Control.com - Nerds in Control

Last outage we intended to do waterwash but failed dismally. Nothing would work. Is there any pre start checks that we can perform on our next one? Our outage will begin 4th Oct-15th Oct.

It is presumed you are talking about an off-line compressor water wash. It is also presumed you have a Water Wash Skid, with a large water tank with a heater, a smaller detergent tank, a Water Wash Pump, and some kind of control panel possibly including a small PLC. Most skids have an eductor valve to draw the detergent into the outflow of the skid, a low pump suction pressure switch, a low pump discharge pressure switch, and a pump flow switch. The Water Wash Skids provided with F-class units also have one or two solenoids for controlling detergent flow.

There is also usually some motor-operated valving located outside the axial compressor inlet ductwork. There may also be several manual valves which must be operated before and after the wash procedure.

There is usually some valving at the discharge of the skid which will allow the flow to the unit to be shut off and test flows to a drain to be established. You can put water in the detergent tank, jumper the input signal from the Mk VI to the skid that starts the skid, throttle the discharge valve to simulate water wash nozzle area, and run the skid to see if you can get it to work. This is difficult because of the cycling nature of the water wash procedure used for F-class units, but it can be done.

Usually there are several problems with the sequencing in the PLC--most notably, the flow switch keeps tripping the pump because flow doesn't get established "soon enough" for the sequencing as written in the PLC. Another major problem is the pump suction pressure switch keeps tripping the pump if flow is established too quickly. The pump suction pressure switch is just there to prevent damaging the pump if there's no water in the tank (a level switch would have been better, but it costs more...). Usually, as long as someone is in the Water Wash Skid to monitor the process who could stop the pump if the tank level was too low, this author just jumpered the suction pressure switch to prevent nuisance trips...sad but true.

The flow switch is a little more difficult because it has to be coordinated with motor-operated valve opening. If the valve opens too slowly and flow doesn't increase fast enough, the skid will be tripped. If the valve opens too fast the pump suction pressure switch will trip the process...kind if a timing thing. Keep a jumper handy at the flow switch also while you're learning to perform the procedure.

Last, review the sequencing in the Mk VI to learn what's supposed to happen when--and then monitor the procedure closely to determine what's happening when. This is not easy, but there have also been some glitches in the Mk VI sequencing...especially if the unit was not packaged by GE....

Your site should dedicate one person on site to be the Water Wash "expert" to study the system, mechanically and electrically, to work with the OEM (GE or ???) to understand the fine points of the process, and to get any sequencing problems worked out--this will take time. That person should then write a detailed procedure for everyone to use when performing water washes--one specific to your site and equipment.

And that's the most important thing about water washing--be prepared for the first few washes to take a LONG time while you work through the glitches in the system. It NEVER works the first time, or the second, or the third--and usually it takes a few jumpers and a thorough understanding of what's supposed to happen when to make it work properly.

Sorry; but that's the truth! If the unit is still under warranty, you should contact the OEM/packager to help with your issues. Be prepared to spend some time the individual--again, this is a complicated process to understand and troubleshoot--made more complicated by the static starter which must be used to crank F-class units.... Patience is a must, and time MUST be allotted for working through problems the first few times water washing is performed.

markvguy

To markvguy
Thanks for the advice.My apologies for the lack of information.
It appears that the wash will go ahead on 12/10/06.Apparently when the wash failed the last time round,we couldn`t get the GT to crank and it was linked to the HRSG Drums being empty.
Further to this,there are 3 permissives i believe when the GT achieves crank speed,these could not be met.

Changing the subject,as i am a new member of the forum i have been very interested in the discussions held and i have a few questions of my own.I hope you could be kind enough to help me.

1.What is meant by "No Reset"when dealing with PID Controllers?

2.Do GE have a glossary document for Speedtronic Mark VI?I have found your descriptions very helpful but in my quest to achieve a better technical understanding of Speedtronic as a subject,this would help me immensely.Unfortunately i was not present on the Speedtronic Training.

3. On our FSR control screen,what is the function of DWCK,CPR?I have noticed that CPR is at a reading of 128 even if it is on load or off load?
Many Thanks

So, which is it: kevine or kevinr?

Something sounds a little odd... On non-F-class units, the sequencing usually requires one to crank the unit and select Off-Line Water Wash ON BEFORE the unit reaches 14HM, but that didn't seem to be a requirement for F-class units (in other words, the sequencing was slightly different for the F-class units versus the rest of the fleet).

Again, the site should initially select one person to familiarize him-/herself with Off-Line Water Washing, learn the skid, the sequencing, the piping systems--everything. Then, that person should develop a site-specific procedure that anyone could use to perform the procedure, including some troubleshooting tips. The procedure should include an outline of what happens during the procedure, so that the reader can observe the process. It was common for the HMI displays NOT to work correctly--sometimes it was bad as the targets (buttons) not working properly, but usually included the "flow-chart" of the process not providing proper indications or no indication at all.

As for your other questions:

1) This would be a good question to ask all by itself separately in another thread if you want the best answer from the real control gurus on this forum.

2) Usually there is a file called LONGNAME.DAT in the unit-specific directory (or, just search the C: drive for the file using MS-Windows Explorer). That file contains the "best" description of the signal names. If you have other questions, ask them here--surely there are others who could benefit from the discussion.

3) DWCK is not a familiar term; have you tried to look it up in Toolbox? CPR stands for Compressor Pressure-Ratio, and is a fancy way of calculating the pressure drop across the compressor taking into account the barometric pressure. (The 96AP transducers MUST be working correctly for this calculation to function correctly! And it is critical calculation that is used for Exhaust Temperature Control!!!) If CPR is truly at 128 in the Mk VI, then there is something seriously wrong!

