MKVIe Ground Detected Above 3300 RPM

J

Thread Starter

J

Hello,

We have a 7FA frame GT that was originally controlled by a MKV and has be upgraded to a MKVe migration i believe it is called. Basically it is a MKVIe controllers with the old MKV hardwired terminal boards. Anyways, I am just curious if anyone has ran into a ground alarm on there control system that only comes in above 3300 rpm. If anyone has, please share with me your discoveries as this has been tough to troubleshoot as it can only be done when the unit is online.

Thank You
 
J.

You will get the detailed reply from CSA. As far as I can tell you must check the SOVs which are actuated just before the grounding alarm appears. You can check the Events, SOEs and historical alarms. If you find that the alarm comes when SOV operates, then you need to check the loop of that SOV.

Hope this will help.
 
hi sir,

Regarding ground issue, check all the junction box inside turbine compartments, gear box, and generator side wires insulation is proper or not. Any shield is touching or near to junction box body, may during running the junction box vibration can, make touch with wires or shield. So check it every thing properly.
 
Hello,

First, try with an online ground fault detector, not very useful as many inputs has the same common wire...

Secondly, you can simply by using Alarms viewer utility of GE, you can see exactly what happen at 3300rpm. If all the time the ground appears exactly at 3300rpm level of speed, then I'm pretty sure that at this level we have some valves limits which change their positions open to close or vice versa.

However alarms/events will show you exactly what happening. Take a note and then the next shutdown you can do your wiring check.

NB: be careful! as alarms appear in delay from valves move (cause alarms appear after reaching the setpoint of 30vdc in 125vdc application). Then you need to look of what is moving/changing state, few seconds before alarms appearing.

Good luck.
 
J,

You've been given some good advice and troubleshooting techniques.

3300 RPM is approximately 92% TNH (speed), so you're looking for something that has 125 VDC connected to it (a solenoid-operated valve (SOV)) or some kind of switch (limit switch; level switch; etc.) that is changing state just before the alarm comes in. Once you identify the device(s) that might be changing at that time, when the unit is off-line you can try toggling them (forcing the solenoid(s)--once you understand any knock-on effects of forcing the solenoid(s) you want to test), or changing the state of a switch, to see what happens. You might just visually inspect any device(s) you identify as being a potential ground source and get lucky and see obvious problem(s) with the device or the wiring to the device.

125 VDC battery grounds can only be caused by devices which are directly connected to 125 VDC. <b>BUT,</b> the device which might be grounded <b>DOES NOT</b> have to be connected to the Mark VI!!! The Mark VI 125 VDC battery ground detector looks at ANYTHING that is connected powered by the 125 VDC battery that is powering the Mark VI. So, if there are devices (such as switchgear contacts in the switchyard) powered by the same 125 VDC battery as is powering the Mark VI, and they develop a ground, the Mark VI senses that ground and alarms. Emergency 125 VDC lighting in compartments is another example of devices that are connected to the same battery as the Mark VI, but are not connected to the Mark VI. (The Mark VI has NO way of isolating devices that are not connected to it--no control system can do that very easily. So, any device connected to the battery (other examples include the Emergency DC L.O. Pump motor, and Emergency Seal Oil Pump motor--these motors are not directly connected to the Mark VI) that develops a ground will cause the Mark VI to annunciate a 125 VDC battery ground.

I believe the LCI and exciter are being switched out at around 90-95% speed, and there are some motor-operated switches with status contacts ("limit switches") connected to the Mark VI. I also believe the motor operators are, in some cases, powered by 125 VDC. So, don't forget to monitor these inputs to try to help pinpoint what might be causing the problem.

