Adding instrumentation to curtail an oil spill on a GE Frame 5 with Mark 5 controls


Thread Starter

Mark Allen

One of the sites that I work at uses four GE frame 5 turbines with for power generation. The units were installed in 1971. The units were rebuilt in 1999 and Mark V controls were added. The site is un-manned when the units are not on-line, which is most of the time. The units are used mostly for summer peaking.

The units have the original hydraulic ratcheting means that sits atop the accessory wheel. On 5 different occasions during the 12 years that I have been maintaining the units, a hydraulic line or some other pressurized part of the ratcheting means has failed resulting in the spillage and subsequent clean-up of all the oil in the lube oil sump. This happens when the site is unmanned and the units are in the Cool-Down mode.

The site where the turbines are is 50 miles from the site I normally work at. I am trying to come-up with a means of controlling/limiting the oil spill as early in the spill as possible.

This is the scheme I am considering. Adding an additional liquid level detection device in the lube-oil sump, the original lube-oil level gauge has given trouble on several occasions on several of the units. I would like to install the new level detector in the access area where the high-pressure hydraulic oil pressure regulator is located. This area, with the access door that lifts, looks like a good location that would provide for easy installation/calibration/ maintenance of the device. I would like to set the level gauge to come into alarm state once the oil level drops an inch below the level that the current low level alarm comes into alarm state.

I would also like to install a RTD on or next to the aft turbine bearing. I would like to connect both devices to the Mark V and configure it so that if the new level gauge came into alarm state and the controls sensed the unit at zero rpm and the aft bearing temperature at a safe level then the controls would halt the cool-down process, The process would only resume if the aft turbine bearing temperature rose to unsafe levels. Once the bearing temperature of the aft turbine bearing had cooled to safe levels due to the resumption of lube oil flow then the cool-down process would be halted again. This would continue until someone arrived at the site and took steps to fix and issues with the unit.

Does this sound like something that would work? Are there any dangers to the unit that I might be over-looking? Is there a better way of accomplishing what I am trying to accomplish?

Mark Allen

Bob Johnston

I guess the plan is do-able but I have a few questions.

1) Do you really need to keep the machines on continuous ratchet once they are completely cooled down? If your start-up is predictable, you could start the ratchet process 24-48 Hrs. before start-up. GE recommendation from GEK 3620 is 6 hours ratchet before start-up on a cold rotor.

2) Wheelspace temp. would be a better permissive to shut-down cooldown as that is GE's permissive (All wheelspace temp.< 150 Deg F/65 Deg C)or you could just use a time after shutdown e.g. >24 Hrs.
Bob Thanks for the in-put, I need all I can get.

The machines are not normally in cool-down mode. The machines are put on cool-down only after the dispatcher notifies us that they will need the units. The units are then left on cool-down for 24 hours after the units are shut-down.

The units are often dispatched until 8 or 9 o’clock at night during the normal summer peaking period. As luck would have it the oil-line or what ever part never breaks while I am still at the site putting the units to bed but normally it happens at some wee hour of the night. I took all the tubing apart that is associated with the ratcheting means on the units and inspected them hoping to spot any weakness and there-by prevent a failure. Every time I think I have looked at everything an unexpected part fails. An example, the summer I took all the tubing apart and inspected it, the Borelian <sp> tube in the forward stroke pressure sensing device on one of the units broke and all the oil spilled.

I had considered using the wheel-space temperature detection, the instrumentation is already in place and used by the Mark V controls. This is my thoughts on why I thought reading the aft turbine bearing temperature might result in controlling an oil spill better.

I thought that if I used the wheel space temp as the permissive then the cool-down process would continue uninterrupted till the temp dropped to some point (I was thinking maybe 200 degrees F.) and that might take a long time, long enough to spill a lot of oil.
I thought if I could sense the aft bearing temperature then the cool-down oil pump could be cycled on and off more precisely and result in the pump running only enough to keep the bearing safe, resulting in less run-time for the pump if a leak had occurred.

You post has triggered another thought I would like your in-put on.
So far all of the oil spills have resulted in some failure of some part of the ratcheting means. All of the rest of the plumbing that is pressurized during the cool-down process is robust and never seems to fail. I am wondering about having the Mark V configured to prevent only the 125 volt dc hydraulic ratchet pump motor from starting if the low lube-oil sump level alarm is in alarm state yet still allow the cool-down lube oil pump to continue to run every 3 minutes.

