Flow loop and pressure loop

ddeshi2

Member
Join Date
Dec 2010
Location
Baton Rouge
Posts
89
Hi all, I have two PID loops in the program: flow loop and pressure loop.

Flow loop: We have a flow meter (magnetic flow meter for water), a positive displacement pump, a VFD to control the speed of pump thereby controlling the flow.

Pressure loop (To create Backpressure with the water flow): We have a control valve (Fail Open)and a Pressure Transmitter.

I'm using a timer to trigger the PIDs (250ms) and loop update times for both PIDs (250ms) and they seem fighting with each other. I read somewhere here that the pressure loop update time should very fast compared to the flow loop and they cannot have same loop update times. Is it true? Can someone please elaborate on this or give me a link to the article about pressure and flow loops.


Also, My flow and pressure loops have the following PV and CV:

Flow loop: PV-- FLOWMETER READING
CV-- VFD SPEED
Pressure loop: PV-- Pressure Transmitter
CV-- Control valve (Back pressre)

I need to get a back pressure of 50 PSI with 3 GPM of water. I'm using two PIDs now and they are not cascaded. Am I thinking in the right direction or should they be cascaded or something? Please help.

Thanks in advance,
Ding
 
If you have a positive displacement pump the flow will be constant regardless of pressure in the system. For a given speed you will get a constant flow rate. An independent flow loop will work fine. I would make this a very slow response loop. You really don't even need a PID for this. Just calculate the speed based on the pump displacement and add a correction to compensate for internal leakage and you should be good to go. I would only use the flow meter to determint that compensastion for internal leakage.

The PID for pressure should be independent. If you are having stability problems your tuning may be too aggressive. I'm not really sure what you need the pressure control loop for, since at constant pump speed you'll get constant flow regardless of the pressure. What is the flow going to?
 
If you have a positive displacement pump the flow will be constant regardless of pressure in the system.
ddeshi2 can control the flow or pressure by varying the speed of the VFD.

The two PIDs may fight each other.
I would use the PI speed control in the VFD itself. I would first try a open loop voltage from the PLC to the VFD so that it the pump supplies 3 gpm and then throttle back on the pressure control valve to increase the pressure. The load on the VFD will increase but the PI control in the VFD should try to keep the speed constant.

I would start with proportional only control for the pressure control valve and increase the gain slowly. Pressure control loops are tricky and need to be fast. At 250 milliseconds you do have a chance unless you increase the hydraulic capacitance which will slow down the rate of pressure change with change in the volume. I would add an accumulator between the pump and the control valve so the pressure doesn't change as fast when the control valves opens or closes.
 
@ Tom Jenkins

You really don't even need a PID for this. Just calculate the speed based on the pump displacement and add a correction to compensate for internal leakage and you should be good to go. I would only use the flow meter to determint that compensastion for internal leakage.

My Pump is rated for 12 gpm, 750 psi , 1450 rpm. I want 3 gpm of water at 150 psi. I'm not sure how to calculate speed based on pump displacement. I looked at the pump manual and I found the following formula:

Rated GPM Desired GPM
---------- = ---------------
Rated RPM Desired RPM

Can I use the above formula to calculate pump speed? But it doesn't take pump's displacement into account. Please help me in calculating pump's speed.

@ Peter Nachtwey

I'm using a Powerflex 40 VFD and I can setup PID in Vfd itself.But I'm not familiar with it and want to go with PID in ladder logic to control the speed of pump (thereby controlling the flowrate of water). I'm planning on increasing the flow loop update time to 1 sec and see how it reacts. The reason I want to use 1 sec is because my flowmeter was set for a Time constant of 1 sec (it updates the flow for every 1 sec). Do you think it should be more than 1 sec for good control?

I would add an accumulator between the pump and the control valve so the pressure doesn't change as fast when the control valves opens or closes.

We have a pulsation dampener between pump and valve. What is an accumulator? Does it serves the same purpose as a pulsation dampener? I think accumulator and pulsation dampener both behave similar. Am I right?

Thank you!

Ding
 
My Pump is rated for 12 gpm, 750 psi , 1450 rpm. I want 3 gpm of water at 150 psi. I'm not sure how to calculate speed based on pump displacement. I looked at the pump manual and I found the following formula:

Rated GPM Desired GPM
---------- = ---------------
Rated RPM Desired RPM

Can I use the above formula to calculate pump speed? But it doesn't take pump's displacement into account. Please help me in calculating pump's speed.
You need to know how fast the motor turns for every volt. Then it is simple matter of using ratios.

I'm using a Powerflex 40 VFD and I can setup PID in Vfd itself.But I'm not familiar with it and want to go with PID in ladder logic to control the speed of pump (thereby controlling the flowrate of water).
Use the VFD PI control. The PLC is TOO SLOW!!!!!

I'm planning on increasing the flow loop update time to 1 sec and see how it reacts. The reason I want to use 1 sec is because my flowmeter was set for a Time constant of 1 sec (it updates the flow for every 1 sec). Do you think it should be more than 1 sec for good control?
NOOOOOOOOOOOOOOOOOO!!!!!!!!


We have a pulsation dampener between pump and valve. What is an accumulator?
A bigger version of the pulsation dampener. There are many kinds of accumulators but the most common have a bladder that is pressurized to about 80% of the operating pressure. When the system is running oil will compress the nitrogen in the bladder

[quote
Does it serves the same purpose as a pulsation dampener?
[/quote]
Much more and better

I think accumulator and pulsation dampener both behave similar. Am I right?

Thank you!

