Control.com - Nerds in Control

I believe the bleed valve in a frame V GE Gas turbine is operating as follows:

During accelaration and decelaration mode of the turbine, servo valve (electrically operated) ahead of the bleed valve is actuated from mark V.
And when the servo valve is energised, a portion of the compressor discharge air passes through the servo vlave line and this air pressure opens the bleed valve (2 nos)mechanically.

Now after reaching a certain speed during accelaration (probably between 90 to 95 % of TNH), servo valve is deenergised, thus the air flow line is blocked and the compressor bleed valves is closed.

Now my query is:
(1) If during accelaration / running (turbine on full load) servo valve coil fails to actuate / deenergise ,is there any procedure where we can manually intervene for a short time?

(2) i am told of a situation when bleed valve trouble alarm has appeared in mark v (when the turbine was in full load), but on actual inspection the operational personnel have found both the bleed valves actually closed. What could be the reason?

(3) Is the above case may be due to any problem in servo valve ahaed of bleed valve?

MOST compressor bleed valves (that's the entire name of the devices you are enquiring about--compressor bleed valve(s)) are pneumatically operated, using the gas turbine's axial compressor discharge pressure through a solenoid-operated valve to close the valves against a large spring which is trying to open the valves. Generally, the valves are normally open, spring-loaded to open, fail open--all of these are valid descriptions of the valve.

Generally, attached to the valve(s) is a limit switch (device 33CB-n, where "n" is usually 1 or 2, depending on the number of compressor bleed valves). The limit switches are to be adjusted to sense when the valve is fully open--in other words, the switch is to be actuated when the valve is fully open, and, generally, the switch contacts are to be closed when the valve itself is fully open.

The compressor bleed valves are to be open during starting, acceleration, and shutdown--to protect the axial compressor against stalling and surging during acceleration and deceleration. (Hence, that's why the limit switches are usually set to sense a fully open condition.)

The solenoid-operated valve, usually device 20CB-1, is generally in the air supply line to the compressor bleed valve actuator. When 20CB-1 is energized (usually with 125 VDC), the solenoid causes the valve to port axial compressor discharge air (pressure) to the actuator(s) of the compressor bleed valve(s) to close the valves.

Since the compressor bleed valves are air-operated and usually located in the (very hot) turbine compartment, if there is a failure of the solenoid-operated valve 20CB-1, there is very little one can do to close the compressor bleed valves manually. The springs which keep the valves open when there is no air pressure from the solenoid-operated 20CB-1 are usually very strong, and there is usually no method for manually operating the valve.

OCCASIONALLY, there are manual butterfly isolation valves upstream of the compressor bleed valves when off-line compressor water wash valving/system is supplied with a unit. If someone wished to brave the high temperatures of the turbine compartment, it might be possible to manually close these valves if 20CB-1 failed. _HOWEVER_, if the unit were to trip for any reason and IF the manual isolation valves were closed, it's very possible that the axial compressor would experience a stall or surge condition during deceleration, causing severe damage--so IT IS NOT RECOMMENDED TO MANUALLY CLOSE THE ISOLATION VALVES IN THE COMPRESSOR BLEED VALVE PIPING WHEN THE UNIT IS RUNNING (GENERATING POWER)!!!

The compressor bleed valve trouble alarms are usually generated by the limit switches, sensing the valve(s) are open when they should be closed, or closed when they should be open. The limit switches do fail, and they do require occasional adjustment (sometimes the nuts holding them in place do vibrate loose).

If the solenoid-operated 20CB-1 were to fail intermittently, it could cause nuisance compressor bleed valve trouble alarms. While it's NOT recommend operating procedure, on conventional combustor-equipped units the compressor bleed valves can be open when the unit is on-line and generating power. HOWEVER, power output will be greatly decreased (since a portion of the compressor flow will be diverted to the exhaust duct through the compressor bleed valve(s)).

In your description of the system, there is very low axial compressor discharge pressure during starting and acceleration--when it's very important that the compressor bleed valves be open. And, that's also why the compressor bleed valves are usually spring-loaded to open, and use air pressure to close, and why they are closed when the unit is running using axial compressor discharge pressure.

There are usually piping system diagrams, called P&IDs (Piping & Instrumentation Diagrams) or Schematic Piping Diagrams, depending on the packager of the turbine-generator. These drawings are CRITICAL to understanding the operation of the unit. If you haven't already done so, you should locate and study these drawings--they have LOTS of good information and can provide lots of clues to how the unit operates.

Finally, a servo-valve is usually a low-voltage control element, say +/- 10 mA or +/- 10 VDC device. In MOST (but NOT all) applications, the current/power applied to servo-valves is modulated to control position or flow-rate of an infinitely positionable device (such as a fuel control valve or modulated Inlet Guide Vanes). This author has never seen or heard of a servo-operated compressor bleed valve--though they may exist. (A LOT of different control schemes have been employed over the years, and while you say your unit has a Mk V, it may be a turbine control system retro-fit panel, applied to an older unit with some earlier control scheme.)

markvguy

Thank you, markVguy; your explanation is is fantastic. Now its clear to me.