But, the better bet is that the CIMPLICITY screen is not configured correctly. Unfortunately, there are several types of CIMPLICITY displays (none of which this author liked!) so trying to describe how to determine what signal is "linked" to the value being displayed for CPR is very difficult. CIMPLICITY is NOT like most HMI software, in that with many programs one can hover the mouse over a value on a display and see what signal is driving it, or right-click on a value to determine what signal is driving it. Nope; not CIMPLICITY!

markvguy

markvguy,

Cimplicity has the ability to display tagname of point of item on the screen along with current value. GE has decided to turn this feature off in the Mark V HMI. If project is opened in Cimplicity workbench, in menu on left side, choose "Security". Double click "Roles" and put checkmark in "Dynamic configuration". This will allow you to right click on value on screen and see point name and value. GE also hides "Point Control Panel" in the GE meatball in screen header. Double click the meatball and a menu will popup, select "Point control panel". You can now add points that you want to see from the database.

So, you've already started your outage; unless you have someone from the OEM/turbine packager onsite who is familiar with the process and the equipment provided with the unit, you will most likely not be much more successful this outage, either, unless you just manually perform a simple water wash.

You haven't provided any details about exactly what went wrong so it's difficult to say with any certainty what to look out for. Also, there is a lot of variation in the exact equipment provided--if it was provided with the unit, as some Customers/architect-engineers opted to build or obtain water wash systems from someone other than the OEM/packager.

It is presumed you are talking about an off-line compressor water wash. It is also presumed you have a Water Wash Skid, with a large water tank with a heater, a smaller detergent tank, a Water Wash Pump, and some kind of control panel possibly including a small PLC. Most skids have an eductor valve to draw the detergent into the outflow of the skid, a low pump suction pressure switch, a low pump discharge pressure switch, and a pump flow switch. The Water Wash Skids provided with F-class units also have one or two solenoids for controlling water/detergent flow.

There is also usually some motor-operated valving located outside the axial compressor inlet ductwork. There may also be several manual valves which must be operated before and after the wash procedure.

There is usually some valving at the discharge of the skid which will allow the flow to the unit to be shut off and test flows to a drain to be established. You can put water in the detergent tank, jumper the input signal from the Mk VI to the skid that starts the skid, throttle the discharge valve to simulate water wash nozzle area, and run the skid to see if you can get it to work. This is difficult because of the cycling nature of the water wash procedure used for F-class units, but it can be done.

Usually there are several problems with the sequencing in the PLC--most notably, the flow switch keeps tripping the pump because flow doesn't get established "soon enough" for the sequencing as written in the PLC. Another major problem is the pump suction pressure switch keeps tripping the pump if flow is established too quickly. The pump suction pressure switch is just there to prevent damaging the pump if there's no water in the tank (a level switch would have been better, but it costs more...). Usually, as long as someone is in the Water Wash Skid to monitor the process who could stop the pump if the tank level was too low, this author just jumpered the suction pressure switch to prevent nuisance trips...sad but true.

The flow switch is a little more difficult because it has to be coordinated with motor-operated valve opening. If the valve opens too slowly and flow doesn't increase fast enough, the skid will be tripped. If the valve opens too fast the pump suction pressure switch will trip the process...kind if a timing thing. Keep a jumper handy at the flow switch also while you're learning to perform the procedure.

Last, review the sequencing in the Mk VI to learn what's supposed to happen when--and then monitor the procedure closely to determine what's happening when. There have also been some glitches in the Mk VI sequencing...especially if the unit was not packaged by GE....

Your site should dedicate one person on site to be the Water Wash "expert" to study the system, mechanically and electrically, to work with the OEM (GE or ???) to understand the fine points of the process, and to get any sequencing problems worked out--this will take time. But, the time will be well spent.

That person should then write a detailed procedure for everyone to use when performing water washes--one specific to your site and equipment.

And that's the most important thing about water washing--be prepared for the first few washes to take a LONG time while you work through the glitches in the system. It NEVER works the first time, or the second, or the third--and usually it takes a few jumpers and a thorough understanding of what's supposed to happen when to make it work properly.

Sorry; but that's the truth! If the unit is still under warranty, you should contact the OEM/packager to help with your issues. Be prepared to spend some time the individual--again, this is a complicated process to understand and troubleshoot--made more complicated by the static starter which must be used to crank F-class units.... Patience is a must, and time MUST be allotted for working through problems the first few times water washing is performed.

If you can be more specific about the problems you experienced we can discuss things in more detail. It would be difficult to share the .m6b file as they are large and sometimes unable to be read without the same version of Toolbox used to create the file, but it would not be impossible to discuss the sequencing.

Again, assign one person to become very familiar with the system; allow that person the time to read all the available documentation and assemble all the schematics, P&IDs, and drawings and develop an understanding of the system and what it's intended to do; then allow that person the time to observe what is/isn't working when you try to perform a water wash. He/she will be able to help with the problems at that point and develop a site-specific procedure which should be very easy to use.

Again, you will most likely require someone from the OEM/packager to help get rid of all the bugs in the sequencing....

markvguy

Of course, compressor washing is required so to clean the compressor and take the output to the level desired.

Online washing is typically not recommended; therefore, the type of washing discussed here is offline.

But do we really have to take the machines off so often?

Well, if we install HEPA air intake filters (HEPA refers to the high efficiency grade), then the compressor remains clean and there isnt really a necessity to wash for a 1 year or 2.

gdearcangelis@tteur.com