Process Control Design and Practice

A Practical Approach to Control and Automation in the Process Industries for the Chemical Engineer

Introduction

There are a number of good textbooks for Chemical Engineers that teach process dynamics, linear control theory in the Laplace domain, and stability analysis.  This text will not repeat that work.  In practice, the work of process control engineers, and more generally chemical engineers applying control technology to their processes, dwells very little on process dynamics, which are often simple and slow (relative to mechanical or electrical systems), and as such are rarely modeled.  Control loop stability is achieved easily in the vast majority of cases with simple tuning rules, or simply trial and error tuning.  Instead, the work of process control is focused on designing and managing the hundreds of sensors, valves, feedback loops, sequences, interlocks, and constraints that work together to automate a given process.  Hundreds of these devices are often under the supervision of the one junior production or process engineer for that unit.  In many companies the process control specialist, often not a chemical engineer, and quite often a contractor, is an expert in the hardware and programming of the control system and its sensors, but not an expert in the process to be controlled.  She deals with thousands of I/O points and many different processes and process engineers; unless the process control specialist arrived at that position after experience as a process engineer (rare), that control engineer relies on that unit’s engineer(s) for process expertise and control system design.  This is quite unlike many aerospace, mechanical and electrical engineering control applications, where an engineer or a team of engineers is more likely to work with relatively few control loops at a time, often with very fast and difficult dynamics, which are well worth their time to model, analyze and optimize a closed loop solution for. Linear control theory applies well to electrical and aerospace problems, but generally not to the few difficult feedback control problems in chemical processes, which tend to be highly non-linear. As such the tools of linear control theory that can be taught to an undergraduate are of limited utility.  In practice those tools are more than is needed for the vast majority of feedback loops with simple dynamics, and less than required for the occasional difficult feedback control problem. And yet process control and automation is an important area of education for engineers. Most chemical engineers will work with automated processes, and must learn to design and operate that automation. Process Control as practiced today by chemical engineers is primarily a problem of design, not analysis.  How do process experts (chemical engineers) work with experts in electronic systems and programming (technicians, programmers, electrical engineers) to create a safe and efficient chemical process?  Do the process experts and the systems experts share the common vocabulary and knowledge base to communicate and design effectively?  In my experience, the answer is generally no.  Enormous amounts of time and energy are spent trying to bridge that gap between process experts and operators, on the one hand, and experts in the tools of control on the other.

The typical 21st century chemical process has an ever growing automation system.  Safety and environmental standards are steadily increasing the detail and complexity of the sequences and safeguards they demand, requiring additional monitoring and control. Intermediates and inventory are discouraged, leading to more dynamically linked processes. The cost of labor, the shortage of trained operators, and the brevity of their typical time in a position means that many units will be controlled by one operator, with limited experience, at a distance, from a control room.  That operator now monitors and assists more than she actually operates.  In fact, to maintain reproducible operation for quality and safety, shift operators are encouraged to let the control system run the plant as much as possible, helping by moving manual valves, making manual additions and supervising safety for maintenance personnel and utility operators.  The modern plant operator spends much of her time monitoring quality (sampling), communicating with other operators, interacting with the accounting and Quality Control lab software, and preparing and executing safety plans for unusual operations.  In such a plant, the productivity, quality, hazard avoidance and environmental protection are a team effort between control system and operator, largely with the operator managing the control system by exception, acting only when something appears to be wrong.  For both the design engineer and the operating engineer, the implication is clear.  To establish and improve the operation of the plant, the design and continuing improvement of the control system is critical. Today’s process control tools are capable of making a plant run like a high performance race-car, safer than ever before, yet at high levels of productivity.  Too many plants instead run with simple, very conservative programming that mimics pre-automation operation, built to satisfy engineers and operators who lack the skills and confidence to fully utilize or trust the tools purchased for them.  This is a huge hidden expense for the process industries. 

This text is intended to be used as part of a Process Control course in a Chemical Engineering undergraduate program, in the senior or perhaps junior year after the students are familiar with the standard unit operations.  Process Dynamics and Control (PD&C) is generally the required core course in most undergraduate Chemical Engineering curricula, a course that teaches how the differential equations describing the movement of mass and energy (some form of the Navier Stokes equation together with mass and energy balances) can be analyzed using Laplace transforms and the tools of linear control theory to predict open and closed loop dynamic response.  Given the constraints of time it will be difficult to do both that analysis course and this design course justice in one term.  I think they should be separate, with one as a core and the other as a technical elective.  My experience as a practicing chemical engineer leads me to the opinion that the material in this text will be useful to any practicing engineer. The material in PD&C, while sound pedagogically as an extension of Transport Phenomena courses, will be of use mostly to those who specialize in Process Control, or who pursue graduate studies in areas that touch on the areas of dynamic modelling and control.  My contention is that Process Control Design and Practice should be the core course taught to Chemical Engineers, with PD&C as an elective for those who wish to gain a deeper understanding of the topic, and the tools of linear feedback control theory.

