PID ??? Tank level switch feed by a variable valve.

[kamenges]

Originally posted by ViperDaSnake:

The payback on this...

You really didn't tell us what you expect the payback to be other than personal satisfaction. What detrimental equipment, throughput, quality or economy issues are you trying to correct? If you are just changing it so your fancy new analog controls don't act in a digital fashion then the payback is simply personal satisfaction.

The push of this post is to make you do a cost/benefit analysis. If you can define what detriment you are trying to address, you can put a cost to the detriment and then attach a cost you are willing to pay for a solution. For example, does the high rate of make-up infeed flow to the boiler thermally stress it enough that you expect a fatigue failure? If so, how long before it happens and how much do you expect this to take off the life of the equipment? There is your detriment cost or, more correctly, your benefit point. If the cost of better control is much less than this then add the sensors you need and move on.

Keith

 

[TurpoUrpo]

Do you measure amount of steam and temperature of steam that is leaving the boiler?

 

[shooter]

nice problem.
use both switches to fully open and close your valve.
however if level in the middle lets assume 50 % valveopening.
the full level is hit. close the valve until level oke. the the opening should be 50%-10%=40%.
if level gets too high again close valve until level is oke then subtract again 40-10 is 30%
obvious same for low level.
now the valve will stay open for 30% ubtil any level is reached so the filling time is much larger.
this has a huge advantage on pressure control, you will see the the pressure much more constant.
please think of the feederpump if a pump is used this pump could use a VFD. same setup as above.

 

[rguimond]

Maybe I'm way off base here, but I think the amount of steam used should equal the amount of make-up water required. If you're willing to spend the $, invest in a really good steam meter and measure the amount of steam you're using. Add the equivalent weight of make-up water (another flow meter). Low and High level switches will act as safety switches.

Calculate the rate of usage and feather speed of make-up water supply pump to match.

 

[milldrone]

Maybe I'm way off base here, but I think the amount of steam used should equal the amount of make-up water required.

rguimond,

I believe you are half right. You also need to consider "shrink" and "swell". "Shrink" and "swell" refer to what happens to the water level when the flow of steam stops or starts.

When the boiler is steaming at maximum the water cavity is filled with a large volume of steam bubbles. If the flow of steam stops (quickly) the bubbles collapse and the water level "appears to drop". But there is still the same amount of water in the boiler. This is one reason why monitoring the steam flow only is not a good idea.

Now consider the opposite end of the operating spectrum. The boiler is at pressure but there is no flow. If the water level is anywhere above the minimum (instead of at the minimum) and the steam demand increases rapidly the water fills with bubbles and can go way over the top limit.

ViperDaSnake has not given us all the details, but I'm guessing that someone in the past at his facility has seen the result of high water levels in the boiler. He has indicated that the steam is used for heating so the results usually show up as slow heating or production loss. I'd even guess that he has a intermittent priming (priming is the term used by boiler people concerning water carry over into steam) issue. The production people have probably noticed it (slow heating) when the demand is high. The slow heating can curiously happen even though the boiler has plenty of pressure and may not even be burning at the maximum fuel level.

I'm not a PID expert, but I do know that when a boiler has priming issues related to shrink and swell it is best to use both analog level in the boiler and analog flow.