Troubleshooting 125 VDC battery grounds can be challenging, but I always start by looking at devices which are NOT in compartments and which might have gotten rain water into them or the junction boxes (JBs) nearby. Many outside JBs are improperly installed during construction and have conduit penetrations in the top of the JBs which are not properly sealed, if they are sealed at all. If the unit has off-base cooling water modules provided by the turbine packager the circulating fans usually have "vibration" switches which are exposed to the elements, and I often find moisture in them (though they shouldn't be changing state during start-up, I'm only using them as an example of devices outside of compartments where I have commonly found grounds).

Again, only devices in the field which have 125 VDC connected to them can cause 125 VDC battery ground alarms. This eliminates thermocouples, 4-20 mA pressure transmitters, RTDs, speed pick-ups, vibration pick-ups, etc. So, things like switches (any kind of switch: level, temperature, limit, pressure, etc.) and solenoids are potential ground sources. And, again, if you've had a lot of rain (or snow) recently, it's always good to check outside JBs and devices for moisture. And if you find the ground is possibly not on something connected to the Mark VI--remember, it can be on any device/circuit that is powered by the same battery as the Mark VI is powered by.

Hope this helps! Please write back to let us know what you find!!!
 
From the information you have provided, I would suspect the problem to be in a solenoid that operates when L14HS (or L14HSX) picks up. You may be able to troubleshoot this when shut down. Compressor Bleed Valves would be one of the first things I would look at, but there are a lot of things that happen at that speed. While shut down (unit at zero speed or on turning gear) just force the relays that actuate the various solenoids on the gas turbine at 14HS. (Be careful about any solenoids on the generator, especially if it is hydrogen cooled.)
 
CSA

I need you assistance in understanding concept about grounding in switches. I am looking at Appendix D (Signal Flow Diagrams) of GEH-6195 for writing this. Contact input (switch) is connected across POS and RET of DTBA.

POS is connected to P 125VDC via a jumper and RET is connected to N 125 VDC via 150Kohm resistor. Lets suppose that the contact is open and there is no grounding in the system then we will get +62.5VDC on POS with respect to ground and -62.5VDC with respect to ground. In this case if the contact closes, we will get +62.5VDC (with respect to ground) on both the terminals of the switch. In the system we will have the potential difference of 125VDC across 150Kohm resistor.

We have two situations for the grounding.

Case 1: If the POS side of contact gets grounded.
If POS side of the contact is grounded we will get the ground alarm (the ground alarm is annunciated when the absolute voltage on either side is below 31.24 V. We can find exact value in IO_CFG in TCQA card definition Socket 1- screen 21/21) no matter if the contact is open or close. In this case we will get 0VDC POS side and -125VDC on the RET of the contact. Am I right?

Case 2: If the RET side of contact gets grounded.
This one is a little trickier I guess. I will discuss three scenarios of this case.

In scenario 1 lets suppose that the contact is OPEN and RET side of one switch is grounded. 33Kohm of PD core and 150Kohm of DTBA will become parallel and now the voltages will not be equally divided. We will get the equivalent resistance of 27Kohm and voltages on POS side will be +68.75VDC and on RET side -56.25VDC with respect to the ground. In this case we will not get the ground alarm as absolute voltage on either side is not less than 31.24 VDC RATHER we have unequal voltages on both sides.

In scenario 2, contact is OPEN and lets suppose that more than 2 RET sides of the contact are grounded. As the number of RET gounds increase the voltages will become more unequal. The alarm will be annunciated when the voltage on RET with respect to ground are less than absolute -31.24 voltages. This will happen when more than 7 RET sides of contact are grounded.

In scenario 3, the contact is CLOSED and even if one RET is grounded we will get the ground alarm because as soon as the contact closes the POS will get grounded. The differential across 150Kohm will still be 125VDC but the voltage on both the terminals of contact with respect to ground will be 0VDC. On HMI we will see the voltage levels of +0VDC and -125VDC in this case.

To troubleshoot if the voltages are unequal we need to check the switches which are OPEN at that time and if we have 0VDC and -125VDC we need to check the switches that are CLOSED at that time.