I know this would leave the unit unavailable for the next day’s run until the leak was repaired and the unit then ratcheted for a period of time but it would very seldom occur.
At the current rate it would be 5 times in 12 years.

I am thinking this may be a good solution.

Mark Allen

here is some more information for you to chew on if it helps. I help maintain a couple Frame 5's and 7's. Our 5's are 1985 vintage, and the ratchet assy. is on the torque converter, not the accy. gear. To date we have not have any tubing failures or major oil leaks, I hope I did not jinx myself.

I would go with Bob on using wheelspace temps to shutdown cooldown. We created logic that stops cooldown after a minimum of 24hrs or after wheelspace temps are below 250 degf. We have a standby cooldown sequence that ratchets the units for 1 hour each day. These machines are also seldom used peakers. We do not ratchet before a startup since we typically have no idea when the units will be dispatched.

I know that there is logic on our 7's to run the emergency lube oil pump in the event of a loss of power for the accessory pumps if bearing temperatures are greater than 250 degf, but I don't think this is an issue on the Frame 5 units since the bearings are not subject to the heat like the #2 bearing on a Frame 7. I am assuming this is a two bearing Frame 5 machine, but assuming has gotten me wrong before. My current logic for our Frame 5's does not have any logic for bearing metal temperature protection. But to be safe just stopping the ratchet sequence would save you from a major oil leak in the event of a leak in the ratchet system, but still keep oil to all bearings, as long as the leak is in the ratchet system. Like you said most other oil system piping is robust, in this case I again assume it is the pipe in a pipe confiration.

I think logic to stop ratchet if a leak is detected should be good. If you truly have an oil leak that needs to stop the ratchet or cooldown sequence then you will have to wait for the 24 hours before you try to spin the machine to protect against blade rubs. But I suspect in this case you will be unavailable to fix the leak anyway.

Good luck, let us know how you proceed.
Thanks for the input Mike. I agree that putting the wheel-space into the logic change/addition is a good idea. Yes the lube oil plumbing is a pipe in a pipe. I also agree that the ratcheting means on the units you maintain is a much improved design over the units I maintain. The units I work on have three bearings, one at the compressor inlet, one just aft of the turbine and one on the collector end of the generator. The reason I am so concerned about the number two bearing is that it is located in a recess in the exhaust plenum and is subject to a lot of heat migrating through the turbine out-put shaft after a shut-down.

We take the extra step of opening the breakers for the cool-down and emergency lube-oil pumps in addition to the breaker for the hydraulic ratchet pump motor. We do this because the Mark V controls are set to default to "On-Cool-Down" mode any time the controls are re-booted. There are times that the power to the Mark V is shut down for one reason or another and when the power is restored people forget to take the units out of cool-down mode and just leave. By opening the breakers this prevents the units from ratcheting for a month at a time and there-by reduces the exposure to the oil-spills.

The company I work for has a plant modification program that reviews and makes an approval or not approved decision. I am going to submit this idea to them.

I will let you know how it turns out.

Mark Allen
I'm not really familiar with these old ratchet mechanisms that don't use the self-sequencer, but I believe there is still an alarm that would be generated if the unit did not "see" a ratchet stroke after a couple of cycles had expired. The alarm message is something like "Hydraulic Ratchet Trouble".

When there is a leak the ratchet cylinders won't be moved, and this alarm (if it's used on the type of ratchet mechanism at your site) would be annunciated after the Speedtronic failed to see some movement for some period of time. I did see one older set-up like this on a Frame 5 two-shaft machine some years ago and while it didn't have a limit switch to indicate the end of a forward stroke it did have pressure switches which would indicate a high pressure when the cylinders reached the end of their stroke (which is exactly what 33HR-1 does on the self-sequencer mechanisms, anyway!) It's likely that would be what the sequencing would be looking for to change the direction, and if it didn't see high pressure after some time the "Hydraulic Ratchet Trouble" alarm would be generated.

You could use the 'Hydraulic Ratchet Trouble' logic that annunciates this alarm to shut off the ratchet pump after some additional time (perhaps another two cycles or so), but you would want to leave the Aux. L.O. Pump running as long as the wheelspace temp was "high", or if the unit has bearing metal temperature T/Cs, you could use that as well, though it may cycle the Aux L.O. Pump as the bearing metal heats up after oil flow stops.

