Mark-vie 125VDC input supply coming from battery charger suddenly having difference between both legs.

Positive (+) phase: 45VDC
Negative (-) phase: -80VDC

Immediately we have isolated Mark-Vie input 125VDC supply from charger and system is lined up on DACA.

Please guide me what problem will occur on Mark-Vie system in case of continuous connect with this difference.

 

VKGUPTA,

GE Speedtronic turbine control systems are designed to run on an ungrounded DC supply. They use a voltage divider network to <i><b>reference</i></b> the DC supply to ground in order to monitor for grounds on either leg (phase) of the DC supply.

When a ground develops on one phase or the other, the magnitude of the voltage of that phase (the one with the ground) will decrease towards 0 VDC, and the magnitude of the voltage of the other phase (the one without the ground) will increase towards full battery voltage.

So, what you see when the voltage split is -80/+45 VDC is that a ground is developing on the positive phase--because the voltage of the positive phase with respect to ground is decreasing.

This is sometimes called a "soft" ground, as opposed to a "hard" ground which would be characterized by a -125/+0 VDC voltage split. Soft grounds are most always caused by moisture (usually water; sometimes L.O.) getting into junction boxes and coming into contact with terminals which have battery voltage (125 VDC) on them, most commonly discrete (contact) inputs or 125 VDC solenoid circuits. (The majority of I/O connected to a Speedtronic turbine control panel are discrete (contact) inputs (pressure switches; temperature switches; limit switches; level switches; etc.), and there are also a large number of 125 VDC solenoids (fuel stop valve/trip solenoids; air-operated solenoids; etc.). ONLY DEVICES POWERED BY 125 VDC CAN DEVELOP GROUNDS WHICH WILL RESULT IN A VARYING VOLTAGE SPLIT AND A "125 VDC BATTERY GROUND" ALARM.

The last thing to know about the 125 VDC battery ground detection circuit of GE Speedtronic turbine control panels is that they monitor <b>every device or circuit powered by/connected to the 125 VDC battery--not just devices or circuits connected to the Speedtronic turbine control panel.</b>

So, for example, the 125 VDC battery is usually used to power the DC Emergency Lube Oil Pump Motor, a DC Hydraulic Ratchet Pump Motor (if the unit is equipped with a DC Hydraulic Ratchet Pump Motor, the Emergency DC Lighting Circuit, generator protective relays in the Generator Protection/Control Panel, and many sites also use the 125 VDC battery to power the fire detection and discharge systems. Some sites also use the 125 VDC battery to power devices associated with the step-up transformer (high-voltage switch operators; protective relays in the high-voltage yard; transformer protective relays; etc.).

I've been to sites where the 125 VDC battery is used for hazardous gas detectors around the plant, or for plant security systems, also. <b>ANY</b> device or circuit connected to or powered by the same battery used to provide power to the Speedtronic turbine control panel can develop a ground that will be detected and annunciated by the Speedtronic turbine control panel. <b>>>ANY<< DEVICE OR CIRCUIT POWERED BY THE 125 VDC BATTERY--NOT JUST DEVICES OR CIRCUITS CONNECTED TO THE SPEEDTRONIC TURBINE CONTROL PANEL--WILL AFFECT THE 125 VDC VOLTAGE SPLIT WHEN THE BATTERY IS CONNECTED TO THE SPEEDTRONIC TURBINE CONTROL PANEL.</b>

So, you've said the battery and charger was disconnected from the Mark VIe. You haven't said what happened to the voltage split when this was done. Did it return to an even split (+62.5 VDC/-62.5 VDC)? Did it remain at +45 VDC/-80 VDC? We also don't know how you isolated the battery and charger from the Mark VIe (by the breaker near the Power Distribution card(s) in the Mark VIe; or by opening the breaker at the 125 VDC Distribution Panelboard that supplies the Mark VIe), so we can't say for sure if the ground is on some device or circuit connected to the Mark VIe or some device or circuit not connected to the Mark VIe.

Now, for your last question. Far too many sites run with +125/-0 VDC splits, or +0/-125 VDC splits. And even more sites run with unequal splits, such as the one you are reporting. Speedtronic turbine control panels <i>are designed</i> to run with one or more grounds on either phase (leg) so that the turbine and generator can continue to produce power while the instrumentation and control technicians work on finding and resolving the ground(s).

