Unbalance 110 VDC Distribution Incoming on the Battery Charger Rectifier / Service Rectifier

Hello gents,
We are having unbalanced 110 VDC (+62 and -46) on all our six Battery Charger Rectifiers / Service Rectifiers. The voltage values were taken Upstream of the 110VDC Distribution Incoming feeder breakers. We also have a Ground Fault Alarm on one of the rectifiers which loads include: 12.5kW Emergency Lube Oil Motor, 1.8kW Hydraulic Ratchet Pump Motor, MkVI-1&2, Filter House Emergency Light and Enclosure Emergency Light.
We would like to clear the GND Fault Alarm and the Voltage Unbalance.
 
Hello gents,
We are having unbalanced 110 VDC (+62 and -46) on all our six Battery Charger Rectifiers / Service Rectifiers. The voltage values were taken Upstream of the 110VDC Distribution Incoming feeder breakers. We also have a Ground Fault Alarm on one of the rectifiers which loads include: 12.5kW Emergency Lube Oil Motor, 1.8kW Hydraulic Ratchet Pump Motor, MkVI-1&2, Filter House Emergency Light and Enclosure Emergency Light.
We would like to clear the GND Fault Alarm and the Voltage Unbalance.
Hello All,

When the problem started ...any work or other fact have been noted..

Clear DC ground fault require some one to be patient and follow the schematic carefully to at least locate the fault..

Start by searching what have been done any work in any area and chase the cat by read the right drawings!

Any time!
James
 
The alarm has been there for a long time, and we were waiting for a Plant Shutdown to figure out why it's there. So far we have taken down all the load one at a time resetting the SRB/1 Board. The Alarm went off once all loads were taken down including MKVI and resetting o button at SRB/1. The voltage from each leg to ground remained unbalanced (+ve 62VDC, -ve 46VDC) and the voltage readings in the meter is fluctuating, not stable.
 
The alarm has been there for a long time, and we were waiting for a Plant Shutdown to figure out why it's there. So far we have taken down all the load one at a time resetting the SRB/1 Board. The Alarm went off once all loads were taken down including MKVI and resetting o button at SRB/1. The voltage from each leg to ground remained unbalanced (+ve 62VDC, -ve 46VDC) and the voltage readings in the meter is fluctuating, not stable.
Seems that this alarm was present for a while and you did not cleared it as it shoud be and now you getting back to it...

You will have to get your drawings and chase it could be a soleinod valve or other device ..I cannot add more for the moment as ground fault clearing need one to be on site...
 
Normally, DC ground fault occurs in one feeder (may be in a dc motor/instrument). Finding it out is tricky.
Two ways you may adopt in a running plant,
1. DC earth fault locator instrument can be used which can be engaged online (in a running plant).
2. You can measure DC voltage (in + and - line) of each outgoing feeder in DCDB, where there is a chance that you can get a feeder where + and - voltage is more than other( a substantive difference), it indicates earth fault in that particular feeder downstream circuit.
+62/-46V may be more unbalanced (say +80/-18) in that particular feeder, where the DC fault is lying.

As far you will reach nearer the fault location voltage unbalance will increase.
If you are lucky you may find out earth fault location in this way during running plant.
But always be extra careful and do it after getting permission from competent higher authority,
Best of luck
 
mickyeng...
I'm aware of only one company that uses an unbalanced +DC and -DC Bus System. The unbalance improves the ability to diagnose unintentional earthing/grounding of DC Powered Control Circuit wiring and devices!
The company is SEL or Schweitzer Engineering Labs!
Regards, Phil Corso
 
Hello gents,
We are having unbalanced 110 VDC (+62 and -46) on all our six Battery Charger Rectifiers / Service Rectifiers. The voltage values were taken Upstream of the 110VDC Distribution Incoming feeder breakers. We also have a Ground Fault Alarm on one of the rectifiers which loads include: 12.5kW Emergency Lube Oil Motor, 1.8kW Hydraulic Ratchet Pump Motor, MkVI-1&2, Filter House Emergency Light and Enclosure Emergency Light.
We would like to clear the GND Fault Alarm and the Voltage Unbalance.
There has to be a ground occuring on the negative side as it is lowered by 9Vdc (55vdc-9vdc=46vdc). Similarly you see the positive side with respect to earth is equally floated up to 62Vdc because 9vdc(gnd) plus 55vdc(+ve side) equals to 62vdc. Check the negative bus side and the cables.
Also if possible try to take the system on the battery bank and see if gnd fault Alarm is still present or not. Doing that you can be sure that the problem is in the battery charger itself and not in the field side.
 