Ding
Yes, they are similar but an accumulator is much bigger.

ddeshi2, you need to discus this with the designers. You don't have much of a chance for success without some rethinking and redesign.

1. Use the VFD PI start with P gain only this will give the system some drop and be easier to get into the ball park.

2. Get an accumulator. An accumulator adds hydraulic capacitance.

3. The pressure control PID needs to be as fast as possible. Every 10 ms may work if you add more hydraulic capacitance.
 
By definition Displacement = Rated_Flow / Rated_Speed, with the result being gallons per revolution in your case.

And yes, you can use the formula you provided to determine the speed just by rearranging it algebraicly to:

DesiredSpeed = DesiredFlow x RatedSpeed / RatedFlow

This ignores pump internal leakage (volumetric efficiency) but is probably close enough for your purposes. (I got 15 Hz at 60 Hz nominal, or 25% speed.)

I don't see why you would want to screw around with a PI or PID loop for flow - the straight algebraic calculation will give you a result as accurate as any feedback loop. If you really need tremendous accurcy just add a "fudge factor" by increasing the rpm with a constant to compensate for the volumetric efficiency.

Similarly, I don't see what good the pressure control loop is going to do for you. The pump produces flow, the system restriction produces pressure. A positive displacement pump will give you this flow even if the back pressure changes.

You need to decide what the system requirement is. If you need constant flow then setting the pump speed will give it to you. This would be the case if you are controlling a chemical feed rate, for example.

If you need the constant pressure then you are probably going to need to vary the flow to maintain pressure as your system demand varies.

You can set the pump speed, maintain a constant flow, and adjust a valve downstream of the pump to maintain a constant pressure at the pump discharge. However, I don't see that this will do you any good whatsoever. What kind of system is the pump discharging into?
 
Last edited:
@ Peter Nachtwey

Thank you for the detailed explanation. I'll talk to my engineer boss about the accumulator and let you know what he says. I'll be at the field tomorrow and try to set up the VFD PID in the VFD itself and see how it behaves. Also I'll try not using VFD PID at all as Tom mentioned above and let you know.

@ Tom Jenkins

Thank you for the help. It is a steam generator.We feed air and fuel to the combustion chamber in addition to water ( tool design is such that the water will cover the walls of chamber keeping the chamber cool and flame is ignited in the vortex core ) It's difficult to explain here but look at our website www.hydroflametech.com and you'll understand what I'm saying. For getting the flame we need backpressure and we are using a Fail open valve to get the required backpressure( difficult to control though). I'll try not using the PID at all for water and see how the system behaves and let you know.

Thank you!
Ding
 
Ok, so I’ve a slightly better understanding of your problem. I’m assuming we’ve been considering water control only of a novel steam generator. It seems that the water is distributed in multiple sections, but I’m not sure. In the top section you probably have a manifold/nozzle arrangement to create the ‘flame containment’ annulus; this manifold probably has a minimum water backpressure requirement to ensure proper water flow distribution. This, in turn, prevents hotspots in the unit. Downstream, there seems to be some quench water flow, presumably supplied by the same pump and common header. (So, maybe you’re controlling total flow to the unit while maintaining backpressure at the top end?)

It would be helpful to have a water flow diagram to illustrate the layout of the piping/pumps/instrumentation along with a narrative of the objectives/constraints of the various streams and process equipment. (Maybe this is proprietary).

By limiting your description of your problem you limit the usefulness of others’ recommendations. I lose interest in otherwise interesting/challenging problems when people can’t adequately describe what they’re trying to do (rant over). Others are more charitable and bless them for it!
 
I don't think a pressure control valve will do anything for you. I assume that you have multiple nozzles for the water spray. I also assume that there is some pressure in the combustion chamber. The pressure at the nozzle entrance will be the combustion chamber pressure plus the pressure drop through the nozzle orifice for the given flow. Changing the back pressure at the pump head won't change a thing.
 
I would be nice to know the goal

Too often people tell us what they want to do but that is often different than what they should be doing to achieve the desired end goal.

Changing the back pressure at the pump head won't change a thing.
It should change the flow rate through the nozzles but now I don't think the control valve is necessary. The pump can be run at what ever speed is necessary to maintain the desired flow through the nozzles. The accumulator would keep the pressure and there for flow relatively stable .

I agree with jamesau, this hasn't been thought out and it would have been better to say what the end goal is. I would think that the flame could only heat to steam so much water so flow control is critical and pressure is what ever is required to achieve that flow. I would also think the temperature of the incoming water would be important if it is being heated to steam. There is no mention of that.
I would also be nice to know the steam out flow because this can be used as a feed forward to the pump speed.

ddeshi2 said:
I'll talk to my engineer boss about the accumulator and let you know what he says.
Your engineer boss should know the answers to your questions. If not then it is time to get a new engineer boss. Really!!!!
 

Similar Topics

Good day all. I just installed a basic pressure control loop the other day. It's in a stand alone CompactLogix that serves no other purpose than...
Replies
14
Views
6,509
Hello All Could we get some expertise on flow control ? -Using a PID loop in Productivity 2000 with an analog output, How can we convert...
Replies
19
Views
1,514
We're implementing a second pump in a piece of equipment that pumps primarily water, or fluids similar to water. This pump is located in a stand...
Replies
23
Views
13,061
Dear All, I need a sample PLC program to count the output pulse of a mass flow meter so that a specific amount of mass (for example 100gm)can be...
Replies
2
Views
77
Has anyone installed Hp laserjet pro 4203 to OMNI 3000 or 6000 via ethernet..Is there a specific Config? I have a problem..Please advise
Replies
0
Views
42
Back
Top Bottom