When I was the lead electrical and instrument (E&I) design engineer for what was then the world’s largest acrylonitrile plant, in my free time (when not going to concerts and bars in Harvard Square) I learned how to write a dynamic model of a compressor. Because one huge compressor was supplying the reaction air for the world’s largest catalytic reactors, I wanted to know more about surge control. I knew that reactor trips in other plants sometimes caused surge and the shutdown of other reactors. I got tables from the R&D group of the control valve manufacturer that enabled me to estimate the pre-stroke dead time and stroking time of large actuators for various positioners and boosters. I got the surge curve from the compressor manufacturer, and momentum balance equations from the literature. I can’t say it was a great model, but it got the attention of Monsanto Engineering Technology process control group leader Vernon Trevathan. One of Vernon’s areas of expertise was compressor surge control. I got a job offer to be a specialist in a department of about 75 of the world’s best in process, materials, and mechanical technology and in process modeling and control. If I had not done that compressor model, I would have stayed a design engineer and generalist. My focus would have been specifying and buying instrumentation and configuring systems. At the risk of self-congratulation, I probably would not have become a Senior Fellow, would have not been inducted into the Process Automation Hall of Fame, and would have not gotten the ISA Life Achievement Award.
The point is that engineers have a difficult time getting recognition unless they can do something special. Today, you don’t need to write programs but you do need to know how to use programs and how to configure applications. In particular, I recommend learning how to do dynamic modeling. In the building of the model you will gain a deeper understanding of the process and of control problems and solutions. Running the model for different scenarios and control strategies will enable rapid exploration and quantification of process control improvements (see Tips # 62, 98, and 99). New modeling packages with graphical configuration capability make the learning process as an extracurricular activity much more manageable. If you don’t have the skills and time, start with generic free online control loop simulators. If you build a virtual plant that successfully relates optimization benefits to golf, you may be invited to play golf with your CEO.
An emerging opportunity that is presently neglected is learning how to mine information from an Asset Management System (AMS) and an Instrument Management System (IMS). These systems are generally underutilized, perhaps because the expertise to get the most of out of these systems falls between groups. The smart instrument and positioner diagnostics and the AMS snap-ons for valve, vibration, ultrasonic, and failure analysis offer incredible capabilities for improving existing automation system performance.
There are many opportunities offered by advanced process control tools. Learning how to use an auto tuner and an adaptive tuner with loop metrics is an essential starting point of any career that seeks to go beyond E&I design and into loop performance.
Concept: Learn how to use dynamic modeling, system analysis, and advanced control tools in your free time. As you demonstrate improvements, management will provide the means for you to continue.
Details: Learn how to get the most out of an Asset Management System, an Instrument Maintenance System, and loop tuning and metrics tools. If you want to get more into designing process control strategies and improving the loop dynamics per Tip # 70 and optimizing the setpoints of control systems, learn how to use dynamic modeling tools and build a virtual plant. To move beyond single loop optimization, learn how to apply model predictive control (MPC).
Watch-Outs: Most out-of-the-box dynamic models do not include transportation and mixing delays, measurement lags, transmitter damping, measurement noise, analyzer sampling and cycle time, wireless update rates and trigger levels, and control valve backlash and stiction. Consequently, the total loop dead time is dramatically lower than it would be in an actual plant. Because the maximum controller gain and minimum reset time are inversely and directly (respectively) proportional to the dead time, controller tuning conclusions are thereby invalidated. If valve pressure drops are not updated by the use of a pressure flow solver and tables of valve flow coefficients cannot be entered, the installed characteristic, and, thus, the process gain and resulting controller gain, will not include the effect of valve gain.
Exceptions: If you have children under the age of 10, you may not have time for personal development unless you can rent a grandparent as offered in the ISA book The Funnier Side of Retirement for Engineers and People of the Technical Persuasion.
Insight: Distinguishing yourself as having special skills requires initiative because companies are often not interested in personal development unless there is a clear impact on the bottom line.
Rule of Thumb: Use models to better understand the process and control system relationships and learn how to use tools for instrument system and loop performance analysis to improve the calibration and maintenance of instruments and the tuning of PID controllers.
Look for another tip next Friday.