The problem occurs when a ground (or multiple grounds) exist on one leg and then a ground--just one--develops on the other leg. That's the equivalent of a dead short across the 125 VDC battery, and that causes arcs and sparks and blown fuses and tripped circuit breakers and turbine trips.

The "long-term" effects of an unbalanced DC supply resulting from one or more grounds on one leg of the DC supply are that if left unresolved should a single ground develop on the other leg of the DC supply that BAD things will happen. Nothing will happen as long as a ground (or multiple grounds) exist on one leg of the DC supply--again, the Speedtronic is designed to run like this so that reliability is improved. But NO DC-powered control system can survive grounds on both legs of the DC supply.

So, the takeaways <i>from the information provided</i> are:

1) A ground is developing on some device or circuit connected to the positive phase (leg) of the DC supply--and that device is <b>NOT</b> necessarily connected to the Mark VIe. Any device or circuit that is powered by a breaker in the 125 VDC Distribution Panelboard that also supplies the Speedtronic turbine control panel that experiences a ground can cause the voltage imbalance and eventually the "125 VDC BATTERY GROUND" alarm.

2) Speedtronic turbine control systems are designed to run with one or more grounds on devices or circuits connected to either phase (leg) of the DC supply--but it CANNOT run with grounds on BOTH phases (legs) of the DC supply.

3) The ground you are reporting is a "soft" ground, likely caused by moisture getting into a junction box where it shouldn't be. The easiest way to troubleshoot grounds like this are to methodically open and inspect every junction box which is OUTSIDE of any turbine-, accessory- or load compartment enclosure--because that's where most of the moisture-related grounds occur.

4) ONLY DEVICES OR CIRCUITS DIRECTLY CONNECTED TO THE 125 VDC BATTERY (and that includes devices connected to the Speedtronic such as discrete (contact) inputs, and 125 VDC solenoid outputs) CAN CAUSE THE VOLTAGE SPLIT TO CHANGE AND RESULT, EVENTUALLY, IN A "125 VDC BATTERY GROUND" ALARM. In other words, magnetic speed pick-ups, thermocouples, LVDTs, RTDs, pressure transmitters, etc., <b>CANNOT</b> cause unbalanced voltage splits and "125 VDC BATTERY GROUND ALARMS. So, checking these devices/circuits that are not powered by 125 VDC is useless and a colossal waste of time, energy, effort, and money and can lead to a LOT of frustration.

Hope this helps!

 

VKGUPTA,

My last post may have seemed contradictory with regard to how Speedtronic turbine control systems are designed, so I'm going to try to clarify.

The 125 VDC supply (be it from a battery and charger or from a DACA) must be ungrounded. That is, one phase (leg) must NOT be grounded.

Many DC systems operate with one phase (leg) grounded (earthed); this actually makes wiring a little simpler as ground (earth) is used as a conductor. And, any grounds on the grounded leg will never be known--and they don't need to be known discovered, because one phase (leg) is grounded.

HOWEVER, when a ground develops on the ungrounded phase (leg) that presents a problem. A serious problem that results in arcs and sparks and blown fuses and tripped circuit breakers. That situation must be resolved before the equipment can be put back in service.

Speedtronic turbine control panels DO NOT use a grounded supply, and they reference the ungrounded supply using a voltage divider network which--when the supply is ungrounded--results in a relatively even voltage split when referenced to ground. (The term "voltage split" is something of a misnomer, as one has to read each phase (leg) with respect to ground to measure the difference (the "split").)

Speedtronic turbine control panels use the 125 VDC to power the main microprocessors and other components of the system, AND they (typically) use 125 VDC as the voltage source for discrete inputs and solenoid-operated devices--some of the I/O (Inputs and Outputs) connected to the turbine control panel.

Now, if a ground develops on one leg or the other of devices or circuits connected to the 125 VDC supply the Speedtronic turbine control panel doesn't really care. It just keeps on running (and the turbine keeps on running) just like nothing ever happened. If the severity of the ground is high enough, the Speedtronic turbine control system will annunciate a "125 VDC BATTERY GROUND" alarm (the magnitude that setpoint is usually around 30 VDC with respect to ground). And that's a signal for the technician(s) who are responsible for the care and maintenance of the control system to investigate and resolve the ground.

