Process control design and practice - A new approach to teaching control to chemical engineers

Modern process plants in the chemical, oil, food, and pharmaceuticals industry are highly automated to meet modern standards of safety, productivity, quality, and environmental protection. Chemical plant operators today generally only 'operate' by exception, stepping in when the automated system encounters a situation for which it is not prepared. As computers, sensors and automation have improved and decreased in price over the last 30-50 years, the process industries have invested heavily in automation. At the same time the one course that most chemical engineering undergraduates receive as part of the standard curriculum has hardly changed at all. Often titled 'Process Dynamics and Control' it teaches Laplace domain analysis of simple time domain dynamic models, an introduction to linear control theory and its tools, with the culmination generally being a derivation of how to tune a single loop controller to achieve stable feedback control for a general dynamic model. A newly minted chemical engineer working as a production engineer in a process unit will often be responsible for technical supervision of a process with hundreds of automated sensors, valves, motors and other I/O elements, including dozens of feedback loops. The stability of those loops will rarely be a concern (loop tuning can be taught empirically in an hour). A much greater challenge is to manage the complexity of this automated process. Why are all of these feedback loops present? How are sensors and control valves placed to absorb the available degrees of freedom? How do safety, quality, and environmental (SQ&E) goals translate to continuous and discrete constraints? How do you manage change and optimization in a plant operated by a control system? These are not primarily problems of analysis, but problems of design. Process Control Design and Practice is a course that tackles the design problem of process automation. It teaches how feedback controllers are applied to manage the thermodynamic degrees of freedom to maintain a single state. The complexity of sequential logic design problems (batch plants, startup/shutdown in continuous units) is broken down into abstraction levels of unit control, equipment module control, and device control to allow modular design. Students learn to think algorithmically, breaking down complex sequences into actions and logical transitions. Batch recipes formulated by a chemist are translated to sequential unit recipes for implementation at production scale. SQ&E goals become logical statements that interpret sensor data and actuate valves, motors, or sequences. The student is taught to translate concepts recorded in process flow diagrams, piping and instrumentation diagrams, chemists' recipes, operator procedures, and accountants' ledgers into a control system design specification that can be executed by programmers and the people who build and maintain your process. By making students think and communicate about how chemical engineering unit operations function in time, this design course links engineering science theory to industrial practice while preparing students to practice engineering on automated processes. Whether as an elective or a replacement for the standard dynamics course, this course will better prepare chemical engineers for 21^st century automated manufacturing.