Ideas needed

[kurt.brinker]

This is the scenario.
We have 3 fluids to mix. One fluid is under constant pressure and the other 2 are pumped out. We have flow meters on all three lines right where they leave their respective containers. The flowrate of the pressurized fluid will vary (not alot, but a few GPM) and I want to incorporate PID loops for the pumps on the other 2 liquids so that the flowrate of fluids 2&3 is always equal to the flowrate of fluid 1.

I am thinking i should ensure the scaling on all three meters is the same and then take the 4-20mA signal from the first flowmeter and have that value create the setpoint for the other meters. The pumps will then increase or decrease acordingly. ( obviously, The PV for the loops will be the 4-20 from the 2 meters. )
Also i am using SLC 500.

One more question. WHen i recieve the 4-20, should i immediately scale it 0-16383 and use that value or leave it as 3277-16383.

I'm not looking to have this programmed for me or anything of that nature. I was somewhat thrown into this position and i am burning the midnight oil every night learning as much as i can. I was supposed to be studying under our main programmer but due to alot of personal problems, he hasn't been around much. Anyway let me know if there is an easier way or if you need a better explanation/

 

[Bob O]

kurt.brinker

Look here http://www.ind-info.com/case-studies.html for information on PID. Also search for PID and the name Ron Beaufort in this forum.

Good Luck

 

[kurt.brinker]

GREAT link. Thanks alot.

 

[Gerry M]

I assume this will be a constant flow from all the pipes.

If so, you need to implement "ratio control". Google that term and you should be able to find some info. on that type of control.

 

[kurt.brinker]

correct. the flow will be constant but the ratios will be different. The ratio base is .15 lb/min of #1, 150lb/min of #2, and 1.5lb/min of #3.



I am going google that term right now. thanks.

 

[rory]

Pid

Should work sweet, since you have basically one constant flow and use that for the setpoint of your another 2 loops you will be ok. I would scale it so it makes sense ie if your flow is 0-10 ls why not scale it 0-10000 or somthing. Just make sure you use the full range of your output ie when you set your output on your channel to run the full 4-20ma ie 6554 to 32768. Be aware of min drive speeds and the like.

Treat them inpendent, get one working then the next.

 

[kurt.brinker]

thats what i thought initially as well. I found out today its a ratio.

fluid 1 is .15 lb/min
fluid 2 is 150 lb/min
fluid 3 is 1.5 lb/min


that throws a HUGE wrench into my plans. I'll get it figured out tho.

 

[Mark Buskell]

Are you using 3 flow meters and 3 variable speed pumps?

If so, its just a matter of figuring you maximum flow and then scaling the setpoint for your PID.

I do injection of chemicals into a water stream. Some times it's fixed flow setpoints based on the water and on other systems, the chemical setpoints follow the water. We work in grams per liter, so its just a matter of multiplying the chemical grams per liter times the water setpoint (LPM) and using that result for the setpoint of the PID.

 

[kurt.brinker]

Fluid #1 is in a pressurized tank and released when a solenoid valve opens, even if we maintain precise pressure on the tank, the flowrate will fluctuate due to viscosity changes and such. Fluid #2 is under pressure as well and its flow will be controled via a control valve. The third fluid is pumped out via a linear pump and controlled by a control valve.

We have flowmeters on all three lines.

 

[Gerry M]

One thing comes to mind for scaling:

Using % helps sometimes. In other words, if Flow1 increases 1%(1% of .15), then increase Flow2 & Flow3 by 1%(1% of 150, and 1% of 1.5 respectively). Check and see if this works for you.