Does it have to be immediately? No. Should it be done immediately? Yes. The sooner the ground is found and resolved, the better. Because, again--a ground on the other leg WILL result in arcs and sparks and blown fuses and tripped circuit breakers and tripped turbine and an inability to re-start the turbine.

So, while the <b>supply</b> is <b>intended</b> to be ungrounded, the turbine control panel can run with one or more grounds on either leg for an indefinite period of time--until a ground occurs on the other leg, and then, well, ...

Troubleshooting grounds is not easy--but it's not impossible, either. It just takes a good understanding of the ground detection system and the knowledge that the ground alarm <b>DOES NOT</b> mean the ground is only some device or circuit connected to the Speedtronic turbine control panel--because when the Speedtronic is being powered by the 125 VDC battery and charger it is monitoring ALL of the devices and circuits being powered by the battery and charger.

The ground detection circuit is not capable of distinguishing between grounds on devices or circuits connected to the turbine control panel and grounds on devices or circuits not connected to the turbine control panel. And, the alarm only applies to devices and circuits which have 125 VDC directly applied to them--so, not to devices which get 24 VDC or 7.0 VAC RMS or 4-20 mA or 10 VDC or which provide their own power (passive magnetic speed pick-ups; thermocouples).

Again, the majority of grounds occur in junction boxes or devices which are outside of the Accessory Compartment, the Turbine Compartment, the Load Compartment, and the Generator Accessory Compartment. And, most of the problems are the result of water (rain water; condensate) getting into junction boxes or devices which were improperly installed. It is NOT good construction practice to run conduits into the top of junction boxes or devices when they are not inside a building or enclosure or compartment.

Most installation "electricians" and their supervisors don't know this and they just run the conduit in the most convenient way--which is quite often into the top of the junction box or device. So, when it rains it's common for water to leak into junction boxes or devices with conduit penetrations on the top--and even on the side.

Some conduit fittings are not suitable even for use on the side of junction boxes or devices in that they don't properly prevent water from getting into the box or device. Such fittings are usually less expensive than suitable fittings, but only the least expensive fittings are used for side- and even top conduit penetrations--which leads to a lot of problems.

And, again, grounds can occur not only on devices connected to the Speedtronic turbine control panel.

In general, one or more grounds on one phase (leg) of the 125 VDC doesn't really cause a problem for Speedtronic panels. However, multiple grounds of varying significance does quite often cause a lot of intermittent and nuisance problems, and can even cause the ground to seem to shift phases (legs) during starting, acceleration, operation, and shutdown--including cooldown. So, it's really in everyone's best interest to find and resolve any ground as soon as it becomes known. Failure to do so can result in multiple grounds which can be much more difficult to find and resolve--not impossible, but more difficult.

Now, for the next question--yes, the 125 VDC output of the DACA is ungrounded, and is used for powering Speedtronic turbine control panel I/O such as discrete (contact) inputs and solenoid-operated devices. So, grounds can develop or will remain present when DACAs are used to power the Speedtronic turbine control panel <i>if the battery/charger is completely disconnected from the Speedtronic turbine control panel.</i>

If you completely isolate the Speedtronic from the battery and charger and all of the external devices and circuits powered by the battery and charger and only connect the output of the DACA to the Speedtronic turbine control panel and the voltage split returns to normal (evenly divided) then the ground is on some device or circuit which is NOT connected to the Speedtronic turbine control panel.

Lastly, I'm not a fan of DACAs for providing power to Speedtronic turbine control panels. They are known to amplify voltage spikes and dips on the incoming power, causing the output voltage to spike higher and lower than one would otherwise expect. That's because there is little in the way of filtering on the output of the DACA. This can cause problems with Mark V and Mark VI power supplies in particular, and Mark VIe power supplies in some cases. (I'm referring to the power supplies that convert the 125 VDC to +/- 15 VDC, and 5 VDC, and 28 VDC, etc., for use by the main microprocessors and other Speedtronic components.)

Hope this clarification helps!