What is the solution if 20CB1 solenoid is passing. The bleedoff valves will close earlier than 95% speed while the machine is under startup.

The solution? Have a spare 20CB-1 handy, then remove the existing 20CB-1 (the unit must not be running...) and try to clean or repair it. Quite often the valve packing gets "sticky" and/or the valve stem gets gummed up preventing the valve from transitioning to the "vent" position from the "pressurize" position, allowing the "passing" you are speaking of. It is believed the valve stem packing can be replaced.

This author has seen broken springs on a couple of very old valves (the springs used to open the valve when the solenoid is de-energized.) There is also an "arm" (mechanical linkage) to the armature of the solenoid which has been known to need some tender, loving care (also known as cleaning and greasing with an appropriate grease--the "mechanical department" should be able to advise on this...).

If that doesn't return the valve to normal operation, install the spare valve!

These valves, most of which are "buried" in the bottoms of the Turbine Compartments of Frame 5s and Frame 6s, require periodic maintenance--but almost never get any.... Several sites have relocated the valves to outside the turbine compartments, under the walkway somewhere where the indicators (YES--there is actually a position indicator on these valves, which can't be seen unless one is directly in front of the valve which is nearly impossible to do when the unit is running!!!) are easily visible and maintenance/replacement is made much easier. Requires some new conduit and tubing--but is well worth the time and expense.

Please write back and let us know what you find!

markvguy

Hi all,

I want to share my experience with the 20CB solenoid. On our three Frame9 GE turbines, the inside of those 20cb solenoids were totally destroyed by rust. This was related to the high humidity of the compressed air. So if the solenoid was actuated or not, the result was the same. We have replaced all these valves after 8 years.

Best regards,
carlos

Both "Mark V guy" and Carlos are correct. The 20CB1 was passing continuously. We had managed the situation temporarily and replaced the SOV. As explained by Mark V the location of the valve is most difficult to access. Hope GE will relocate the valve in future units.

Regards to all.

Thanks for the feedback.

If the unit is out of warranty, you are free to relocate the device to a more convenient location. As was mentioned earlier--many owners/operators have done so!

Of course, you don't want to relocate it fifty feet (16 meters) away from the current location, but putting it outside the turbine compartment, under a walkway, for example, is perfectly acceptable. The determining factor has to be the time required to VENT the pressure from the actuators when the bleed valves are commanded to open. If the tubing run is very long, it will increase the time required to vent the pressure to the point the valves will open.

markvguy

One more consideration: drainage of condensate. If you look at the Cooling & Sealing Air Schematic Piping Diagram, or P&ID (Piping & Instrumentation Diagram), you should see a small "bleed" orifice in the low point of the supply air piping to the solenoid-operated valve. Sometimes, by design, the low-point is the poro-stone filter upstream of the solenoid-operated valve. The purpose of this orifice is to allow a means to get rid of condensate from the line through a very small orifice. (Come to think of it, the internals of the valve should not rust if the bleed orifice is working properly.... This is one of those things which quite often gets overlooked during maintenance outages--they are very important, though!)

So, if you relocate the valve and there is a low-point in the air supply piping, make sure there is a bleed orifice in the low-point to get rid of the condensate.

markvguy

Hello, Mr. Mark V guy

You have mentioned that, the Bleed valves will open during deceleration to protect the axial compressor against stalling and surging!!!!!

Assume that the machine is running at full load hence, the Bleed valves are closed. Also assume that the packing of solenoid-operated valve 20CB-1 got sticky so, the solenoid valve remains always open supplying axial compressor discharge air pressure to the bleed valves independently of whether the 20CB-1 is energized by 125 VDC or not. Further assume that an emergency shutdown is initiated (due to detection of HIGH- HIGH Bearing Vibration, for example). Then, the machine starts decelerating with the solenoid-operated valve 20CB-1 fail to close and the Bleed valve remain in closed position while the machine is decelerating!!!!!!!!!
What will happen to the axial compressor in this case?

Is there any other alternative method to protect the axial compressor in this case?

Thank you

Hello, Mr. Anon Y Mous,

If the compressor bleed valves fail to open because the solenoid-operated valve fails to open when de-energized then there will be some fairly noisy deceleration--accompanied by high vibration and very sharp loss of speed. This author has experienced it twice, and both times the units (a Frame 7EA and a Frame 6B) survived with no detectable damage (no rubs, no rotor lock-up--nothing). Once, the unit was very throroughly borescoped and no damage was found whatsoever--except in the opinion of the Customer who felt that GE should provide a brand new axial compressor, even though the unit was still under warranty and if it failed after the borescope GE would have made good on the repairs, or replacement, as necessary. (This does nothing for the lost revenue, of course, but in GE's experienced opinion the risk was low and that unit has been in flawless operation for more than 20 years as of this writing!)