I'm fairly confident based on the information provided in the original post the gas turbines are GE-design Frame units--but that's just a SWAG (Scientific Wild-Arsed Guess). Since the packager of the units was not named we can't really know who makes the battery charger(s). BUT, if the turbine control systems are GE Mark* units AND the battery charger(s) has(have) a ground detection card that is enabled and working AND the ground reference jumper in the Mark* panels is IN then finding grounds can be VERY difficult. VERY difficult, indeed.

I have seen corroded battery terminals (NO maintenance or housekeeping) cause this kind of problem. I have also seen a ground in Emergency L.O. Pump motor field circuit cause this kind of problem. I have seen bad brushes in the Hydraulic Ratchet pump motor cause this kind of problem. AND, I have seen where the GT battery(s) was(were) used for OTHER circuits (such as in the high voltage switchyard and transformer controls and protection) and grounds in that circuit cause problems. And, finally, battery charger output capacitors can also cause similar issues; they require maintenance and replacement from time to time.

But, finding "a" battery ground which has been around for a LONG time can be really problematic--because there may actually be more than one ground.

Good luck! You'll need it, it sounds like.
 
Most ALL ground detection circuits work similarly, and when the Mark* system is "enabled" (by the ground reference jumper being IN (INstalled)) AND there's a working ground detection circuit/card in the battery charger, then the two ground detection systems are in parallel with each other--and that causes problems with BOTH of them. Generally, it means they BOTH indicate grounds at incorrect voltages. And, most sites are loathe to disconnect the charger ground circuit(s) (read below).

MANY people MISTAKENLY believe the Mark* ground detection system ONLY works on devices connected to the Mark*--and that's just plain FALSE. There's no inexpensive, easy way to isolate the Mark* I/O from other devices connected to the battery (such as generator protection panel devices; emergency compartment lighting; DC pump motors; some fire detection/protection schemes; etc.). The ground detection system on the Mark* looks at EVERY device and circuit connected to the same battery as is powering the Mark*.

What MOST people don't also realize is that ONLY devices and circuits powered DIRECTLY by the battery (connected to the Mark* or not) can be the source of a ground. So, thermocouples cannot be the source of a battery ground; RTDs cannot be the source of a battery ground; speed pick-ups cannot; vibration sensors cannot; 4-20 mA transmitters cannot. DC solenoids can be the source of a battery ground; a temperature switch, or a limit switch, or a pressure switch, or a level switch, or a vibration switch--if powered by a nominal 125 VDC battery source through a Mark* turbine control panel--can be the source of a battery ground. But, low-level signals cannot cause a battery ground (unless the power supply itself is grounded--the ones that convert nominal 125 (or 110) VDC to low-level voltage and currents.

I have tried to write about how to troubleshoot grounds on a Mark* turbine control panel before on Control.com. It's NOT easy to describe. And, without circuit drawings (sometimes called "loop sheets") it's very difficult for people to visualize how things are and can be interconnected, which makes troubleshooting grounds very difficult for most people.

Without ANY load on the battery/charger, it's likely that the voltage split will not be perfectly balanced. Again, ESPECIALLY if the ground detection circuitry on the Mark* and the battery charger are BOTH enabled. If the ground detection system in the Mark* is enabled and annunciating grounds, the ENTIRE battery/charger and all the devices powered by the battery/charger are protected. Most people just DO NOT BELIEVE that--because they don't understand basic circuitry, AND there are generally no circuit drawings for the Mark* or the devices connected to the DC distribution panel. And, because the ground detection card/circuit annunciator is connected to the plant DCS--and because most people believe (falsely!) that the Mark* can distinguish between devices connected to it and those that aren't--they ARE NEVER going to disconnect the charger ground detection circuit.

And those people are doomed to a life of ground alarms (insignificant--and difficult to find) and most ALL of them will blame the Mark* (because it IS, after all, the root of all evil at a power plant).

Without knowing a LOT more about the configuration of the plant and who packaged the turbines and auxiliaries and the battery charger(s) there's really not much more we can add.