This design course attempts to teach chemical engineers the communications tools and design principles to create and maintain a control system that will operate a process as specified by a process design, incorporating all of the many constraints of quality, safety, and operability. The chemical engineer will be taught to manage the complexity of hundreds of loops and thousands of I/O points by organizing a design hierarchy.  Students will be taught how automation design integrates with and executes process design and its various unit operations; they will become capable of communicating effectively about automation with operators, managers, chemists, control specialists, and the control system computer. They will learn how to maintain and control their processes at thermodynamic states that their process design specifies through the use of pumps, valves, and heat exchangers.  They will learn how a chemist’s batch recipe becomes a unit control recipe, and how that high level of abstraction is broken down into detailed equipment modules and control modules that control temperature, pressure, or batch and continuous charging. They will learn how a continuous quality constraint or a discrete safety constraint is translated from a process design concept to a constrained feedback loop or an interlock.

In my years working as a process control engineer, a process and design engineer, and then as a supervisor of both, the most common process control challenges I encountered were failures to communicate between chemical engineers and operations staff, on the one hand, and the people who created and maintained their control systems on the other. Chemical engineers often fear making changes to the control system.  That fear stems from a lack of understanding, the feeling that the control system is a “black box”, essential but not to be touched. Highly paid outside process control consultants spend days and weeks in control rooms and meeting rooms talking to engineers and operators trying to figure out what it is that they are being hired to do. Startups can be tortuous affairs as failures of communication are revealed.  Designing and operating a 21st century chemical plant requires a well-designed, flexible control system that collaborates transparently with operators and which changes as the plant and its processes change.  Today’s chemical engineer must understand how that system works at a high level, how to communicate in the design language of control to control specialists, operations staff and other professionals, and how to specify control system designs and changes to accomplish process goals.  The objective is not to make chemical engineers into control system programmers. In many cases that would be a waste of their talents; programmers can be hired as needed; most companies are short of capable engineers.  What is critical for the design and operation of a process is for the specification given to that programmer be precise and authoritative, communicating exactly how the chemical engineer needs the plant to run.  The objective is to make the chemical engineer an effective designer and manager of the automation system that runs her plant.

Dr. Tom Meadowcroft

A Note on Prerequisites and Educational Background

This text teaches how to automate chemical processes. The target student is a chemical engineer who has completed or is completing their courses in heat, mass, and momentum transfer, thermodynamics, reactor design and separations processes, or one who has finished their bachelor’s degree. It assumes that the student is familiar with the standard unit operations taught to chemical engineers, in particular heat exchangers of various sorts, reactors (batch and continuous), and distillation columns.  For example, the text will use the fact that the temperature and pressure of a flash tank boiling a solution of two liquids will determine the composition of the liquid and vapor product.  It will not try to teach why this is so, how distillation works, or how flash tanks and distillation are related.  That having been said, an electrical or mechanical engineer working in a process industry with a working knowledge of the processes in their plant should be able to follow the text and most of the examples presented.  The key process understanding required is the qualitative knowledge of how key process outputs will change when key mass and energy inputs are manipulated, and an ability to understand the implications of the conservation of mass and energy in practice.  A few questions to a chemical engineer colleague should help to get the non-chemical engineer past the conceptual gaps.  Beyond that, try “Chemical Engineering for non-Chemical Engineers”, by Jack Hipple, as a companion text.

Creative Commons

Process Control Design and Practice is licensed under Creative Commons (BY) 4.0.  All are free to use, share and modify the text and images herein.  Attribution to the original author must be given.  This work was not published as a normal textbook because I wanted it to spread as widely and quickly as possible, because any new text is quickly copied, and because students are forced to buy too many expensive texts by their professors.  Please contact Tom Meadowcroft if you wish to contribute improvements or additions and have them included as part of the “official” version.

Dedication

This work is dedicated to the memory of my late wife, Stephanie Lopina, PhD PEng, who worked in industry and academia as a chemical engineer, and was a leader and mentor to the women who followed her in our profession.  In her honor, and as a gesture to the women who work in the process industries and suffer a shortage of female mentors and exemplars, all of the engineers and operators in this text will be referred to using feminine pronouns.  If male engineers find that jarring or unrealistic, please seek me out to share your thoughts.  I will read them while thinking of the laughter such sentiments would have engendered from Stephanie.

TM 2021

Meadowcroft@Rowan.edu