Without compressor bleed valves, units would continually experience potentially damaging stalling and/or surging during acceleration to- and deceleration from rated speed. This would most likely be very damaging if allowed to continue over time. So, compressor bleed valves are employed to eliminate--to the extent possible--the cumulative affects of continued starting and stopping without mitigating the stalling and/or surging.

GE is now implementing (through a TIL, Technical Information Letter, it is believed) a sequencing scheme which tests the compressor bleed valves and solenoid during each start attempt. It requires instrument air (since the normal source of supply is compressor discharge pressure). It's automatic, and increases the time from START initiation to FSNL/loaded operation. It does NOT test the solenoid-operated valve during loaded operation of the unit, which is when it's most likely to stick, but it provides a measure of security--as best as can be provided.

To this author's knowledge, there is no alternative scheme which is currently available to protect against the failure mode you have described (with multiple exclamation points!!!!). However, read on, please.

People can punch holes in protection schemes all day long until the cows come home to lay down and chew their cud--it's a fairly popular sport in many parts of the world, even commonly practiced by GE field engineers, though usually amongst themselves.

It's as simple as this: ENGINEERING IS A SERIES OF COMPROMISES. GE salespeople are continually being told that GE equipment is extremely expensive, and adding redundant protection schemes and devices--while adding protection--is costly: costly to engineer and costly to manufacture. GE has been doing lots of things to try to drive cost out of their equipment to try to provide a quality product at a competitive price--as have most manufacturers in every industry around the world.

This author has said this many times in this forum: Once the unit is out of warranty, the owner and/or operator is free to modify his unit and its control- and protection schemes as is seen fit. However, they assume (as in "take responsibility for") the liability for the consequences of their modifications--and so most Customers don't make modifications since they, and their insurance company(s), are highly "allergic" to risk--the risk that comes with having to fix some collateral damage which might occur when every possible contingent situation was not considered and something was damaged as a result of a "site-developed" modification.

One can ask GE to engineer and provide such schemes through their "CM&U group" (Conversions, Modifications, and Uprates)--but generally the cost is prohibitive, and most of that is because GE must add some amount of money to the price in order to cover potential risks associated with one-off and non-standard schemes, so most Customers opt not to purchase the schemes

And, this is where experience is most valuable--and why people pay for experience when hiring individuals or consultants, or choosing suppliers. Experience provides a level of risk-management that isn't possible with a lack of experience--the ability to recognize what is most important and critical and how much attention (monitoring and protection) the situation requires or demands.

Risk management--it is a very large component of engineering. Understanding what the consequences of failing to have redundant schemes are. This author knows of one site which had several compressor "events" during shutdown (the exact number was never determined, but it was more than three) because someone failed to check into the alarm "Compressor Bleed Valve Position Trouble" during shutdowns, related to the bleed valves not opening quick enough. Their compressor survived the operators just initiating a 'Master Reset' to be able to re-start the unit; and the problem was only discovered when a field engineer was on site during the shutdown prior to a maintenance outage and asked, "What the he!! was that?!?!?!" when the unit tried to shake itself off the foundation.... The operators responded that was "normal" and when asked about the alarms, they said the I&C department told them it was a nuisance alarm.... So, even with normal protection, one cannot protect a unit against operator error! And, the compressor was not even borescoped and the unit continues to run to this day (that is one SUPREMELY lucky Customer who needs to get some new operators and I&C technicians!).

Lastly, this author does not design axial compressors, turbines, generators, the auxiliary systems or the control and protection schemes used to monitor and protect the systems. He has, on many occasions, been forced to justify the schemes; some times it's difficult. But, there's no such thing as a free lunch and you can rest assured that GE employs some very intelligent and talented people who have used the Company's collective experience over decades of producing combustions and equipment to provide a level of control and protection that they feel minimizes their warranty costs, primarily, and provides their Customers with a reliable and adequately protected unit--otherwise known as risk management.

When considering the effects of not having redundant devices or being able to anticipate and protect against every possible failure, try to consider other factors, such as risk management and cost when designing and manufacturing engineered systems,

This author has never been at a site which routinely removes and inspects the 20CB-1 solenoid-operated valve--until they do so for a failure analysis! Someone at their site has determined--knowingly or unknowingly--that it's not cost-effective to do so. This is also risk management--of a sort--not spending money until a device fails.

The name is markvguy--no salutation required; this author works for a living just like most people on this site.

markvguy

Markvguy, I appreciate your informative posts. I have seen on Frame7 machines compressor bleed valves not opening fully during a fired shutdown and cause the unit to flame out/trip. It usually occurs on the upper valve, it will go about 50-75% open and stick. I believe moving the valves outside the enclosure will be a better environment for the valves and alleviate the stinking problem, but is there anything else that can be done or reviewed?

We once had a problem of malfunctioning 20CB. Our bleed valve continued in he close position even until we reach around 90MW and stalled there. Then we plug in plant air after the 20CB to simulate CPD. The bleed valve closes and we manage to go to base load.