In our GE Frame-6B & 9E machines compressor bleed valves are having the limit switches to indicate the valve position. In frame-9E we are receiving bleed valve trip from the top two bleed valve limit switches due to faulty switches. It is suspecting that temperature affect the switches as bottom two limit switches are working normal for long time.

Is any site using other than the below,
Mfr:Crouzet,
Model:SP 3969 -10
GE # 328A7435P001

or how will be the response if micro switches used in place of limit switches.

Moreover the price also too high around 3K USD/ switch.

take care
g.rajesh

 

G.Rajesh,

You are free to use any switch, including a non-contact proximity switch, as long as the switch meets the zone/area classification requirements, and the voltage rating (normally 125 VDC, sometimes 24 VDC)--and, of course the temperature--always the temperature. If you can adapt a suitable switch to the application you are free to do so--as long as the unit is not under warranty or some kind of long-term service agreement which precludes purchasing parts from third-party providers.

I mention the voltage rating because there are LOTS of different switches and micro-switches that would work for the application--but many of them are not rated for 125 VDC service. There is only micro- to milli-amperes of current flowing through the contacts when closed, but they must still be capable of the voltage level.

And, most GE-design heavy duty gas turbines provide field devices and instruments for Class I, Div. II, Group D classification (sorry; I get my Ex and EXd and EExd and EXXd all confused)--check your local or regional area requirements.

I've seen many sites replace these particular limit switches with proximity switches (non-contact) and they work VERY well and last a LONG time. You need to find ones, again, rated for the required voltage and temperature. Do be aware though--that some Speedtronic turbine control panels (Mark IV and Mark V) don't work well with every proximity switch--some have a very high leakage current which the Speedtronic determines as continually closed. Most Mark VIs and Mark VIes work with most proximity switches--but it's still probably best to buy one or two as a test before buying as many as you need only to find out they have too much leakage current.

To use proximity switches it may be necessary to fashion some kind of suitable attachment/mounting bracket, but I think you'll find it very worthwhile in the long run. Just be sure to take the unit vibration into account when making the bracket(s).

It's always something, isn't it?

 

Good advice CSA,

If one has a MK5 or below, a positive 125 VDC contact is needed.

I have found that the switch is sometimes NOT the problem. the wiring from the switch to a conduit connection near the "split line" is problematic.

I have found "cheapy" terminal strips in the conduit connection.

nothing more than an observation.

 

Thank you CSA once again.

Yes agreed, we have noted your valuable suggestions, we'll workout towards them.

take care
g.rajesh

 

I would also suggest you make sure that the problem is with the switches and not with the valves themselves. The two top Bleed Valves tend to get stuck sometimes, specially after a longer period out of operation.

You should also check the wiring, the calibration of the switch, if the tongue of the switch moves freely (if not, just spray some WD40 on it and move it back and forth several times until it springs back freely to neutral position), and check the proper installation of the mounting bracket. Sometimes the bracket becomes loose and the valve pushes the whole body of the switch, not only its tongue.

Oh... and, if your installation is similar to what I've seen, check the small junction box next to the switch. Although high temperature resistant, due to high temperature, sometimes the terminals in those JBs become very brittle and break due to vibrations.

 

In frame 9e it is a common problem. All three components are problematic. Bleed valves stucking, 20CB-1 seat passing as well as limit switches adjustment gets out. Check all the 3.

GE also has raised some TIL on relocation of bleed valves outside the turbine compartment. Whats the arrangement there at your site?

Regards,
Owais

 

Owais,

Our site bleed valves are inside turbine compartment and we too heard it can be placed over the turbine compartment.

Our issues mostly with limit switches malfunction and few cases valves struck. We are using lubricant for valves to improve its performance.

take care
g.rajesh

 

g.rajesh,

I've worked on quite a few GE-design heavy duty gas turbines, and most of the sites that have problems with compressor bleed valve <i>actuators</i> use axial compressor discharge air for the pressure source. AND, most of them don't look after the poro-stone filter or the low-point drain orifices (also known as a continuous blow-down orifice) in the lines, and most don't do proper valve line-up during off-line water washing. Humidity and moisture making their way into the actuator is the biggest cause of problems. Unfortunately, most packagers, including the OEM, use a poro-stone filter which is made of carbon steel--and while they put a low-point drain orifice on the bottom of the poro-stone filter rust forms and collects in the poro-stone filter and eventually blocks the orifice in the low-point drain.

And, when there's rust in the lines mixed with moisture/humidity and that's applied to the actuator, well, the results are pretty predictable.

Another cause of problems with bleed valve actuators is that they are not rebuilt properly--meaning the alignment of the various components isn't always proper which contributes to sticking, especially when hot.

