Pressure Control - Integrating or Non-Integrating?

It has been said in previous posts here that Pressure Control is typically an integrating process (non-self-regulating).

However I am having trouble wrapping my head around a particular setup.

I realize in the following ideal scenario (split range PID) that the pressure control process would be integrating, with a final control element that increases or decreases pressure in a vessel:

CV: < 50 % - Release pressure in the tank at a rate inversely proportional to the output.

CV = 50% - Idle, remain at current pressure.

CV > 50% - Increase pressure in tank at a rate proportional to the output.

However, let's say another vessel has a relief PCV that vents pressure in the tank: 100% is Open, 0% is closed, the valve merely vents the tank.

Additionally, there is a separate mass flow controller regulating the rate of air entering the tank.

Assuming the MFC is left at a constant rate of air into the tank, is the Pressure Control process now self-regulating (non-integrating)?

Let's say air is entering at a fixed rate of 300 SLPM.

PCV PID loops kicks in and eventually the valve reaches 25% open, steadying at a desired setpoint of 2 PSIG.

Setpoint now increases to 5 PSIG. PCV eventually starts to close and will now settle at a different CV to compensate.

Isn't this a self-regulating process?

I think you have the right idea

kdcui said:

It has been said in previous posts here that Pressure Control is typically an integrating process (non-self-regulating).

Click to expand...

Yes, but that assumes there is one control element that controls all flow in and out. Your system as more than one source and sink of flow.

However, let's say another vessel has a relief PCV that vents pressure in the tank: 100% is Open, 0% is closed, the valve merely vents the tank.

Click to expand...

I am assuming the other tank is relieving into the controlled tank. The tank is still integrating the flow from the relief valve and the controller. So far so good.

Additionally, there is a separate mass flow controller regulating the rate of air entering the tank.

Click to expand...

Yet another controller.

Assuming the MFC is left at a constant rate of air into the tank, is the Pressure Control process now self-regulating (non-integrating)?

Click to expand...

I don't see how. So far you have only mentioned flow into the tank. Pressure will build until equalization or bursting.

Let's say air is entering at a fixed rate of 300 SLPM.

PCV PID loops kicks in and eventually the valve reaches 25% open, steadying at a desired setpoint of 2 PSIG.

Click to expand...

It must be letting flow out. If the pressure is to remain constant the flow out must be equal to the flow in.

Setpoint now increases to 5 PSIG. PCV eventually starts to close and will now settle at a different CV to compensate.

Isn't this a self-regulating process?

Click to expand...

The PCV PID is doing the regulating. If the valve didn't close then pressure would not increase would it?

Thanks for the reply Peter,

I may have made the situation seem more complex than I intended. Both of the tanks in my hypothetical situation are completely independent:

The first vessel adds/relieves pressure with a single split range control element.

The other vessel relieves pressure with a PCV, and is pressurized (at an assumed constant flow rate) by an additional controller.

I guess what I was getting at (and I think you implied) is by having that 2nd controller adding air into the system, the response of the Pressure Control will behave as a non-integrating process:
The PCV output will change according to the setpoint of the pressure.
In an integrating process we would expect the output to spike an return to a nominal value once setpoint is reached.

Is this correct or am I still missing something?

Originally posted by kdcui:

I guess what I was getting at (and I think you implied) is by having that 2nd controller adding air into the system, the response of the Pressure Control will behave as a non-integrating process:

Click to expand...

Not really. What happens if you set the command to the PCV to 0% but the MFC maintained it's rate? You may want to do that VERY remotely, by the way.

Just because a system appears to self-regulate through a portion of it's range doesn't make it self-regulating. For example, consider a shallow water tank with a small hole in the bottom being fed by flow control valve that can supply much more liquid than the hole can let out. Through a very small control range I can make that system look like a self-regulating system. But ultimately, if I close off the flow control valve completely the tank will eventually empty. If I open it completely the tank will eventually overflow.

Keith