 

CSA,

Thank you for your detailed reply. I am currently trying to understand this same circuit.

Here is my question. How can it be "un-grounded" and still maintain a perfect +/- 62.5 VDC with respect to ground? Won't the voltage divider have to be grounded to earth in the middle to maintain that reference?

I've seen a similar circuit with light bulbs used as a voltage divider and fault indicator. Two equal light bulbs bridge the circuit and are grounded in the middle. If the + side goes to ground it will drive the + side to 0VDC (with respect to ground) and short circuit the first light bulb making it go out or dim, and put the full voltage across the second light bulb. But I think the engineers called it "centrally grounded" instead of floating since the voltage was not truly allowed to float to far above and below ground potential.

Would the voltage divider used to supply the GE Mark VIe be centrally grounded like this or would changes in the ground potential (i.e. lightening strikes, faults, etc) cause fluctuations in the measurement between earth ground and the +/- terminals?

Thanks,
Frank

 

Frank,

Quote from my first response to this thread:

"...They use a voltage divider network to reference the DC supply to ground in order to monitor for grounds on either leg (phase) of the DC supply."

>Won't the voltage divider have to be grounded to earth in
>the middle to maintain that reference?<pre> ----- - - - - ---------------- - - - -
| |
| ---
| | |
----- | |
--- 125 VDC --- Voltage
----- Battery | Divider
--- & |------ Network
----- Charger | | Referenced
--- --- ----- to
----- | | --- Ground
--- | | - (in
| --- Speedtronic)
| |
----- - - - - -------------- - - - - -</pre>
This circuit is exactly the same as the one you described--two equal value resistors in series with the center grounded. And it works exactly as you said.

The circuit resides on printed circuit cards in the Speedtronic turbine control panel, but it monitors ANY device or circuit powered by the 125 VDC.

It's very difficult to draw, but the output of the AC-to-DC converter (DACA) is connected in parallel with the battery/charger, and it is separated by a diode scheme that performs a "high-select" of the supplies so that when the battery/charger output is higher than the DACA the DACA is blocked from supplying the Speedtronic. And, it also prevents the DACA from powering any other devices or circuits outside the Speedtronic turbine control panel. In other words, the DACA can't power the Emergency D.C. L.O. Pump or any other circuit the battery normally supplies. And, when the battery/charger is isolated from the Speedtronic the Speedtronic also can't monitor any of the devices/circuits normally fed by the battery/charger for grounds.

And--again--the Speedtronic ground detection circuit doesn't just monitor devices and circuits connected to the Speedtronic turbine control panel. It also monitors every device and circuit powered by the 125 VDC battery--when the battery/charger is also supplying the Speedtronic.

Some steam turbine applications do not have a 125 VDC battery and charger to power the Speedtronic turbine control panel--they only have DACAs (usually two of them) to provide 125 VDC to the Speedtronic. In this case, the Speedtronic only monitors those devices and circuits connected to the Speedtronic--and this thread was started about a gas turbine application, which is the overwhelming majority of the type of turbines Speedtronic turbine control panels are controlling and protecting.

But--for this type of circuit to work the 125 VDC must be ungrounded. That means if the power is coming from a battery and charger the battery/charger output must be ungrounded when it's connected to the ground detection voltage divider network that is <i><b>referenced</b></i> to ground. Neither phase (leg) of the DC supply is actually grounded by the ground detection circuit--the resistors "isolate" each phase (leg) from the ground connection, which makes it a reference circuit.

Hope this helps!

 

Thanks again CSA. Impressive text to graphic skills!!

Does this mean if we have a single fault to ground, let's say on the (+) side for sake of example, that we would have a current through the ground fault to the ground into the middle of the voltage divider and back to the (-) terminal of the battery or inverter?

Presumably it would be a small current due to the size of the resistance in the half of the voltage divider it would travel through, or are there diodes preventing any actual current flow through the grounded center in the voltage divider?

Thanks again,
Frank

 

Frank,

There are no hidden diodes in the circuit as shown--just as there were likely no hidden diodes in the light circuit you described.

The only difference from the circuit shown and the one in Speedtronic turbine control panels is that there is a physical "jumper" which can be used to disconnect the ground from the resistors; otherwise it's exactly the same.