I've actually never understood why GE didn't stop the ratchet pump when this alarm was generated. It's mostly likely either the result of a leak in the hydraulic ratchet tubing/piping, or a rub that's not allowing the shaft to be turned, or some other condition that needs investigating before ratcheting can be continued. A failure to ratchet is troublesome and if the shaft can't be turned, then there's no point in continuing to run the ratchet pump for more than a few minutes, especially if there's an oil leak.

MIKEVI is too young to know, but one of the Frame 5 units at his site did have a hydraulic ratchet tubing rupture and subsequent leak in the first couple of years of operation that caused most of the L.O. to be pumped out on the ground. They used bacteria to remove the oil from the soil; it was a new and revolutionary idea at the time for cleaning up the spilled oil from the soil--and successful!
CSA, guess I am guilty without even knowing it!!! I will have to talk to some other folks who have been here longer than I and learn more about the past.

I like your suggestion to Mark on using the alarm for ratchet system trouble. For Mark any well thought out logic that includes lesson learned from others should be a great addition to current practices.

Another user group that I am part of has just started a new part of the site that will allow users to post current turbine logic files for sharing. It is just starting so I don't know how it will progress yet. My hope is that it will take off and allow users to compare logic for machines and see what others have done to improve and enhance system features, without stepping on the the OEM's proprietary code.
CSA while I like your idea of using the Hydraulic Ratchet Trouble alarm it would fall short of stopping one of the oil spills we had. The semi-circular pressure sensing tube that senses sufficient forward stroke pressure cracked. The crack was minor enough that the system kept enough pressure to actually ratchet the turbine. The leak was such that all the oil was pumped out on the ground. The Hydraulic Ratchet Trouble alarm didn’t come in till the pump lost suction due to all the available oil in the sump being spilled.

Below is the request I sent to the plant modification team for review. I had considered adding an alarm to the Mark V control but after some site searches I found a thread by CSA explaining how to add alarms. It is way beyond my expertise so I settled on an alarm light on the MCC panel for now, I will request that the next occasion that we have a GE controls man at the site that they set-up the alarm then.

I will let you all know how it turns out.

As the Mabelvale units are currently set up, during the cool-down process the unit ratchets once every three minutes while the cool-down lube oil pump runs continuously.

The hydraulic ratcheting means on GE frame 5 gas turbines serves the same purpose as a turning gear on a steam turbine. The ratcheting means uses hydraulic oil pressure to provide the force needed to rotate the unit 1/6th of a turn every 3 minutes. The only purpose served by the ratcheting means during the cool-down process is to assure that the turbine shaft remains straight as it cools to ambient temperature. If the ratcheting means is stopped prematurely during the cool-down process the turbine shaft may acquire a slight bow to it resulting in a 24 hour delay of a unit re-start. The shaft will return to a straight condition after the 24 hour cooling period.

The Cool-down lube oil pump provides the needed lubrication during a unit shut-down. The pump also provides cooling to the aft turbine bearing preventing over-heating and damage to the bearing once the unit has coasted to a stop. The Cool-down lube oil pump is typically left in the run position for 24 hours after a unit shut down.

On 5 different occasions since 1999 a pressurized oil line associated ratcheting means has failed in one manner or another. This has resulted in the loss of 5,500 gallons of lube oil and 5 different environmental violations.

The modification to the current cool-down process that I propose is to install one additional lube-oil sump level low indicator per unit and to wire the indicator in series with the hydraulic ratchet pump motor. The new indicator should be set up to come into alarm state if the oil level in the sump reaches the level that the currently installed oil level indicator comes into alarm state. The reason I am not considering using the currently installed level indicator for the modification is that it has not proved reliable enough over the 12 years of operation that I have witnessed.

The new level indicator should have a normally open set of contacts and a normally closed set of contacts. The normally closed set should be connected in series with the 125 volt DC contactor for the hydraulic ratchet pump motor. The normally open set should be connected as to complete a circuit for an alarm light to be located on the unit MCC panel. The normal Cool-down lube oil pump operation would be unaffected.

This arrangement would stop the ratcheting process during the early stages of any oil spill that resulted from a ratcheting means oil line failure. This should reduce the size of the spill by almost 1,000 gallons. It would also likely result in a 24 hour delay in a unit re-start after the oil spill, but the oil spill itself has normally caused a several day delay for the unit re-start.
Mark Allen,

Thanks for the feedback! "Feedback is the most important contribution!"(c)

Very good choice to take all the known failure modes into account in your decision.

rajendra mehta

do you have bearing no-1 thermocouple oil seals GE part number? because oil leakage is an issue at my place