging mechanisms, a cabinet ventilation fan and an anti-condensate heater. There is no interconnection from this circuit to the control boards in any way. The controls derive their power from the 24 VDC genset battery and go through a power supply board to provide 9vdc for the logic circuits. This voltage only varies by a few tenths of a volt up or down during transfer switch operation. There is also an alternator on the engine which is in parallel with the batteries to charge them and also supplies power during genset operation to the xfer switch controls, since it is in parallel with the battery to charge it. The battery is also connected to a trickle charger which seems to be connected and operating fine.

To answer some specifics from other's posts:
1. There are no significant motors on the primary circuit.
2. This is an open transition switch, with the 5 second delay on retransfer to utility as mentioned earlier. It takes about 6-9 seconds to transfer to generator, waiting for frequency and voltage to come up.
3. The transfer switch does not try to go back to emergency. It does physically transfer reliably-the problem is in the failure of the control boards that have to be manually reset.
4. The transfer switch is not switching a neutral. It is a 3 phase system, with grounded neutral at each pole, made wye by virtue of the way the local utility connects their step-down xfmrs at the substation.
5. I have used a Hioki 3196 power analyzer and noted nothing unusual about the 480 volt genset output other than the usual current waveform distortion while the generator takes up the load. No voltage spikes of any significance.
6. The transfer switch breakers are made by Cutler-Hammer and the control side is made by Lakeshore electric, out of Cleveland.

 

[davidbeach]

Look at the control voltage, and any transients there.

 

[waross]

"3. The transfer switch does not try to go back to emergency. It does physically transfer reliably-the problem is in the failure of the control boards that have to be manually reset."
I think that this is the place to concentrate your efforts.
1> If a manually applied signal applied on the close terminals will reliably close the breaker then we look at the control circuit. If not, we look at the breaker. Note, I mean the same terminals, not an alternate set of terminals, for remote operation or over-ride operation.
2> Are we getting a close signal from the control circuit? If this is a pulse, is it of sufficient voltage and duration? There may be a capacitor going bad or something draining the charge. If the output is a pulse, it may not be of sufficient energy for the switch it is being used with.
respectfully

 

[kyphil]

PROBLEM SOLVED! I thought I would let the members who have been helpful on this thread know what the solution was. The controls on the 15Kv tranfer switch were goofing up and they were getting all their supply from the 24VDC batteries. What was happening was that the two #12 in pvc conduit feeding this switch were routed in a common trench from the generator under a 500kva padmount xfmr. When the field from the xfmr fluctuated, it induced noise on the DC line, creating havoc in their controls. All they had to do was install a lowpass filter on that dc line, twist the wiring tightly and install ferrite chokes on the last point where the dc lines fed the main control board. Problem solved and lesson learned! Just wanted to let you all know and thank you for your help.

Kyphil

 

[waross]

Hello kyphil
Thanks for the feedback.
The next time someone has a problem with a transfer switch, I am sure this thread will be cited.
Respectfully

 

[itsmoked]

I wounder id steel conduit (in the first place) would have prevented the magnetic coupling.

Thanks kyphil.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[kyphil]

Kyphil to the list & itsmoked: Steel conduit might have alleviated the problem, but it is difficult to shield a strong magnetic field. Having designed several MRI facilities, I am familiar with this. I think next time, I would use steel conduit, twisted shielded DC wiring and certainly kept it as far from the transformer as possible.

Thanks / Phil

 

[itsmoked]

I would think that steel conduit would line up the flux with the wires length wise and hence reduce the field-thru-the-loop flux between the wires, a lot. But you also then have steel conduit in probably wet concrete, a lousy situation too. I am also amazed that much flux leaks out of the transformers. Interesting...

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com