The diodes I described were located elsewhere in the Speedtronic turbine control panel and are used to protect the DACAs from the battery/charger and vice versa--remember in the Speedtronic turbine control panel there can be multiple sources of 125 VDC so it's necessary to isolate them. I didn't draw the DACA source (the DC output of the AC-DC converter). (By the way, incandescent lights are purely resistive loads.)

So, yes, there would be some small amount of current flowing--just as in the lighting circuit you described.

 

>So, yes, there would be some small amount of current
>flowing--just as in the lighting circuit you described.

Thanks CSA. Now that I think I confidently understand this 125VDC system, the other thread I started on our neutral fused cabinet, brought up another possibility: a 125VDC battery system where the batteries are connected to ground in the middle, generating a +/- voltage. I think my emergency 125VDC system for my switchgear breaker control is this way, but as it's not part of the Speedtronic discussion, I'll leave it be.

Thanks,
Frank.

 

Hi, we have had several similar issues with earth leakage on a pair of Frame 9E machines. Over the years we have seen several chaffed wires (compressor bleed valve wiring, AA temp switch) and some leaking switches in the accessory compartment.

Although I'm unsure how to explain it, we also had a issue with the batteries themselves leaking. They were identified by the installer and once replaced the earth leakage monitor on the Speedtronic once again returned to its balanced state.

One procedure I have found usefull for identifying device leakage is to go into JB2 in the accessory compartment and physically disconnect the devices as many have common supply points here.

Good luck searching.

 

Hi,

If these two posts were found to be not helpful, it would be helpful to me (the author) if I knew why they were not helpful. I'm always trying to improve my explanations--many times at the sacrifice of too many words--but I also try to anticipate future questions and I'm always cognizant of others who read these posts looking for information and so I try to provide as much information as possible. Sometimes too much for the original poster, I'm sure--but for others who are new to the field and these concepts it might not be near enough. If I err, it's on the side of being verbose and trying to provide as much information as possible.

There's a saying that goes, "If you give a man (or a woman) a fish, you feed him (or her) for a day. If you teach a man (or a woman) to fish, you feed him (or her) for life." My intent is teach people to fish--although I know many just want to be given a fish (they just want the minimum information possible to solve their problem). But, the beauty of these forums is that many people can read the questions/problems and see the answers--and quite often we do not get enough information to provide a simple, concise "do this and your problem's solved!" answer. So, I try to provide sufficient background to understand the concept and intent and functionality so that absent critical information people can troubleshoot their problem without a lot of back-and-forth. (And, truth be told--many times when we ask for clarification, we get little or nothing in the way of answers that are of any use. So, to try to reduce the back-and-forth I try to put as much information out there so people can troubleshoot their problems by understanding concepts and intent and apply it to their situation which wasn't very well explained or detailed in the original post.)

So, please--if you feel something is wrong with the posts or have constructive criticism then let me know. I'm all about improving the quality of my posts--but I can't do that if I don't know what was not helpful. (There is one person who finds many of my posts unhelpful, and doesn't provide any feedback to help improve them--but that's okay. I just consider the source and let it go at that. I'm not offended by those unhelpful votes in the least.)

If there's anything constructive that can be offered to improve my posts--and in particular these two (which were difficult to write and describe without being able to post schematics)--it would be most helpful, and I would be grateful, if those who find them unhelpful would take a couple of minutes to provide the feedback.

Thanks!

 

I just wanted to thank you for the extremely detailed responses. We are having issues with intermittent DC grounds at our substation and this message board has been very helpful. Thank you so much for sharing your knowledge.

LFord

 

LFord,

Thank YOU for the feedback! It is most appreciated by all who read these threads because it helps others to know what was useful and what was not. We like to say here at control.com, "Feedback is the most important contribution!"(c) because it's true. Without the feedback we get from people--good or constructive, or even bad--others won't know if the information provided was good, constructive or even bad.

I firmly believe that feedback is one of the reasons why the Speedtronic control community here at control.com is one of the best on the World Wide Web. And, being a long-time contributor here at control.com, I especially hearing when people find the information useful, or even not--as long as they can be constructive so I can improve my responses.

Hope you will continue to read and learn--and even add your experience and knowledge!