PID for damper position control

geniusintraining

Lifetime Supporting Member + Moderator
Join Date
Jun 2005
Location
SC
Posts
8,242
Damper control: air to air heat exchanger, pre-heater

We call this a block and bypass dampers, I have process air coming into this preheater at 200f leaving @ a set point of 500f, the means of heating this air is the same air has to be incinerated @ 1400f, so we take a portion of this hot gas and recover the heat to preheat the incoming air

I have a few PID controllers that I’m having issues with at best I can only get +/- 40deg variance, so I am going to “roll my own”

They are controlled by a forward and reverse motor starter, so just cycling/pulsing the contact to get the proper damper position, they both have (now) a 0-5vdc position feed back to the plc

They also need to be in the same mirrored position, due to creating back pressure, if one is open 20 % then the other needs to be closed 80% +/- 5%



Here’s what I plan on doing,

Set up a PID to control a timer preset, depending on how far out of span the PV is compared to the SP, we will increase or decrease the timer preset, but I also need to put a max rate of rise 25 deg per/min

Then slave the other damper with a PID loop, looking a actual position for the other PID control

So what do you think? is there a better way? the controllers that I am using now are no longer made and there is no info on them, I have tried for several days to tune these, but no luck so keeping them is not going to happen.

Damper 116 and 117 are first then if it works I'll go back to the oil heating damper

4000.jpg
 
I would have a outer PID control the temperature by setting the positions for both damper PIDs. I think the outer 'PID' will be a real challenge. Getting the balance between the two dampers will be tricky enough. Tuning the output of the outer loop will be tricky if the is much of a dead time or a difference in the dead times between the 116 and 117 legs.

There are too many variables to tune by hit and miss. This takes a plan, data and some math. I would find what ever info I can before tweaking. A theory of operation and some why things were designed this way would be good.

Ramping up the SP is the easy part. This is not a simple SISO system.
 
Thanks Peter,

I think I understand what you are saying, I will give that a try, would you still use the output from the other two PID's to control a timer and pulse the contact that way or would you drive to a given position?

The dead time will be a factor, for other reasons, the speed of the two dampers are not close, thats the reason I was going to use the second one as a slave, and just mimic the movement/position

Thanks again,
 
Thinking out loud :-
Could you not get rid of the blue damper and install a weighted damper to control backpressure ? also , one damper 80% open + another 20% open will not give a net result of 100% opening - much more to take into account , the back pressure of the heat exchanger etc , not withstanding that there is nothing linear about the airflows through these dampers - the volume of the gas passing through the green damper will be less than the blue damper in any case , so the opening ratio of one damper to another won't be 1:1 .

Could you not run through the heat exchanger at all times , and control the make up air temperature with a fresh air blending damper fed from the same fan , but on the air off side of the heat exchanger - this certainly could work .

I think if you can get rid of one of the dampers , then your project will be much simpler to implement .
 
Thinking out loud again - is the chimney natural draft , or is there a booster fan ?
You might make a little test , leave the blue damper wide open and see what air off temperature you can get from the heat exchanger . If the heat exchanger is big enough , then you can modulate the green damper only to obtain control . I personally like the idea of doing away with the heat exchanger dampers altogether , and obtaining temperature control with a blending damper .
 
10BaseT. said:
Thinking out loud :-
Could you not get rid of the blue damper and install a weighted damper to control backpressure ? also , one damper 80% open + another 20% open will not give a net result of 100% opening - much more to take into account , the back pressure of the heat exchanger etc , not withstanding that there is nothing linear about the airflows through these dampers - the volume of the gas passing through the green damper will be less than the blue damper in any case , so the opening ratio of one damper to another won't be 1:1 ..

That would take a lot of engineering and much more cost, this is a working system (just not as efficient as I would like them)



10baseT. said:
Could you not run through the heat exchanger at all times , and control the make up air temperature with a fresh air blending damper fed from the same fan , but on the air off side of the heat exchanger - this certainly could work ..




No, the more air I run, the more gas I will burn to achieve 1400deg SP, the cost to run this system is 150k USD per-month thats my savings, I just replace the dampers (116 and 117) they were 30k each (plus labor), but I have cut the natural gas usage by 30%, but still looking for more



10BaseT. said:
I think if you can get rid of one of the dampers , then your project will be much simpler to implement .

Yes but it would not be as efficient, when you run the numbers and you calculate the back pressures you need the all the dampers, I have added some static pressure switches I will also use these to help my final calculations/tuning

Remember that the less air the better, I have to run at 1400deg, but I want to recover that heat and use it, I will on a normal day I run around 15,000scfm

But thank you for your suggestions
 
I think you may be complicating your life a little more than it has to be with the inner loops on the dampers. Unless these are something different that what I think they are the dampers will maintain their position if the motors are not run. If this is the case I would just use a simple hysteresis control with a shift opposite the direction of travel to account for drift as the motor stops. Putting a PID regulator on something driven by a reversing motor starter probably won't get you much other than a burnt out motor.

As Peter said the outer loop is the hard one. If I were doing this I would try to get some data on exactly how damper position influenced heat exchanger temperature. Use this data as a damper command feed forward so you get the damper positions close before you even start doing any control. Then let your PID tweak the master damper position to get what you want.

Are you confident of your master/slave damper relationship and that this relationship holds over a broad range of damper positions? I could see this as a sticking point if this relationship isn't constant.

Keith
 
I wondered if there was a piggy back fan for the exhaust draft .

Please note that using an air blend damper doesn't decrease your efficiency - you still provided exactly the same amount of combustion air to maintain the stoichometric ratio , some goes through the heat exchanger and some bypasses it , the M101 combustion air fan does the same amount of work .
OK , this is a big lump of kit , but again , what happens if you run the thing with the blue damper only , controlling the amount of flue gas running through the heat exchanger using the green damper . This heat exchanger probably has a huge lag which would mean you have to tune the loop very dead to get it close to setpoint , and it will take ages to do it .

