The existing Frame 7E Gas Turbines. A retrofit on the control system was carried out migrating from Mark II to Mark V Speedtronic Control System in the year 2000.

We have been working to improve our starting reliability and have studied the causes of failure to start. We have identified that failure of the clutch to engage within the 5 second allowed time is the most significant cause of failure to start. We have checked all the system components and system settings and rectified any faults found. We have achieved some improvement but still note that failure of the clutch to engage remains the most significant cause of failure to start.

To understand the cause of the unreliability we have studied the installed starting sequence in the Mark V CSP. The control sequence is that the Auxiliary Hydraulic Pump start command and the Starting Clutch Solenoid Valve (L20 CS1) energization are simultaneous with the Start Command and the Master Protective Relays energization. We have observed that in this sequence the hydraulic pressure takes approximately 8 seconds to build up to 105 bar as the accumulators fill. We observe that this means that during the clutch operation it is therefore not supplied with 105 bar as suggested by the Hitachi detailed hydraulic drawing 10R181-732. We propose that this situation means that the reliability of clutch operation is compromised.

We hold the view that the Auxiliary Hydraulic Pump should be started before the clutch solenoid is energized and the hydraulic oil supply pressure (105 Bar) shall be actually made available before the Clutch Solenoid Valve (L 20CS1) energization.

Our view is supported by the improved clutch engagement reliability that we achieve when the auxiliary pump is started in manual before initiating the start sequence. In tests we have carried out we see that we achieve a much more positive clutch engagement and significantly reduced time between clutch solenoid energization and clutch limit switch activation when the auxiliary pump is already running and system pressure is already 105 bar.

I am look forward to hearing from you in due time.

 

Have you checked the hydraulic dump valve? Many times this valve will stick open, causing low hyd pressure during startup, thus causing the clutch to not engage. On 7B machines, it's located on the aft section of the lube tank in the acc compt-on the right side--I think it's designation is 20 HD.

 

Ayed,

Marafiq; those units are getting old--but it sounds like you're taking good care of them!

From the information provided, it would seem you've proved your theory--and my recommendation would be to use the 'Hyd. Pressure Low' permissive to actuate the clutch engagement solenoid--and the 5 second timer.

So, the sequence during start-up would be to start the Aux. L.O. pump (a permissive to start the Aux. Hyd. Pump), start the Aux. Hyd. pump and when the low hydraulic pressure permissive is satisfied (L63HQL is a logic "0") the starting clutch engagement solenoid is energized, starting the timer to check for clutch engagement.

Please write back to let us know what you decide and how you fare!

 

thanks for those given advice. However, we need to lesson

1- did someone face like this problem in Gas Turbine frame 7E?
2- what is the root cause?
3- what was the action taken to rectify the problem?

let us hear from you gentlemen

 

Ayed,

There are stories like yours all the time of small glitches in sequencing causing problems such as you have described for years, decades even. If you own the machine and it's not under warranty or a long-term service agreement and you've analyzed and tested as you have described then it's perfectly acceptable--recommended--to make the changes necessary to improve operation and reliability.

I visited a site once that had horrible problems for years with their diesel starting motor failing to start if the unit had been running. It started just fine after a maintenance outage--no problems whatsoever. But, if the unit was started and ran for a few hour or even a few weeks (it was a Base Loaded machine) and tripped, or they had to shut down for a couple of hours for some reason the diesel wouldn't start on the first, second, third, fourth or fifth try--sometimes more. They tried starting fluid sprayed into the air intake of the diesel engine; they even tried putting a block heater on the diesel--all with no luck. They had diesel mechanics out to look at the problem (during maintenance outages, of course) and they could find nothing wrong.

This unit was located in a cooler norther climate. During a maintenance outage I was testing the compartment cooling fans when I discovered the gravity dampers on the Accessory compartment doors were installed backwards as well as the compartment vent fan was rotating in the wrong direction. I asked and was told this was the original configuration and fan rotation. The result was that cool outside air was being drawn into the compartment by the vent fan directly over the diesel engine--cooling the engine block to the extent that it made it so cold it was hard to start. In fact, all the site personnel complained the Accessory Compartment was always very cold--which was odd for a unit of that vintage.

I also found the compartment pressurization switch wired incorrectly and the inversion mask to be wrong--and they had been changed from the as-shipped configuration (which was documented in the as-shipped drawing the site still had in the Maintenance Department Library). Further, we found that the direction of air flow versus that shown on the Heating & Cooling Piping Schematic Diagram (P&ID) was opposite that shown on the drawing.

The gravity dampers were reversed, fan rotation was corrected, the switch wiring and inversion mask were corrected, and as far as I know the site never had a problem starting the diesel engine again. I actually ran into a gentleman who was the original commissioning engineer for that site a few years later and told him about the problem, and he told me they had had the same problem during commissioning (difficulty starting the diesel engine). He couldn't recall exactly why the vent fan rotation was incorrect, but thought it might have occurred because the gravity dampers weren't working correctly when they first ran the unit and instead of considering the gravity dampers were installed incorrectly they assumed the fan rotation was incorrect.

One other possibility we didn't discuss was have you checked the manual valve positions on the hydraulic accumulator assembly to be sure they are in the proper positions? There are usually two accumulator assemblies, both of which should be in service at all times, and each assembly has a small hand valve and a large hand valve. The small hand valve is the "bleed" or "drain" valve, only used for maintenance. It should be closed, tightly, at all time. The large hand valve is the "block" or "fill" valve, and should be fully open at all times (except during maintenance). If the manual valves are not in the proper positions, particularly the small bleed/drain valves, it might take longer than it should to establish hydraulic pressure.