Again, best of luck! It would be really nice if you could tell us, mickyeng, where you finally isolate the ground(s) to. A simple, "It was a device connected to the Mark*," or, "It was not a device connected to the Mark*," would suffice (for me).
 
mickyeng,

If you're calling a large UPS (Uninterruptible Power Supply, sometimes called an "inverter) a "Service Rectifier" which is ALSO supplying the gas turbines (and turbine control systems) in addition to a 110 VDC battery and charger then the "Service Rectifier" is probably supplying AC to an AC-to-DC converter for the turbine control systems. (GE calls the AC-to-DC converter a "DACA"--their choice of terminology.)

AND, you wrote that when ALL the loads were disconnected from the unit(s) with the battery ground the voltage "split" was still +62 VDC and -46 VDC--and that included disconnecting the Mark VI that would mean that the ground reference circuit of the Mark VI was also disconnected. The GE Mark* system wants a "floating" DC power supply--meaning the DC power supply should not be grounded on either leg or through a ground detection reference circuit.

The snippet below (from GEH-6421, Vol. II, 'Power Distribution Modules') has the ground detection reference jumper highlighted on the TCPD card in the <PD> Power Distribution Module in the Mark VI.

When the 125 VDC (110 VDC, apparently, at your site) supply is disconnected (probably by opening the DC breaker that supplies the DC power from the battery and charger to the "Power Filter Board" (GE often calls this the "Power Conditioning Module") the entire Mark VI is disconnected from the DC power supply--including the ground reference circuit highlighted.

Hold a 9 VDC battery in your hand without touching either of the terminals. That is a floating DC power source. Use a voltmeter in DC mode and connect the negative lead of the meter to a proper ground. Now touch the positive lead of the meter to the positive lead of the battery in your hand. You will see the meter display indicate a voltage somewhere around 9 VDC initially and then it will start decreasing, pretty quickly, probably, to around 0 VDC. It might indicate a few hundredths or a couple of tenths of a volt (depending on how humid it is and how humid your hand is while holding the battery)--but for all intents and purposes there is zero potential between the positive terminal of the 9 VDC battery and ground.

Now, touch and hold the positive lead of the meter to the negative terminal of the 9 VDC battery. You will see an initial voltage of around -9 VDC or so, and it will start to decrease, again--pretty quickly, probably--to near 0 VDC. It might go to a couple of hundredths or tenths of a volt again (depending on conditions) but for all intents and purposes there is zero potential between the negative terminal of the 9 VDC battery and ground.

It's also entirely possible that the voltmeter might indicate 2.4 VDC when the display stops decreasing while the positive meter lead is in contact with the positive terminal of the battery. Or, the meter display might indicate -3.5 VDC. A floating power source--with no reference to ground--will not be evenly split, nor will it likely ever be zero, either.

1638131113817.png

Now, those two rectangles to the left of the BJS jumper are actually resistors--of equal value. SO, when there is DC voltage at the input to the <PD> that voltage is EQUALLY divided (that is called a 'voltage divider'--two equal value resistors that is center-tapped, and in this case to ground).

BUT without any ground/ground reference there is NO GUARANTEE the voltage from a floating power source will be equally split. In fact, it's only the resistances in the voltage divider network that are equal that make the voltages equal in magnitude as long as there are NO grounds in the circuits powered by that power supply.

The programming and configuration in the Mark* turbine control system expects the voltage divider network to have a certain resistance. IF the battery charger has a ground detection card/capability it most likely uses a voltage divider network--which means the resistors of the Mark* and the battery charger ground detection will be in parallel--which changes the resistance the Mark* AND the battery charger circuitry sees--which means the sensitivity of BOTH ground detection circuits will NOT BE AS DESIGNED. So, only one ground detection circuit can be used--NOT two (2).

Finally, it IS possible to have a large voltage converter feed a second DC power supply to a Mark* turbine control system (so maybe the "service rectifier" outputs DC and not AC--we don't know, you would have to tell us). And, if that's the case, and the "service rectifier" has its own ground detection circuitry then it can also negatively impact ALL the ground detection circuits.

That's about all I can add to this thread without a LOT more information and explanation and definition of terms.

Best of luck--and, PLEASE, write back to let us know what you find and how you resolve the issue(s).
 
Top