PID question

[Vidmas30]

Good day,

We got heating unit where v570 controls the condensate boiler (temperature PID loop). First issue was, that when we started to tune the unit, heating system was closed, left working only hot water boilers, so it's small contour. Because of that, PLC couldn't finish PID autotune process, because system heats up quickly and temperature after second wave, stays above setpoint very long time.. So we tuned it manually. The question is: beacause of the same reason (fast heating up) PID block status shows integral wind-down message (when there's no hot water usage). What should we do about it? Is there a chance that because this wind down the integral error will accumulate and PID block stops working? Should we use force integral error block to reset this error? if so, when it would be best to do this? Or maybe someone could share better solution? Thank you!

[Emil]

Hi,

I'll be glad to see log files from PID server with data from the process!

Don't forget to add the block "Read Contgrol components with factor 100 and log CVp, CVi and CVd too...

[Vidmas30]

Good day, Emil,

Yesterday played a bit with PID, logged data with PID server, but in the end of the day my laptop (left for some time..) hybernated and when came back PID server crashed and all the data gone.. So I have nothing to show.. Good news is that all went well and now PID works good. Although, there's one thing that bothers me:when changing SP a large step (for example valve CV 50% and when SP lowered CV becomes 0%) valve closes to 0% but there is some wave of CV near the end I mean CV 0% valves positinion goes to 0% and near the end CV 16% then 4% then0% (like some wave). It made a valve to go crazy couple of times.. I think it's not of the sygnal, but of the PID. All, except this issue, works well. (PID parameters P-50.0; I-27; D-37.)

[Emil]

Thank you for the feedback!

... That's why I wanted PID server log. It will let me see what you're explaining, but with more details.

What I can guess:

To prevent strong undershoot in case of changing setpoint (case you describe) the algorythm frces Integral error before reaching below setpoint value. It seems in your case PV drop was too slow and PID with the parameters you describe is quite "agressive". As result, probably PID recalculated the values and returned back to 0.

What you can do? Make PID less aggressive?! Maybe there is a set of good PID parameters, which will work very good with your system, but more "calm". Then, there is a great chance you will not see the behavor you describe (again - it's what I'm gessing on blind!).

How to make PID more "calm" - increase both PB and Ti and, maybe decrease Td.

I hope my tips can be helpful.

[Walkerok]

Hello All,

I have been watching this post and waiting to see where the conversation went. I have done industrial hot water heater and boiler controls for a long time, but I am not understanding your process good enough to offer any advice, so am asking if you could just describe what is happening physically (thermodynamically) a little more.

Are you injecting condensate water into a DA tank to be reused by the main boiler or is this a completely closed loop system?

Is your source of heat a burner or steam injection?

How much condensate do you return in percent will be fine?

What is the size of your main boiler?

How long are your low demand times or all the way off in a standby mode?

Since you are using a modulating control system what is the control mechanism and what is its travel time from low to high?

It sounds like you have a closed hydronic heating loop for building heat, but I am unsure.

Thanks

Keith