Greg McMillan ([email protected]) and Stan Weiner, PE, ([email protected]) bring their wit and more than 66 years of process control experience to bear on your questions, comments, and problems.
Tuning Trashed by Teflon?
There was a process that made extremely corrosive chemicals. The project manager was totally tantalized by the prospect of not paying for tantalum parts. To reduce the use of tantalum in the valve trim, it was decided to use Teflon-coated stainless steel. When the plant was started up, why was it impossible to tune any loop that had a control valve with Teflon-coated seat rings, plugs, and stems?
[email protected].Send an e-mail with your answer to the Puzzler, control questions, or comments to
Greg:
Stan:
Greg:
Stan:
Greg:
Tim:
Charlie:
Mark:
Jerry:
Stan:
Greg:
Stan:
Greg:
Stan:
Greg:
Stan:
Greg:
Stan:
This Months Puzzler:
I think the pool guy comes every Wednesday and takes them out with a big net. The project manager was totally tantalized by the prospect of not paying for tantalum parts. This is the stuff that performance reviews are made of. Now for this month's disclaimer: This column was prepared while we wondering why there were no fish in the swimming pool. There was a process that made extremely corrosive chemicals. To reduce the use of tantalum in the valve trim, it was decided to use Teflon-coated stainless steel. When the plant was started up, why was it impossible to tune any loop that had a control valve with Teflon-coated seat rings, plugs, and stems? What about another puzzler? I don't know if you ever heard about the largest distillation column that we ever built. The process engineers kept running calculations to optimize the overhead line. They finally decided that 72 in. would do the job. However, the plant engineers and operating people wanted a 100-in. line. Because of costs, the 72-in. was built. The column design capacity could not be achieved--until the tarpaulin was removed from inside the overhead condenser. Middle signal selection is the only foolproof way of riding out a single failure of any type including a failure at setpoint. Gosh knows there are more than enough fools in the control room. It could have saved the operator's reputation in this puzzler. When slurries meet a branch in the pipeline, there is no way you can guarantee an equal split of the solids. We learned this the hard way from feeding parallel centrifuges. The best way to help the electrodes see about the same process conditions is to put them all in series in the same line with appropriate methods to isolate and bypass an individual electrode. Of course, Jerry is right that three pH electrodes will never read exactly the same unless the protective caps are still on. Some people find this too disturbing and remove one or more of the sensors. Personally, I am disturbed enough to find the differences interesting, including the reproducibility of the art of glass blowing and the scatter from imperfect mixing and overcalibration. I know some people who have failed at the beginning of their lifetime. Another point is that pointing the pipelines down may leave you and the electrodes with a partially empty feeling. The mid select concept has been used in safety systems to prevent a single failure from shutting down a unit. I have concluded that people usually fail at the end of their lifetime. Why do people fail at different rates? It is a similar reason why pH electrodes fail. We are not all made the same and, despite illusions to the contrary, we don't have identical environments, except perhaps if you live in Naples. This operator is tops. He was told to control at a certain pH and found the best way to do so. If the inlet and outlet lines are both above the sensors, in a little time, they will all be in the same neighborhood. Actually, if you rotate the sensors' outlet line to point down, they'll all be seeing the same representative stream. Now, if you want all three to read the same, forget it. It's never been done. This is pH, guys. Mid select technology is the vendor's way of never having to say he's sorry. It also sells a bunch of sensors and transmitters. Now, a real pH problem to solve is: Why do sensors in identical service fail at so widely disparate lifetimes? Of the three pH electrodes, the one installed in the bottom line is most susceptible to coating from heavy sludge, solids, etc., in the slurry stream. A coated electrode has a slower response time than a clean probe, so it will respond in a sluggish fashion. Operators may like the smooth signal on their trends, but the control system will respond too passively because real fluctuations in the slurry pH are lagged and attenuated by the coating. Slurries are notoriously heterogeneous (by definition!) and I would expect the lowest line would have more solids (less water) so the pH will not be representative of most of the water. I bet the flows are not equal in the three lines as the thicker slurry has a higher viscosity. Changes in acid addition upstream would not be seen as soon as the pH would change in the total flow. If the solids in the slurry are reactive, well, so much the worse. I especially like the Zen-like suggestions to be the best engineer possible. These could be applied to many spheres of life. The sensor in the bottom line should give a smooth trend. The bottom line is plugged. Not much going through it to change the sensor's output. There's nothing like controlling from bad data to hose things up. (Along with all that other stuff that hoses things up.) By the way, I surmise that you are Red Green fans from your column, therefore: If the women don't find you handsome, they should at least find you handy. Talking about smooth measurements, we had correct answers from Mark Chatterton, Charlie Demarest, Jerry Donovan, and Tim Nett to the February puzzler that asked why the bottom line pH sensor in slurry service was so smooth: We have contemplated great issues such as the high temperature in Naples every day and have concluded it is always 84* despite forecasts of 83, 84, and 85. We don't attribute this as much to round-off error as much as the need for the weather person to attempt to make it seem like it is a real job. I needed some decompression after trying to teach university students something practical they can use on the job. Of course, this is all theoretical in that there are no jobs. It has taken a week of intense relaxation by the pool to reach our creative peak. Here we are in Naples, Fla., working up to this literary moment.
