Hi Users,

We have GE's Frame V Gas Turbine Unit with generation capacity of 20MW. The Control system of the machine is Speedtronic's MARK-VI based control system. Recently, we faced a problem in the control system and due to communication failure between the cores, the machine got tripped. We didn't have spares with us, so the downtime ran into approximately 96 hours. The spares were arranged somehow from other sites and the control system was revived. But this is besides the point.

The main issue is related to Ratchet Sequencer Mechanism. Since, the MARK-VI control system was completely out of the picture, the ratchet mechanism or the 'cooldown cycle' could not be operated from control system. In such situations, we are supposed to use jogging option from the field.

But, for the convenience of operator's we designed a small circuit using a timer. In this circuit, a run command to ratchet motor and close command for 20CS was given at an interval of 3 minutes for a period of 15 seconds. Though this circuit bypasses 33HR Limit Switch Feedback, but the purpose of Ratcheting was served without the operator having to visit the field, every 3 minutes for the job.

My question is, can this circuit be built and kept ready as a prototype? Are there any downsides to using this circuit? We may have developed a circuit without looking at the repercussions in details. Are we missing out on any downfalls?

Regards
Sadiya

 

Most GE-design Frame <b>5</b> heavy duty gas turbines have the Young & Franklin Hydraulic Ratchet Self-Sequencer, which controls most of the ratchet operation, but still relies on the turbine control system (Mark VI; Mark V; Mark IV; Mark VIe; etc.) to initiate the cycle. AND, the local Hydraulic Ratchet Jog Pushbutton just sends a signal to the turbine control system to initiate the cycle when the button is pressed, and to stop the cycle when the button is released. So, the system still relies on the turbine control system, no matter.

My recommendation would be to use a small programmable logic controller which has suitable inputs/outputs (which can handle 125 VDC) to accept the signal from the turbine control system to initiate the cycle and to stop the cycle (L4HR), and from the jog push-button (43HR-JOG). And put the logic for energizing 20CS and for monitoring 33HR into the programmable logic controller, as well as connecting 33HR to the programmable logic controller. I would also suggest using one of the programmable logic controller discrete outputs to mimic the 33HR signal to the turbine control system, if for no other reason than to be able to monitor it using Toolbox.

By putting all of the logic for the ratchet operation and control into the programmable logic controller, and by writing a little bit of code to accept a switch input (maintained) to start the logic should the turbine controller be unavilable you would have the redundancy/back-up you desire. Under normal operating sequences, the turbine controller would be sending a signal to start the operation, and the 43HR-JOG push-button could also still be used. And by adding a switch input from a switch mounted somewhere convenient for the operators the ratchet operation could be initiated and maintained for as long as the switch was closed. The entire sequence, including the 3 minute timer, could be written into the programmable logic controller--which is independent of the turbine controller.

The thing you want to make sure of (and this could be done relatively easily with a couple of additional inputs) is to make sure the Aux. L.O. Pump is running and there's sufficient pressure for the bearings. It could even be possible to drive the Emer. L.O. Pump in the event that the Aux. L.O. Pump was not in service (something the turbine controller sequencing can also do). There needs to be L.O. to the bearing when the unit is being turned by the ratchet.

So, yes--this is very doable. I would suggest making the programmable logic controller as independent as possible by connecting the ratchet I/O to the controller, and where necessary, mimicing any outputs to the turbine controller.

33HR is used by the logic in the turbine controller to determine when the unit is in a forward stroke or a retraction stroke. The sequence is that after the unit completes a forward stroke the limit switch changes state and the self-sequencer initiates the retraction stroke. When the retraction stroke is completed the limit switch changes state again, and the self-sequencer starts another forward stroke. This is all done by the self-sequencer. When the turbine controller senses the self-sequencer has completed a retraction stroke and has started a forward stroke a short time is started and then the ratchet sequence is stopped and the 3-minute cycle timer starts again. The reason for the short period after the forward stroke starts is to try to ensure the jaw clutch teeth remain engaged (or the SSS clutch mechanism remains engaged). Even if the ratchet isn't on a forward stroke when it's stopped (such as during operation using jog push-button), when the ratchet is re-started 20CS is always energized to close or to keep the jaw clutch halves closed.

So, your "circuit" will work without even using 33HR, but it's kind of a good idea to continue to use it.

When connecting 33HR to the programmable logic controller, it may be necessary to do a little re-wiring to make sure the turbine controller discrete input wiring is properly disconnected from the turbine controller before connecting it to the programmable logic controller. And, again, the outputs of the programmable logic controller need to be capable of supplying the power (volts and current) required by the devices (solenoids; motor starters, if desired; turbine controller discrete input circuit).