FOR SOVs it is very different. If SOV is grounded we will get alarm as soon as they are grounded reason being one side of SOV is connected to ConnSO and the coil resistance is very low. Am I right in this case?
 
Esoteric_Stone,

I really only see one question, about ground in 125 VDC SOV (Solenoid-Operated Valve circuits (including the solenoid). My experience with grounds in solenoid circuits is that a ground in the circuit on either side of the solenoid will result in a 125 VDC Battery Ground alarm. If you draw out a typical solenoid circuit, the side of the solenoid that is connected to the SO terminal of the DTBC or DTBD will always have a nominal -62.5 VDC on it. And that nominal -62.5 VDC passes through the coil and through the wire that goes all the way back to the DTBC/DTBD NO terminal (usually the NO terminal). So, when the solenoid output of the Mark V is not energized the NO terminal will read nominal -62.5 VDC. So the entire circuit (both leads of the solenoid coil--from the DTBC/DTBD out to the solenoid and back to the DTBC/DTBD) has nominal -62.5 VDC on it (when the solenoid output is not energized).

When the solenoid output is energized, the NO contact of the relay in the Mark V closes and applies a nominal +62.5 VDC to one side of the solenoid coil, and with a path for current to flow through the solenoid coil the solenoid is energized with 125 VDC.

Now if a ground develops on any wire connected to the solenoid or the solenoid coil itself, my experience is that when the solenoid output of the Mark V is de-energized the battery ground will appear to be on the negative leg of the battery (when measuring the DC voltage at the <PD> core input terminal board). And, often, when the solenoid coil is energized <i>if the ground is on the "positive" side of the solenoid coil/solenoid coil wiring (the wire connected to the DTBC/DTBD NO terminal)</i> the ground will appear to shift to the positive leg of the battery. (This drives many people crazy--when the ground appears to shift from one battery leg to the other, and rightfully so.)

BUT, the nature of the ground also has something to do with the shifting--in my experience. As you note, any time the magnitude of the voltage with respect to ground of either leg of the battery drops below approximately 31 VDC a ground will be annunciated. So, moisture which is getting into a junction box or field device will generally start out as a higher voltage (closer to approximately 31 VDC) and will get worse (closer to 0 VDC) over time. And this "resistance" can cause the ground to behave differently (sometimes it shifts when the solenoid coil is energized, and sometimes it doesn't....); it can intermittent, and that can be more frustrating.

As for the contact inputs ("switches") and your various scenarios, I think you are predicting things and not speaking from actual experience. And my experience is slightly different from your predictions. I've never understood the reason for the presence of the 150 kohm resistor, and I've never really given it a lot of thought. I've never found more than two or three grounds in the wiring connecting a Mark* panel to its I/O (and usually they are in contact input circuits--because they are the largest number of I/O connected to the panel), and yet I've seen "hard" grounds (0 VDC with respect to ground) with just two grounds on one side of the circuit.

In my experience the thing that most people don't understand or realize when they are troubleshooting grounds in a Mark* panel is that the ground can ONLY be on a device that is powered by 125 VDC. I have seen many people power down the entire Mark* panel and take an ohmmeter and start checking every terminal in the Mark* with respect to ground, and they find grounded RTDs and grounded T/Cs and grounded mA inputs--and they spend a LOT Of time "troubleshooting" those grounds, which CANNOT cause a 125 VDC battery ground. Only devices which are powered by 125 VDC can cause battery grounds.

And, along with that is the fact that ANY device powered by 125 VDC that develops a ground can cause a 125 VDC Battery Ground alarm--and the device DOES NOT have to be connected to the Mark*! I have seen grounds in DC motor field circuits, grounds in 125 VDC emergency lighting circuits, grounds in fire detectors (not flame detectors--fire detectors), and grounds in switchyard junction boxes (protective relay circuits). The 125 VDC that powers the Mark* panels quite often also powers the Generator Protection Panel protective relays and devices (transformer over-pressure protection; breaker status; etc.). So, people spend a LOT of time troubleshooting devices connected to the Mark*--and not finding a ground or grounds--when the ground is on a device that is not connected to the Mark*.

