VORTEX AIMing

[Conrad]

Section 2 on changing the settings of a Control System.

Continued from Section 1

The 10” Liquor Feed Line had a 6” Equal Percentage Control Valve to control the liquor flow to the Chip Chute (CC). . .Upon inspecting the design specifications I concluded that this valve was more than big enough and it operated at less that 50% open at the design flow for 750 TPD. . . add 50 % flow capacity. . .which would get the valve to be 50-60 % open or so. . . and the mill would run at 750 + 375 = 1125 TPD and there would be room for running to 1200 or maybe 1250 TPD, on that 6 inch valve. . .and if that would not be the case then an 8-inch valve would be required at the very most. A 10” valve would have a flow capacity of 278 % of the 6”valve capacity and the Huge Chip Cute Pump would not be able to handle that much flow. I advised the CO of this and he said:
“I would tend to believe you on account of what you have already achieved, but the Instrument Man says the 10" valve would solve the level control problem, so we are just going to replace it. . .we have a spare 10” valve so it will cost only cost some time to install it in a few days. . .the work is planned already for the next Maintenance Stop.
I was freaked out a bit by this. . .The Instrument Man obvious knew nothing about Control Systems and Flow Systems. So, I took the initiative to look at the 6” valve in more detail: The air pressure driven actuator was fed via a Spool Valve. . .for most of you this will be irrelevant, in regards to how that works, but the information I got out of the specifications was that the air feed and the orifice diameter was 1/8-inch and that was good enough to let enough air through it to operate the valve. . and do so fast enough. . . to allow the multi million Dollar digester to run at 950 TPD. . .a tiny hole of 1/8 “in diameter!!!!. So I looked at the Spool Valve and it was about 7 @ 8 mm thick and I figured this. . .the Liquid Level in the Chip Chute changes too fast for the valve to provide enough response to adjust the Flow variations fast enough. . .so the level goes too high or too low. . .suppose now that I make that 1/8” hole to be a 1/4“ hole. . .it will allow the air to enter the Actuator at a theoretical 400 % greater rate. . .and this will make the actuator respond left & right a great deal faster, maybe ever 300% faster. . .THAT should solve the problem!

To have more ammunition to convince the CO that the 10” Bal Valve would NOT solve the level control problem, I went to the Instrument shop where the 10” Fischer Ball valve was standing ready on the table for installation in 1-1/2 days. I asked a technician to hook up the Fischer Monster to an air supply at the same pressure as was used in the Mill and I tested this valve on an instant ON- sequence withy the valve 100% closed. It took ~10 seconds to 100% open. . . .
Slow as hell . . . but that is not the main issue. I checked how far the valve was open for a flow representative for 1000 TPD and I discovered that the valve would be less than 10% open. . .and at that point the Flow Curve for dQ/dx was almost horizontal . . . (x=1 is 100% open). . .a tangent of almost 0. . .and THAT was the critical thing. At the target flow at which the control would have maximised symmetric dQ/dx flow deviation response. . .the dQ/dx-line would be about 45 degrees instead of about 2 degrees at it was at 10% open. The result of installing the 10”valve would be that the Digester would not be able to operate at all, and that the entire Pulp Mill would stop running until the 6”valve would be put back!. . .I showed the CO the drawing I had made of the Equal Percentage Flow Curve for the 10” valve. . . .and that this valve would take 3 to 4 seconds to open enough to create a positive dQ to prevent the pump from running dry and about 1 second to close it to prevent the liquor from overflowing the mixing vessel. . which would prevent flooding the Digester Basement. The 10” valve would solve only 1/2 of the problem. . .This operational and control information was the result of knowing and understanding the details of the system and what would result if the settings were very wrong. . . As long as they were running on a “straight road” @ 750 TPD everything was fine. . .even the “wrong settings” of the control system allowed the production of 750 TPD but with my adjustments of the settings of the control system the Mill could run at 950 TPD easily.

The CO saw the light I showed him: “Holy Smokes. . .I believe you. . it is obvious that you are right. . .we ought to educate our the Instrument Man on how System Control ought to be done! I will cancel the Work Order on the big valve, but how do we get the 6” valve to move faster in such a short time?”. . I said: “That’s a Piece of Cake. . .you give me authorization to let the machine shop people make a new Spool Valve with 1/4” orifices. . .I will make a sketch for it. . .and you write out a Work Order that this Spool Valve is to be installed in the 6”control valve Actuator, using a 1/4” supply fitting on the air feed tubing. . .I am 100% sure the internal tubing from the Spool Valve to the Actuator piston is large enough to make this work for perfect Chip Chute Level Control, with the settings on the controller we have now, or even tuned a bit better”. . .The CO appeared convinced:. . .”Damn. . .I will get on it right-away. . .it makes perfect sense to me”. . .And so it happened! Making the new Spool valve was easy and was done quickly, and its installation during the Stop that night took no more than 30 minutes total. . .I was told afterwards.

In the morning after the Maintenance Stop at 8 AM I entered the Control Room and the “boys” there were sitting there doing nothing, leaning at ease in their chairs an looking content. . .the Digester was on line. . .purring like satisfied cat, not a thing was stirring. . .
The red trace on the Lever Controller Chart was gently moving up and down between a less than 1,5 cm range. . the Digester was running at 950 TPD. The boys welcomed me and gave me the Thumbs Up Signal. . .I was pleased I made it so!. . .I fine-tuned the controller a bit more . . .especially the D-knob I tweaked. . . and the range of the up-down level movement reduced to. . .I kid you not. . about 7 @ 8 mm. . .the red trace average very slowly moving a bit up/down indicating that the level average showed only a small cyclic deviation form the Set Point . . . Evidently the proportional, the integral and the derivative settings were virtually perfect. Then the CO appeared for his day Shift at 8:30 and he first checked the TPD rate which stood at 950 and then came right over to the level controller where I was. . .he looked at it and said:

“Surely you are faking this somehow. . I do not believe this can be what is happening in the Chip Chute in the Basement. . .”

I assured him that it was reality that he was looking at. The little 2-bit Spool valve, plus my fingespitzengeful http://en.wikipedia.org/wiki/Fingerspitzengef%C3%BChl had done the trick! The CO said then: “This is a tremendous achievement! I would not have believed this was possible if I had not seen this with my own eyes. . .my compliments. . .We are going for the 1000 TPD today!”

I went on with my normal work there for the Vancouer Consulting Company I worked for and some time later I inquired how “things were going” with the Digester and was told they were operating at the fringes of 1250 TPD. . .without making any more fundamental changes to the basic system. . . Maybe they had done a bit more fine-tuning on the Level Controller for the Chip Chute to get to the 1250 TPD, but I have no information on that.

This experience was the Cherry on the Cake for my engineering experience. Many people say things that are not true and when I hear such falsehoods I jump on them to try to make it clear to them that looking beyond the length of their noses to learn more is very useful, but also it makes life exiting. . .one can escape the confines of boring box that one might be in.

This is the Short Version of my story