PID tuning on SLC 5/05

[Robb B]

Originally Posted by drbitboy

Whoa, thanks!

Is the stilling pipe used for a level measurement?

no, only for pressure, vacuum and level switches (ignored by operators )

Is the stilling pipe slotted or perforated down its sides inside the treating vessel?

no, it is external to the vessel, and is plumbed in near the bottom on a fitting.

Does the oil level reach the top of the treating vessel, or even go up into the extension of the stilling pipe above the vessel?

see above re: operators, but ideally, no oil should be present above the highest point of the vessel, ie: the ceiling.

[drbitboy]

What is the rate of pressure rise just before the control valve opens? e.g. how long does it take for the PV to increase the last 5 or 10psi before it reaches the setpoint?

[drbitboy]

Originally Posted by Robb B

[the stilling pipe] is external to the vessel, and is plumbed in near the bottom on a fitting.

see above re: operators, but ideally, no oil should be present above the highest point of the vessel, ie: the ceiling.

So is there air compressed into region at the top of the vessel, separate from the air at the top of the stilling pipe?

Because the treating vessel is initially full of air e.g. when the wood to be treated is placed inside, and then the vessel is somehow sealed before the process starts pumping oil into it? If yes, then once the oil level inside the vessel covers the stilling pipe fitting near the bottom, there are two separate air pockets, one in the vessel, one in the stilling pipe?

[Robb B]

Originally Posted by drbitboy

So is there air compressed into region at the top of the vessel, separate from the air at the top of the stilling pipe?

Because the treating vessel is initially full of air e.g. when the wood to be treated is placed inside, and then the vessel is somehow sealed before the process starts pumping oil into it? If yes, then once the oil level inside the vessel covers the stilling pipe fitting near the bottom, there are two separate air pockets, one in the vessel, one in the stilling pipe?

That's very likely, as the vessels sit with a very slight (~1 degree) slope downhill to the drain, which is approximately the same place as the stilling pipe, 2/3 of the way down the vessel. If I had to guesstimate, I'd say there was about 6" of elevation difference over the 150 feet or whatever the vessel is long. There is a vent valve which opens while the vessel is filling or draining.

There is a level switch that is supposed to automatically shutoff the fill pump but I think they've been running it in manual for years now. I did test those switches last year and they worked fine.

[Robb B]

Originally Posted by drbitboy

What is the rate of pressure rise just before the control valve opens? e.g. how long does it take for the PV to increase the last 5 or 10psi before it reaches the setpoint?

Hopefully I can catch a few pressure cycles over the next 24 hours, and we can estimate it from that. The general shape at 165 psi was a distinct S curve when pressure reached the setpoint, at 125 psi, it's much more vertical at the point that the pressure reaches the setpoint.

[drbitboy]

Originally Posted by Robb B

That's very likely, as the vessels sit with a very slight (~1 degree) slope downhill to the drain, which is approximately the same place as the stilling pipe, 2/3 of the way down the vessel. If I had to guesstimate, I'd say there was about 6" of elevation difference over the 150 feet or whatever the vessel is long. There is a vent valve which opens while the vessel is filling or draining.

WHUT?


how does the vessel pressurize if there is an open vent valve?


What is the diameter of the treating vessel?


What is the diameter and length of the stilling pipe?


- Dr Bitboy, DMD.

[drbitboy]

Originally Posted by Robb B

...The general shape at 165 psi was a distinct S curve when pressure reached the setpoint, at 125 psi, it's much more vertical at the point that the pressure reaches the setpoint.

That sounds like 165PSI is near the upper limit of head for the pump i.e. behind the knee of the curve with low flow and very large fractional change in flow for a small change in head, while 125PSI is past the knee where there is still a large flow.

[MaxK]

Originally Posted by Robb B

The pressure applied is not constant, as it builds from 0 PSI to setpoint in a bit of an S curve.

Is pressure curve looks this way: almost linear at the beginning then k(1-exp(-t/T))-shaped ?

[jamesau]

If 125psi is the new normal target operating point, consider a smaller impeller for the pump. If you expect multiple pressure operating targets, consider a vfd for the pump motor.

[Robb B]

Originally Posted by drbitboy

WHUT?


how does the vessel pressurize if there is an open vent valve?


What is the diameter of the treating vessel?


What is the diameter and length of the stilling pipe?


- Dr Bitboy, DMD.

The vent valve closes while doing the pressure or vacuum cycles. It is only open while the vessel fills or empties. The diameter is approximately 8', length is approximately 180'. Stilling pipe is 3" diameter, runs about 15' from the vessel fitting to the pressure/vacuum sensors.

[Robb B]

Here's a typical pressure cycle. The initial pressure is where we have issues, the end of the cycle is a stepped bleed-off of pressure, which is working fine. As the graph shows, there is no positive CV action until the setpoint is crossed, calculated error is ranged -200 to +200, I could re-do with greater ranges if it would help.

[Robb B]

Originally Posted by jamesau

If 125psi is the new normal target operating point, consider a smaller impeller for the pump. If you expect multiple pressure operating targets, consider a vfd for the pump motor.

The 125psi target might (will!) change in the future, I'd like to design the system to respond better to any future setpoints that might be implemented.

[Peter Nachtwey]

Originally Posted by drbitboy

A 40HP constant-speed pump (re-read the OP) for a PILOT valve? Really?

my first assumptions where wrong. It was hard to visualize the system after just the first post. The differential equation that I posted before doesn't apply. After seeing the diagram it is clear that the P*V=n*R*T is the way to go The air in the treatment vessel is more like an accumulator in hydraulics.
Anyway the pressure doesn't change linearly with level and if the pump is a centrifugal pump, the pump flow will vary with pressure. At the new lower pressure the pump flow will increase which will increase the pressure faster but this may be offset by the increased air volume will tend to reduce the rate of change in the pressure.

It should be possible to get reasonable control with just a P gain. There will be an offset but then an integrator can fix that.
I don't see why the system is so difficult to control.
Any this thread has exceeded the 20 questions limit.

weird system.

[drbitboy]

Originally Posted by Robb B

[Vessel] diameter is approximately 8', length is approximately 180'. Stilling pipe is 3" diameter, runs about 15' from the vessel fitting to the pressure/vacuum sensors.

The bulk modulus of the air in the stilling pipe may contribute, but it is likely less important than bulk modulus of the oil, I would guess the relative oil:air contribution to the pressure vs. volume behavior is somewhere between 2:1 AND 20:1, unless the oil is hydraulic oil, in which case it may be 1:1.

[drbitboy]

Originally Posted by Peter Nachtwey

my first assumptions [were] wrong.

Join the club; we're having jackets made up.

Originally Posted by Peter Nachtwey

It should be possible to get reasonable control with just a P gain. There will be an offset but then an integrator can fix that.
I don't see why the system is so difficult to control.
Any[way] this thread has exceeded the 20 questions limit.

weird system.

^this

Yeah, I suggested stiff P-only control over two dozen posts ago (#18).

Originally Posted by Peter Nachtwey

After seeing the diagram it is clear that the P*V=n*R*T is the way to go The air in the treatment vessel is more like an accumulator in hydraulics.

I thought so too for some time, but consider that the vessel is vented until it is full of oil, and look at the final liquid:gas volume ratio: 10⁶-10⁷; PV = nRT is out the window.

This is more like a hydraulic system, perhaps with some undissolved air, so it's solidly in @Peter Nachtwey's wheelhouse.