Accumulator charge integrity might also be a contributing cause; if the accumulator bladders aren't properly charged or have ruptured, it might take longer to establish hydraulic pressure. Charging the accumulators with nitrogen requires careful attention to procedure otherwise it's very easy to lose all the existing charge in the accumulators and be unable to charge them properly. Usually, there is a label (if it hasn't been lost or painted over) on the top of the accumulator cylinders very near the charging adapter that lists the proper sequence for checking and charging the cylinders using the charging hose/gage assembly provided with the unit when it was new.

The changes recommended are in line with the information you provided, and would only extend the start-up time for the time required for hydraulic system pressure to get up to satisfy the low pressure switch setting--at which time the normal sequence of events would take place. Instead of actuating the clutch and starting the failure to engage timer at the same time the Aux. Hyd. Pump is started, you would actuate the clutch and start the failure to engage time AFTER the hydraulic system pressure was at least high enough to satisfy the low pressure switch setting.

Another option you could try would be to increase the failure to engage time setting from 5 seconds to say, 10 seconds, or slightly longer. The problem with this is that if, for some reason (say after hydraulic filter change or a long shutdown) it took longer for the hydraulic piping lines to fill and develop pressure the clutch might not engage fast enough--but it would engage faster on subsequent START attempts. Better to wait for hydraulic pressure to build up knowing that when there is sufficient pressure the clutch will close than for just choosing a timer value based on gut feelings.

The reason for the failure to engage logic is that if you have the typical jaw clutch you don't want to apply torque to the jaw clutch until the clutch has engaged--or serious damage to the jaw clutch can occur. If you have the SSS clutch, you don't even need to worry about that--there's no clutch engagement solenoid or hydraulic rams; simple centrifugal force from the starting means will engage the clutch. (Converting to a SSS clutch has other very useful benefits--mostly that if the unit trips during starting or when running and you need to re-start it quickly you don't have to wait for the unit to coast down to zero speed to engage the jaw clutch--you can start the unit again as soon as it has dropped below 14HM speed level!).

Your apprehension is understandable--but, there are many stories of sequencing which has been "working" for years, decades even, that was found to be in error, and once corrected operation and reliability was greatly improved. Sometimes, it's a simple adjustment to a temperature switch setting. The process you outlined for investigation and proof of your theory was excellent--and a small change such as recommended will not have any knock-on effects other than the few seconds it takes for hydraulic system pressure to increase above the low hydraulic pressure switch setting. There is no danger of causing unintended problems.

If you still have a copy of the Mark II Speedtronic Elementary have you compared this bit of start sequencing in the Mark V to what was in the Mark II? If so, what did you find? It's very common for GE to miss small things like this when performing a control system upgrade because they do not copy what was in the old control system; rather, they start using new unit sequencing and make modifications as they deem necessary--which aren't always correct. Usually, the commissioning personnel in the field will catch the error(s), but not always. Unfortunately, GE doesn't teach their field service personnel how GE-design heavy duty gas turbines operate or are to operate; the field service people learn on the job, and some learn better and faster than others. (Most leave the field before they get a real grasp of operating philosophies and sequences.)

 

thank you very much for your recommendations. However, let us check the following

1- to check the accumulators' valves
2- to check the logic of Mark II and Mark V

one more things and it is very staring to us. when we started the aux.hyd. oil pump before we start the unit no failure in the starting clutches. This indicates that the accumulator needs more time to be filled and it does not fill within the 5 second. also, the jaws was cleaned.

 

Whenever a Aux. Hydraulic Pump is started it takes a short amount of time to build up pressure--the amount of time depends on many factors including what servo-valves are trying to do, how dirty the hydraulic filters are (a higher dp usually requires a little more time to build pressure downstream), what the positions of the actuators are when hydraulic pressure is building up, and how long it takes to build up pressure in the hydraulic lines (how "empty" and long they are--especially if there are any off-base skids using hydraulic pressure/flow, such as an off-base Gas Fuel Module). And, there is also the time required to fill the hydraulic accumulators.

So, there can be lots of reasons why it takes some time to build pressure in the hydraulic system. You are right to be checking the accumulators to be sure they are properly valved in and charged, but there are other factors which can't be controlled during every start of the hydraulic pump.

You have done a very good job of proving if there is sufficient pressure in the hydraulic system when 20CS-1 is energized that the jaw clutch will close in less than 5 seconds. Once you've verified everything is okay with the hydraulic accumulators, you will have enough information to make your decision. It will be very interesting to see what the Mark II logic was for this bit of starting sequencing, and for when the Aux. Hyd. Pump was started during a normal START vs. what is in the Mark V.

Lastly, remember, that when starting the Aux. Hyd. Pump, there must be pressure in the Bearing L.O. piping because the suction of both hydraulic pumps (the Main and the Aux.) is pressurized/fed by the discharge of the L.O. pumps and there is an interlock to prevent starting/running the Aux. Hyd. Pump without sufficient pressure in the Bearing L.O. piping.

Please write back to let us know what you find and how you proceed.

 

Dear

I will divided your reply into points.

1- Suction of Aux. hyd oil pump.
It is always, and we are sure that there is sufficient oil pressure in the bearing header otherwise, we will receive an alarm. Also, we have frequently replace the lube oil filters.

2- Oil quality
We are using lube oil purifier and oil checked by lab and on issue at all.

3- Mark II logic vs. Mark V in starting sequencing
There is no change.

So,
What if we put low pressure check = false?