By Greg McMillan and Stan Weiner, PE
Stan: From a barrage of spam, CONTROLs editors were able to sort out some excellent answers to Mays Puzzler from Bruce Seaba and to Junes Puzzler from Vince Mattione.
May's Puzzler: "Ping, Pong, Ping, Pong, Plonk" asked "what happened when 24 ultrasonic level transmitters were installed in a tank farm, and after startup some meters didn't read at all, and nearly all drifted up and down over a 24-hour period?"
June's Puzzler: "Shifting the Zero" asked "is a plant stuck with constantly recalibrating the zero every time its tank empties, or will wholesale transmitter replacement solve the problem?"
In conjunction with our retirement motto of better late than never, we offer them now. The answers from Hunter Vegas were listed in Puzzlers Gone Wild, CONTROL, July 06, p. 71.
Bruce: Twenty-four ultrasonic level transmitters installed over the objections of the instrument contractor leads me to believe that what was spotted was a misapplication. Possibly, the engineer specifying the instruments failed to realize that these were floating top tanks. When you say the meters drifted up and down, you literally meant the meters were floating up and down, and not that the indication from the meters was doing so. The meters that failed were those that traveled past the limit that the cable attached to them could stretch, or the ultrasonic level transmitters were unable to compensate for the change in temperature in the vapor space of the storage tanks. A warmer vapor space would have increased the velocity of the ultrasonic waves, reduced the flight time, and indicated a higher level than actual. This could have occurred because the ultrasonic meters werent temperature compensated, or were only compensated at the transmitter. A large storage vessel with no exchange of vapor volume could develop a temperature gradient from solar energy that single-point temperature compensation couldnt handle.
Vince: The 25-ft capillary is filled with a liquid which expands with increasing temperature. The effect of this is made worse with smaller diameter diaphragms, larger diameter capillaries, and longer capillaries. A 0.03-in., 25-ft capillary on a 2-in. flange with a temperature change from 25ºF to 50ºF and DC 200 silicone may see an additional low-side pressure of 5-6 in. of water column from the change in temperature alone. Changing to a 15-ft capillary, if possible, would cause this to only be a 4-in. water column change. Changing to a 3-in. would cause the original scenario to change 1-2 in. of water column. Changing the fill fluid to inert (Halocarbon) would help because inert is about 20% less temperature dependent than DC 200. If the capillary is 0.07 in., then changing it to 0.03 in. would help.
If possible, maintain the capillary at a constant temperature (inside versus out). Keep the capillary out of direct sunlight and away from heat sources. Changing to another manufacturer would probably not help.
Using the HART parameter to monitor case temperature and assume that is roughly the same as fill fluid temperature, the control system could compensate for changes in ambient air temperature. Since these tanks have d/p cells to compensate for tank vapor pressure changes, theyre obviously not open to atmosphere.
These tanks most likely have a conservation vent that relieves at a given amount of overpressure or under pressure. When the tank is emptied, it will reach some level of negative pressure before the vent opens. (This negative pressure will remain until the tank is filled or vented to atmosphere). The cell on the 4-in. bottom flange has a surface area of approx 12.5 square inches. The 2-in. top cells area is about 3.1 square inches. Any change in tank vapor pressure will have four times the effect on the bottom cell as the one on top. The transmitter was most likely calibrated (zeroed) at a vapor pressure of 0 psig. Any change in vapor pressure from this will cause the transmitter reading to be incorrect. The error probably wouldnt be noticed until the tank was emptied and the tank was under negative pressure. These two d/p cells need to be the same size!
Greg: Sitting here listening to Moody Blues, I am having a flashback to my days as a parent, and wondering if proportional-integral-derivative (PID) or model-predict control (MPC) would be better as a control algorithm. PID excels at handling the unknown upsets, non-stationary behavior, and the unpredictable nature of a teenagers response, and derivative action provides some preemptive action. However, the abrupt action by PID could amplify noise in a teenagers behavior. However, then there is the dead band from backlash and the resolution limit from friction. Maybe a lot of dither from the PID could keep the teenager off balance, and minimize inaction if it doesnt wear you out. There is inverse response to contend with where the teenager starts out doing the opposite of what was requested. Theoretically, MPC could build this behavior into its knowledge of the teenagers future trajectory. Also, the patience of a MPC is inline with modern parenting. But Im not sure your can do enough pseudo random binary sequences (PRBS), or know the time to steady state of a teenager, so my initial thought is PID with an adjustable bias to deal with the potential unstable response. Reader views and experience here are appreciated as always.
Stan: Since we have some room to fill, we offer some reviews of our book The Life and Times of an Automation Professional. Please buy a copy so our cartoonist Ted Williams can go to college.
Greg: We are talking about lunch money since Ted works at Duke.
This book could not have been written/drawn 7,000 years ago.Society for Realizing How Far Partial Differential Equations Have Taken Us
Classic Stan and Greg.Stan and Greg
A must read for anyone who cherishes those adorable automators.AAA (Association for Adorable Automators)
The complete guide to the love and care of automation professionals.SPCA (Society for the Prevention of Cruelty to Automators)
The mother of all automation cartoon books.MAD (Mothers of Automation Design)
Greg: For a timely end, we offer the Top 10 Signs Your Project is Behind Schedule by Lance Dusing.
