APC and Wireless Rabbits

Dec. 22, 2008
By Greg McMillan and Stan Weiner, PE

Greg McMillan and Stan Weiner bring their wits and more than 66 years of process control experience to bear on your questions, comments, and problems.
Write to them at
[email protected].

Greg: To me advanced process control (APC) is any technology that incorporates process knowledge and puts it all on the line online. Advanced regulatory controls, such as feed-forward control, use process knowledge to come up with the right feed-forward gain and dynamic compensation. It is more evident in the higher level technologies that include process performance monitoring, model-predictive control, property estimators, rampers or pushers, and optimizers. The software has gotten less expensive and easier to use, but with the limited resources today, what do we need to do to insure success and quantifiable benefits that will justify additional APC projects? All of these upper level technologies need to be based on a foundation of comprehensive, smart and sensitive measurements and final elements. If the foundation is not broad or solid enough, the implementation cost and performance of the APC technologies become problems to the point where both the APC job and you can get a bad name.

Stan: One of the best things you can do is to replace differential head meters with Coriolis flow meters, and if this is not feasible, magnetic flow meters or vortex meters. Differential head meters have a long-term installed accuracy of about 5% of span. Coriolis meters have an accuracy that can be as good as 0.1% of rate that is independent of process conditions and profile with no straight run requirements.

Greg: Coriolis mass flow meters provide an incredibly accurate density measurement that can be used to calculate the concentrations of a two-component mixture directly or infer the composition of a complex mixture. This composition information enables online component balances and composition control along with the more obvious online mass balances and mass flow control. If you realize that what you really want to know and control is often the composition of process streams, you can appreciate the significance and scope of the opportunity for more intelligent control of unit operations. The use of Coriolis meters for column control enables more accurate ratio control and inferential measurements of feed, distillate and bottoms compositions for tighter enforcement of material and component balances. These lead to better optimization strategies to operate closer to constraints and minimize energy use and maximize production. Similarly, the use of Coriolis meters on feeds and recycle streams to reactors, evaporators, crystallizers, centrifuges and dryers can infer component and solids concentrations for advanced control strategies and metrics.

Stan: The more measurements you have, the less guessing you have to do and the more process knowledge you can put into the APC. Wireless transmitters can enable a greater use of measurements by reducing wiring costs that are the biggest part of the installation cost for meters without sensing lines and analyzers without sample systems. You probably can install ten or more wireless pressure transmitters close-coupled or with diaphragms mounted directly on piping or equipment nozzles for the cost of one wired pressure transmitter with sensing lines. A more extensive use of pressure sensors would allow online monitoring of pressure drops and the tracking down of those fast insidious pressure disturbances from field pressure regulators. The source of disturbances to headers could be narrowed down to particular branches and supply, letdown and feed valves. Pressure drops could be monitored across catalyst beds, filters, column trays, heat exchangers and coils to provide an inference of fouling.

Greg: To see the future, consider the bench top, pilot plant or single-use bioreactor (SUB) that has probes to measure pH, dissolved oxygen, carbon dioxide, cell density (optical density), cell viability (capacitance) and substrate, nutrient and product concentration (NIR). Here wireless connections not only greatly reduce the installation costs, but also make the bioreactors on carts portable. What if lab analyzers on carts could transmit their results to control system data historians wirelessly?

Stan: Wireless transmitters enable the measurement of temperatures on rotating equipment. For example, they enable the measurement of product temperature on rotating reactors for alginate production and rotating calciners for industrial chemical production. Wireless transmitters enable more measurements in remote areas, such as tank farms and waste treatment facilities.

Greg: The use of more temperature measurements combined with flow measurements enables better process performance metrics to track down disturbances and provide online energy balances. For example, if the temperature is measured going into and out of a coolant coil for reactor or crystallizer along with the coolant flow, the heat removal rate can be computed online. It can be used for data analytics and an inferential measurement of crystallization or reaction rate suitable for control of the batch reaction profile. If an impurity or concentration change in the reactant feed occurs, it would be nice to at least know about it before the batch is done, and even better to compensate for it by controlling an inferred reaction rate or slope of the product concentration profile. (See “Unlocking the Secret Profiles of Batch Reactors.”) 

Stan: Without sufficient measurements, APC projects can become black holes, sucking in resources to a point of no escape and causing breakdowns in the laws of thermodynamics.

Greg: Talking about holes, Randy Reiss pulled a rabbit out of one by conceiving wireless access to the interface in the control room for an online principal component analysis of batch profiles and projection to a latent structures end-of-batch prediction he developed. This can be extended to automated maintenance systems, so when you are in the field tracking down problems, you can have access to metrics and analytics on the process and the automation system.

Stan: You also need good control valves to affect the process. I don’t mean on-off valves posing as control valves, but true throttling valves. If the valve was originally manufactured or marketed by a piping valve company instead of a control valve company, or has exceedingly good tight shutoff specs, the valve is best used for interlocks or sequences, not loops.

Greg: For more information on how to improve the foundation for fun and profit, see the recent presentation titled “Control Loop Foundation for Batch and Continuous Processes.” ( )
Can wireless rabbits be the key to your success? To help answer this timely question, we conclude with the top ten reasons from Randy Reiss why your New Year’s resolutions should include a wireless rabbit.

Top Ten Reasons Why You Will Go Wireless Next Year

10. Your astrologist predicted it, and she’s the one who also predicted that Britney would make a come-back...sorta.
9. Plant initiative to remove gofers is the most common reason for not getting tank level readings.
8. You’ve used the excuse that engineering and maintenance “got their wires crossed” too many times.
7. 50 million teens can’t be wrong!
6. Suppliers are threatening to discontinue wired transmitters.
5. Someone sold management that “once-a-second” data transmission over wireless is faster.
4. Last year’s wireless initiative was exposed to be a farce when Jeb’s cell phone died. This year, they’re watching!
3. Who, in their right mind, would trust a battery-operated wireless device to run the plant, unless it looks cool?
2. After a transposition error caused that incident last month, Rob just keeps repeating the phrase “Boy, you can clean that rabbit up, but you don’t know where it’s been,” over and over.
1. There are bad ways to realize that the phrase, “Turn on pump 42” sounds a lot like “Darn that horny toad!”