We have 2 GE 6B dual fuel Ct's (gas & fuel oil) DLN units with MARK V controls. Our issue is that while on fuel oil and at the point where water injection is called for the MARK V has 3 seconds to see flow. This is a control constant point (k4wnz). The transmitter is a 3 element and we have denim water, pump is operating, stop valve is opening. Check valves are fine.

Question: What value is acceptable for k4wnz? We want to increase that value. We will be contacting GE, just wanted users input.

 

In past posts you've said the units rarely run on liquid fuel. So, one has to wonder if this is just a result of lines "draining" down and having air pockets develop between liquid fuel operations. This is really a difficult question, and situation.

Those units have been there for years; is this something that just started recently? Or is it that the units are being run on liquid fuel more now than in the past and it's become a problem because they are can be or are started and run remotely?

One thing you probably already know is that usually when water injection is started it's typically "early" and results in NOx levels below guarantee for a short period.

The timer should not start until the stop valve opens, and once the stop valve opens water injection should start fairly quickly. Any delay should just be filling lines from the skid to the check valves and spray nozzles.

Do you have any sense of how long it actually takes for stable water injection to be established? I mean, if it misses on the first try, does it make it on the second try? Or the third?

I would think a value of no more than 4-6 seconds or so would be reasonable for a unit with a variable speed pump. Because the main concern would be that the pump would ramp up to a high speed and then water injection would start suddenly and blow the flame out, tripping the turbine.

And, in thinking about this even if the pump was a fixed speed pump with a variable control valve, the concern would be the same: not having the flow really high when it finally starts and then blow the flame out and tripping the turbine.

So, I think you should definitely ask this question of GE, but you should also provide them some data from your site. Run some trends and try increasing the timer to four seconds, or five seconds, or six seconds. Record flow rates (from the three flowmeters), and stop valve opening, and flame intensities, and exhaust T/Cs (to see if spreads develop, and if they do, if they always develop in the same combustors). As much data as possible, and then you can really make an assessment on your own and evaluate any response that GE gives you.

But, if the units rarely run on liquid fuel, then be aware that you just might not ever get one single value that works every time.

Lastly, there is usually some period of time for start-up that most permits allow, and as long as you have a few minutes from the first water injection attempt to try again, and once the lines get filled the second attempt will be successful and you will still be within the limits of the permit.

Hope this helps! It's really not a simple problem, unless there's something that can be done to ensure there is no air in the lines, and that's pretty hard to do. Or, is it that it takes the variable speed pump that long to ramp up, or the control valve that long to open? Again, the concern is not to wait too long for flow and have too much flow very suddenly and blow the flame out and trip the turbine.

Again, we don't know if it's air in the lines, or the ramp rate of the variable speed pump or control valve.