Hope this helps!

 

Thanks a lot for the prompt reply, CSA. I will try to build an elaborate prototype as suggested by you.

In your post, you have mentioned that, 43HR-Jog sends a signal to MARK-V, VI, VIe, etc controller only. Then, in case of a failure of control system, like the one we faced at site, how would one ensure ratcheting for the machine, which is important from the point of view of the machine?

Regards
Sadiya

 

Sadiya,

Yes; the manual jog feature relies on the turbine control system. So, if the turbine control system is not working, then the manual jog feature isn't going to work, either.

This topic--an inability to maintain cooldown (be it hydraulic ratchet, or turning gear, or slowroll)--has been covered MANY times on control.com. Gas turbine rotors are very different from steam turbine rotors, and the internal clearances of gas turbines are much larger than the internal clearances on steam turbines. When a gas turbine rotor--more correctly, the axial compressor section of the rotor, because the compressor rotor is MUCH longer than the turbine section of the rotor--is warm the metal is in its elastic range, which means it can stretch but will usually return to near normal shape when it cools. (This is opposed to the plastic range, when the material irretrievably bends and does not return to anything near it's normal shape. If the rotor can't be turned when it's hot and reaches zero speed it will sag under its own weight (that elastic range phenomenon) and it is common for rotating parts to come into contact with stationary parts--which is why one doesn't want to keep trying to turn the rotor when it has been stationary for more than about 30 minutes or so, for a period of about 24 hours. But, <i>eventually</i> it will begin to un-sag, and <i>eventually</i> will return to a very nearly straight condition--at which time it's safe and desirable to turn the rotor to help it return to a near perfectly straight condition.

So, while it's much more preferable to keep the rotor turning (even using ratcheting) while it cools, if it can't be kept turning as it cools <i>eventually</i> it will return to near straight. This may take as long as 36 hours if the rotor was hot when it was shut down and couldn't be turned at all. But, <i>eventually</i> the rotor will return to a near-straight condition--and that eventual time is usually about 24 hours. During which time it is NOT advisable to try to turn the rotor--because it HAS sagged and there IS contact between rotating and stationary parts and efforts to turn the rotor can result in catastrophic damage.

So, patience is key. Yes, it means the unit can't be restarted as quickly as it could have been had the unit been on cooldown when it reached zero speed, but it will be possible to re-start the unit and damage will be negligible--AS LONG AS THE UNIT ISN'T DAMAGED BY CONTINUED EFFORTS TO TURN THE ROTOR SHAFT WHEN IT'S HOT.

I have been working on GE-design heavy duty gas turbines for more than 30 years (three decades), and I have seen and heard of LOTS of machines that couldn't go on cooldown (for ALL kinds of reasons--including control system problems, though that RARELY happens) and with one exception the rotor was undamaged. And that exception was a site that kept trying for hours to get the shaft turning and used hydraulic jacks to turn the rotor. And, because the rotating parts and stationary parts were in contact with each other when they finally succeeded in turning the shaft the blade broke, which caused other blades downstream to break. And, it took more than USD1.6 million dollars to repair the damage, and longer than one month. Which, when compared to 24 hours of lost generation revenue was HUGE. If they'd just waited patiently for 24 hours--as they were advised--there would have been NO problems.

It's unfortunate that you experienced a control system problem, but I have to think there are or were some other circumstances which we aren't aware of which contributed to the control system problem. Because, it's just not normal for a TMR control system to fail as described.

The second most important thing to do when it's not possible to get or keep a unit on cooldown is the make sure the Aux. L.O. Pump keeps running to maintain a flow of oil to the bearings. This is for cooling, because if there's no oil flow to the bearings then heat from the compressor and turbine sections of the rotor will cause damage to the bearing material. So, even if you can't keep oil running continuously you need to perioidically cycle oil flow to keep the bearings cool.

But, the MOST important thing to do when it's not possible to get or keep a unit on cooldown when the rotor is hot after it reaches zero speed is to just WAIT. 24 hours is the minimum amount of time to wait. If the unit can't be turned within about 30 minutes of reaching zero speed, then stop trying to turn it. (That's a guide line, 30 minutes, but if the rotor was running at Base Load and tripped or was shut down and couldn't get on cooldown within about 30 minutes of reaching zero speed then the axial compressor rotor has already started sagging.)

Hope this helps!