And, finally, the air pressure regulator which is usually provided (not always) in the supply line to the bleed valve actuators is usually not working, mostly because of moisture in the lines and lack of maintenance.

The best thing one can do is to use a source of clean, dry instrument air for actuating the compressor bleed valves. If the plant has an instrument air compressor, the amount of air required to close the valves is relatively minimal, and it's only a short "blast" of air required to fill the actuator. Other than that, there's no flow, so it's really not much of a load on a plant instrument air system.

Moving the bleed valves outside the compartment--usually to the roof--results in another maintenance issue. They have to be removed to remove the roof for maintenance. That means tubing and conduit and wiring.... And, most mechanical people are quick to make the modifications--and then just as quick to complain about the extra work during maintenance outages (because they didn't think about it before-hand).

If you're burning-up limit switches, well, that doesn't seem exactly correct. Either the wrong limit switch mechanisms are being supplied (wrong temperature rating) or there's something amiss with the compartment cooling/ventilation. Another thing that happens quite frequently is that someone (usually someone in the mechanical department) unilaterally decides to rotate the valves to make them easier to access--which can reduce the air flow in the compartment for "cooling" the actuators/valves.

Lastly, poor adjustment of the valves during refurbishment, and poor installation of the valves after a maintenance outage can result in mechanical binding of the valves and a failure to completely close. Most GE-design heavy duty gas turbines of the B/E class only use limit switches to indicate full open--since that's the position they need to be in during start-up and shutdown to protect the compressor. If the valve isn't fully closed during normal running operation, then air flow through the valve can add to the heat of the valve and actuator mechanism which can cause further mechanical binding and sticking. Improper installation (including the unilateral decision to rotate the valves to make them easier to remove and install) can cause mechanical binding resulting in a failure to completely close. Small leaks won't always be recognized in a lower-than-normal CPD, so they're hard to detect. But, hot air flowing through the valves at running speed is NOT good.

Oil is good to use on the actuator mechanisms but there are high-temperature greases that can be used and don't require such frequent application. Most of the sites I've been to that use oil require someone to lubricate the actuator mechanism prior to a start--and then someone has to lubricate the actuator mechanism prior to a normal shutdown, when the compartment is HOT. And, usually, the "roller" in the actuator mechanism (some of the actuator mechanisms, anyway) is a bearing, which can seize over time.

Recognizing the requirement for clean dry air, making sure actuator mechanism alignment is correct after refurbishment, and general good maintenance practices are usually all that's required to keep the valves in good working order. If the poro-stone filter can't be properly maintained, and the low-point drain (continuous blow-down) orifices can be kept open, then the first thing to do should be to find a suitable source of clean, dry air for the actuator. The only reason CPD is used is so that a stand-alone air compressor doesn't have to be supplied by the packager, or the EPC. BUT, using CPD means people have to understand the need for proper maintenance and care.

And, remember--while it seems to be a good idea to relocate the valves to the roof of the compartment, now there's rain and sand and dirt and other things which can contaminate the actuator mechanism.

Please let us know how you proceed--and how you arrived at your decision (to move or not move).

 

Improper installation (including the unilateral decision to rotate the valves to make them easier to remove and install) can cause mechanical binding resulting in a failure to completely close. Small leaks won't always be recognized in a lower-than-normal CPD, so they're hard to detect. But, hot air flowing through the valves at running speed is NOT good.

Could you comment further regarding techniques to detect CB valve leak through at load?

 

Curiousone,

When the compressor bleed valve is closed, there should be no flow in the piping--upstream or downstream of the valve. Therefore, the temperature of the piping should be no more and possibly even a little less than the ambient temperature in the compartment when the turbine is running (for those valves mounted in the turbine compartment). Proper use of an infrared heat detector can show that compressor bleed valve inlet and discharge piping temperatures are higher than their surrounding--indicating flow through the valve. This is about the simplest way to detect leaking compressor bleed valves--by properly using an infrared heat detector. (I keep saying "properly using" because just pointing an infrared heat detector at something for a couple of seconds isn't going to produce a meaningful result--unless the holder has experience taking this specific measurement at this specific point (or points) when trying to determine temperatures. Many times, this can lead to a lot of wasted time, after one or two quick measurements are taken and some assumptions are made based on those measurements without analyzing the measurements to see if they're within the realm of acceptable readings for the system.)

I have also installed T/Cs on the compressor bleed valve inlet and outlet piping--and used them to monitor the temperatures to check for leaks. This is actually a better method, because periodic readings can be taken and recorded/trended to establish normal operating temperatures to use to compare against abnormal temperatures. I've used spare T/C inputs to Speedtronic turbine control panels for this in a couple of cases, and just used the T/C connectors and a VOM to measure, record, and trend the temperatures on a periodic basis to get a profile to use.

Those are about the only ways I know of.