You mention that the two damper speeds are different , so in normal operation the back pressure changes - that is why I suggested just using one as a mid point diverter if you like to vary the volume of flue gas through the heat exchanger - the back pressure would stay fairly much constant , being dependent on the final exhaust stack if there is no fan on the exhaust .
You can test this in five minutes by forcing the bypass damper slowly open , and leaving the blocking damper to control . Another thing you should be aware of is that in many applications , dampers such as yours would have 25% free area to prevent the exhaust from being obstructed if there was a failure - if this is the case , you are fighting a bit of battle to get temperature regulation with them - certainly worth checking - I have done loads of systems with safety dampers , and the airflows are all over the place after they are opened more than about 15% (by angle) , and looking at your screen shot , this would also appear to be the case with your dampers . The bypass damper is nearly shut , and the blocking damper wide open , and yet you are still overshooting the SP .
(A load of assumption on my part regarding the screen shot ! )

Do you consistently overshoot or undershoot , or oscillate around the setpoint ?

If you wack the controller into manual , can you obtain a stable temperature ? and could you for test plot damper position versus air off temperature ?
 
10BaseT said:
...note that using an air blend damper doesn't decrease your efficiency...

Agree but the more air I send in the more air I have to incinerate (heated up to 1400deg f) this is done with a natural gas...look at the whole screen shot, I have two processes dumping into a collection chamber, this collection chamber is 0 static pressure set point, achieved by the ID fan, right now only one of by process running, so blending means more and more means more gas consumption


I have done several DOE's and I have data coming out my ***, I have forced dampers closed/open to see what could happen

I have forced 117 open and modulated 116 but I could never achieve set point (due to path of leased resistance) but I also have to becareful of over-temp and over pressure, over-temp could and does happen if I force 117 anything other then open, if needs that by-pass route to cool, now I have limited the amount close that it can travel, this helped reduce the over pressuring and gave a slower rate of rise


Keith said:
I think you may be complicating your life a little more than it has to be

This would no different then my whole life, but it makes it fun.

Keith said:
Unless these are something different that what I think they are the dampers will maintain their position if the motors are not run.

Yes this is how they work now, close close, open open, shorter longer pulses, this is depending on temperature SP (TIC114) and PV (TE114) each damper weighs 1500lbs, they will not move/drift without the motor/gear box

Keith said:
If this is the case I would just use a simple hysteresis control with a shift opposite the direction of travel to account for drift as the motor stops. Putting a PID regulator on something driven by a reversing motor starter probably won't get you much other than a burnt out motor.

Now, they cycle 100 X hr (aprox) for the last 10ish years no troubles. but thats what I was trying to do Keith, just put a large deadband to try and hold +/- 10deg

Keith said:
Are you confident of your master/slave damper relationship and that this relationship holds over a broad range of damper positions? I could see this as a sticking point if this relationship isn't constant.

I have a 0-5vdc position feed back, using this from the master and try and control the slave

Thanks guys this is the conversation that I need to have

Edit: the screen shot
40001.jpg
 
Last edited:
USE 2 INDEPENDANT PID's

This application looks to me like a classic COARSE control
FINE control application. Two different outputs each effecting
the same process variable.

I control similar processes and use 2 totally independant
PID controllers. One controller simply controls the Process
Variable whilst the other is used to trim the first controller.
You can use the Output of the first controller as the PV for
the second controller with its setpoint set to say 50% it will
ensure that (if it is possible) the first controller remains in
control at all times.
If you don't like the idea of using the output of one controller
as the PV of the other then the second controller can usually be
set up to control an intermediate PV. In your case it looks like
the exhaust or stack pressure would be a likely candidate.
 
Last edited:
Thanks John and others,

I will be off on holiday until the end of next week, this should give me some time to think about it when I'm out fishing :D

I will let you know what I did and how it worked or did not after the new year

Merry Christmas and Happy New year to all of my friends at PLCS.net :D
 
Happy holidays to you too !
One last thing for you to contemplate
"Agree but the more air I send in the more air I have to incinerate (heated up to 1400deg f) this is done with a natural gas...look at the whole screen shot, I have two processes dumping into a collection chamber, this collection chamber is 0 static pressure set point, achieved by the ID fan, right now only one of by process running, so blending means more and more means more gas consumption"

That is the whole point ! you do not change the amount of air you use as combustion air - the fan passes the the same amount of air , but instead of passing it all through the heat exchanger , you split the ducting before the HE , and again just after it , a diverter control the ratio of hot to cold air .
This is the way it is done with unmodulated steam heater banks , bypass the heater to obtain temperature control .
The way you have it is a nightmare - increase the process air volume , the flue gas dilutes and cools , air off temperature from the HE drops , dampers open to control it , flue gas temperature rises again - you need something constant .
 

Similar Topics

I have S7 1512C controler for controlling 48 PID temperature loop, the output is PWM. Please I need the best, most efficient way to write the...
Replies
5
Views
212
Hi all, I'm having trouble solving a problem I've been working on for several months, and thought you might like a stab at it. The machine runs...
Replies
22
Views
840
How can I connect PID Output to a valve. In ladder logic program is there any logic do I want to add between valve and PID? PV=SP What will be the...
Replies
7
Views
383
I'm setting up a PID for a 50hp RO pump in a water treatment plant. When I go to manual mode on hmi I set the swm bit, I also ons the pid output%...
Replies
5
Views
523
I am setting up control for Hypochlorite dosing. The easy part is the dosing calculation for flow pacing but I would also like to setup trimming...
Replies
8
Views
889
Back
Top Bottom