Help with a control system

I've taken what you guys have suggested and thought of the following solution.

How about if i have PID 1:
Setpoint: Pressure
PV: Pressure sensor
CV: Airflow rate (min and max limited)

PID 2:
Setpoint: CV from PID 1
PV: Airflow from blowers
CV: % speed of blowers

Do you think this would work? Would tuning be difficult?

Something I see to have forgotten to mention earlier is that I have 6 blowers connected in parallel, feeding 4 valves which are controlled by their own DO sensor PID loop.
 
I've taken what you guys have suggested and thought of the following solution.

How about if i have PID 1:
Setpoint: Pressure
PV: Pressure sensor
CV: Airflow rate (min and max limited)

PID 2:
Setpoint: CV from PID 1
PV: Airflow from blowers
CV: % speed of blowers

Do you think this would work? Would tuning be difficult?

Something I see to have forgotten to mention earlier is that I have 6 blowers connected in parallel, feeding 4 valves which are controlled by their own DO sensor PID loop.

I think I sketched this idea below; we'll call this the camel - "A control system designed by (customer) committee" (and hope they don't read this forum;)). I assumed four control valves means four tanks, but I don't think much changes if there is one tank and all four control valves receive the same control signal.

My thoughts, for what they are worth (disclaimer: this is far from professional advice):

  • Since the response (gain) of total flow to pressure will be dependent on situation- and time-varying positions of the four DO control valves, I think you are right to wonder if tuning will be difficult.
  • That said, an algorithm that looks at the four DO control valves and lowers the pressure until one of them is over say 80% open might be straightforward; perhaps it is a PID itself?
  • I wonder if this approach over-constrains the system, at least if either flow limit is hit:
    • If the four DO control valves pass less total flow than the lower flow limit, then the pressure will increase as the speed control increases the fans' speeds to fight for the lower limit, and the DO control valves will pinch down even more, until the fans are running at 100%.
      • The valves win here i.e. the actual total flow goes below the lower limit.
    • If the four DO control valves want more total flow than the upper flow limit, then they will go wide open as the fans are ramped down and the pressure drops.
      • The upper flow limit wins here i.e. the actual total flow does not go above the upper limit.
      • However, if there are four separate tanks, one per valve, then any of them might go well over their share of the total limit.
    • So in the end this approach does not meet the original goal of upper and lower flow limits
      • I really think the only way to put limits on flow is
        • to cascade a DO-controlling PID output to a flow PID,
        • to have that flow PID output to the control valve position, and
        • to put limits on that DO-controlling PID flow output
      • The pressure control is a canard: if they have an algorithm setting a pressure setpoint via fan speed to open the valves as much as possible, that algorithm could just as easily work directly from valve position to fan speed; cascading through a pressure control will only introduce a delay and make the system harder to tune.

xxx.jpg
 
Last edited:
[Update: I take it back: models range from 13 to 19 variables and 8 to 36 processes. a model would be non-trivial]


One more idea: it might be straightforward to write a coarse first-order, but reasonable, simulator for this process, and from there to experiment with different control schemes, tunings, etc. If convincing the customer that taking a different approach (e.g. drop the pressure control for direct speed control) is a feasible option, then that would be the way to do it.


I worked for a large petrochemical company in the early 80s. We wrote a simulation for a process unit for refinery on a VAX/VMS system, connected to a physical Honeywell TDC-2000 (-3000?) for operator training. The unit had been designed but not yet built, so we were hoping to get some feedback about how accurately the simulator responded (we were ivory tower, so few of us had never seen one of these beyond the differential equations involved; the visitors were real-world folks). I remember we were quite amused when, instead of giving any feedback, they started twiddling the controls, watching the simulator respond, and discussing among themselves what it meant for future operating procedures. I guess they had never seen the real thing either!
 
Last edited:
I apologise for the late reply.

With regards to:

If the four DO control valves pass less total flow than the lower flow limit, then the pressure will increase as the speed control increases the fans' speeds to fight for the lower limit, and the DO control valves will pinch down even more, until the fans are running at 100%.
The valves win here i.e. the actual total flow goes below the lower limit.

On any specific pressure setpoint, the valves have min opening (which are set during commissiong) - I am hoping the system would reach steady state eventually if the minimum flowrate is required.
 
Drbitboy is pretty close. If you have access to any EPA documents, the Fine Pore Aeration Manual has some god info:
https://nepis.epa.gov/Exe/ZyNET.exe...ge&MaximumPages=1&ZyEntry=1&SeekPage=x&ZyPURL

Search for Figure 6-10 page 166.

Again, I strongly recommnd using a flow transmitter instead of controlling valve position directly fom the DO signal. I aslo recommend against controlling the blower speed directly. Send a flow command to the HST and let them manipulate the blower.
 
What is HST?

HST is the Sulzer model indication for a High Speed Turbo blower. Earlier the OP indicated this was the blower his aeration system was using. Originally marketed by ABS in Europe it was one of the first packaged systems to include a very high speed (>10,000 rpm) synchronous motor with a direct-coupled centrifugal blower. The package includes an integral VFD operating at several hundred Hz and an integral controller.

https://www.sulzer.com/en/shared/products/hst-turbocompressor
 

Similar Topics

Dear All, Basically there are 8 conveyors as shown. Each conveyor will be controlled by a Siemens Micromaster 420 Drive. And each conveyors...
Replies
6
Views
7,854
Hello, I am trying to get a 32MV4341 Lenze AC Tech VFD running a conveyor to be controlled by a 24VDC photo eye on the nose end of the conveyor...
Replies
3
Views
1,886
"Hello, I am a beginner learning about PLC. Could you please give me some advice? I want to write PLC instructions as follows: When the sensor...
Replies
18
Views
3,233
Hello automation experts, I'm trying to learn about servo axis control in AB ControlLogix PLCs (L33ERMS). Attached is an example of a servo...
Replies
4
Views
1,050
Getting a trend of the control output, target position and actual position is a must. For a velocity control application substitute velocity for...
Replies
8
Views
2,301
Back
Top Bottom