The next thing that frustrates most people when troubleshooting grounds in a Mark* panel is that GE and many of its packagers usually only run a few wires out to the field for the BUS side of Mark V contact inputs, and for the SO side of solenoids. And, then in junction boxes in the field they jumper many devices to those few wires. What happens is that +62.5 VDC from <CD> can be run out to a JB and used to provide one side of devices that are connected to <CD>, or <QD1> (or <QD2>, if so equipped). And, similarly, the SO power (-62.5 VDC) power for solenoids can be jumpered to solenoids that are connected to any of <CD>, or <QD1> (or <QD2>, if so equipped). This make troubleshooting more difficult--but not impossible. It would have been better if two wires for every device (such as a contact input (switch) or solenoid) were run all the way from the Mark* to the device and back to the device--but, that would requires a LOT of wires, meaning the wire troughs in a Mark V would have been even FULLER and MESSIER!!!

Again, it's nice to be able to predict things and make statements that would help make troubleshooting grounds in a Mark* panel easier, but, my experience is that that's harder than just digging in and finding and resolving the ground. I've troubleshot hundreds of grounds over 30+ years (approaching 40 now!) and while some seem to follow some kind of "logic," most don't. Or maybe it's just that when I get to site to troubleshoot a ground that time is so short I just dig in and get to the task and generally leave and travel to another site without time to really dissect and understand what I found!.

Hope this helps.
 
CSA

Wow. Excellent. Very well explained. I have entered in Gas turbine controls world in 2014 and I do not have experience like you. You have more experience than my age!. I consider myself as your student as I have learned a lot from you via this forum and want to learn more.

>As for the contact inputs ("switches") and your various
>scenarios, I think you are predicting things and not
>speaking from actual experience. And my experience is
>slightly different from your predictions. I've never
>understood the reason for the presence of the 150 kohm
>resistor, and I've never really given it a lot of thought.

Yes you are right. My discussion about these scenarios was purely theoretical and has nothing to do with the practical. It was just for understanding. I will try to explain why 150 Kohm is used. The contact in MARKV is connected as follows<pre>
+125VDC----------------------------------------------------
| |
(33 Kohm) |
|____________ (Ground) (contact)
| |
(33 Kohm) (150 Kohm)
| |
-125VDC----------------------------------------------------</pre>
If there is no 150 Kohm resistor, then when the contact closes we are INTENTIONALLY shorting the positive and negative terminal of battery. We can not do that, so we use the 150 Kohm resistor. The scenarios in my previous post were based on this diagram.

>And, along with that is the fact that ANY device powered by
>125 VDC that develops a ground can cause a 125 VDC Battery
>Ground alarm--and the device DOES NOT have to be connected
>to the Mark*!

Yes we have been through this problem. There was a 125 VDC battery ground alarm and the problem was found in batteries. We were able to diagnose that the problem is not in MARKV by reading your several posts on grounding on this form, but still it is very difficult to convince the people that although MarkV is annuciating alarm but it is not the culprit.

>Hope this helps.
Your posts are always helpful!
 
Esoteric _Stone,

>it is very difficult to convince the people that
>although MarkV is annuciating alarm but it is not the
>culprit.

Yes, operators and their managers, mechanical department personnel and Plant Managers will always insist the ground simply MUST BE and CAN ONLY BE on a device connected to the Mark* because it is such a miraculous control system and can do something no other control system can do. Except when it is not a miraculous control system, which is most of the time, and then it is the root of all evil!
 
The diagram in the above post is<pre>
+125VDC----------------------------------------------------
| |
(33 Kohm) |
|____________ (Ground) (contact)
| |
(33 Kohm) (150 Kohm)
| |
-125VDC----------------------------------------------------
</pre>
 
Top