I had the pleasure of chatting with Sigifredo Nino, an electrical engineer with more than forty years of experience in the process control industry. His career began in the 1980s at a coal power station in Colombia, after which he spent nearly a decade working at a pulp and paper mill. Later, Sigifredo moved to Canada, where he became a leading expert consultant in process control. Our conversation delved into his professional journey, various projects, and significant contributions to the field.
The core of a coal fired power plant is the boiler, where energy from coal-heated water is transformed into mechanical energy that drives turbines, ultimately generating electrical power. According to Sigifredo, starting a career in process control at a coal power plant is ideal due to the complex and interacting multivariable processes involved.
Using the pneumatic-based automation technology available at the plant, Sigifredo gained a deep understanding of how controllers and other devices function and interact. He then optimised these systems, applying a technique he used throughout his career. This approach involved creating Piping & Instrumentation Diagrams (P&ID) by simplifying them into a sketch that emphasised only the essential components and their interactions.
One of the essential control loops in a power plant is the drum level control, which is designed to maintain the desired water level within the drum while adapting to variations in steam flow that exits the boiler to drive the turbines. During this period, Sigifredo gained familiarity with various control loop strategies, including the widely adopted three-element feedwater regulator, which is particularly effective for handling significant fluctuations in steam demand. This critical process involves feedforward control to adjust the boiler feedwater flow in response to changes in steam flow at the outlet, while feedback control monitors and stabilises the drum level. The three-element level control system integrates cascade and feedforward control, where the inner loop manages the feedwater, steam flow acts as the feedforward input, and drum level adjustments are made to fine-tune and maintain the desired level.
Sigifredo developed a disciplined study routine, immersing himself in books and technical articles. During his early career, while living in the on-site accommodation provided by the remotely located power plant, he devoted his time to studying both before and after his work shifts. One of his favourite books during this period was the second edition of “Process Control Systems” by Greg Shinskey. Shinskey's approach, which emphasised first principles, a deep understanding of chemical processes, energy and material balances, and his unconventional time-domain analysis over traditional frequency-domain methods, was particularly influential in Sigifredo's learning journey.
A few years later, Sigifredo had the opportunity to transition to a new company and work at a pulp and paper mill. This proved to be an invaluable learning experience for his development in controls engineering, as the processes involved were both complex and challenging. In the winter of 1993, Sigifredo had the chance to meet one of his process control heroes, Greg Shinskey, in person while attending Shinskey’s “Process Control Systems” seminar in the USA. They quickly formed a bond, and Shinskey became a mentor to Sigifredo, guiding him in his professional growth.
Their relationship evolved into a collaboration and mentorship that spanned decades. When Sigifredo moved to Canada in the late 1990s, their partnership grew even closer due to the reduced distance between them. As a process control expert, Sigifredo frequently tackled complex and exciting projects, providing the perfect opportunity to apply and explore some of Shinskey’s ideas in real-world scenarios.
One notable example of their collaboration is documented in the paper “PID-Deadtime Control of Distributed Processes” by G. Shinskey. While Sigifredo was working on a power generation problem, which involved controlling the superheated steam temperature of a 500MW power boiler, Shinskey had an insight that for distributed interacting lag applications with significant dead time, such as this temperature control, introducing a time-delay to the integral term of the Proportional-Integral-Derivative (PID) algorithm could enhance the control system's response to load disturbances and performance. This approach was unconventional, as textbook PID control typically relates these three elements—P, I, and D—to the present error. Specifically, P is directly proportional to the current error, I is proportional to the integral of the current error over time, and D is proportional to the rate of change of the current error. The idea of using a delayed error was unusual, but the results were astonishing and twice as effective as a conventional PID controller when all controller tuning settings were scheduled in function of the steam flow.
As a process control consultant working across various industries, including oil and gas, petrochemical, mining, and pulp and paper, Sigifredo decided to establish his own engineering company in 2011. One of the areas where he has gained significant recognition is in the control of distillation towers in oil refineries and petrochemical plants. Although he encountered these applications later in his career, the wealth of experience he had accumulated allowed him to thoroughly understand the processes and apply the lessons learned from his mentor. He has authored papers, published successful case studies, and contributed to chapters on “Distillation Control” and “Refinery Control” in the renowned Process / Industrial Instruments and Controls Handbook, 6th Edition, by Gregory K. McMillan, P. Hunter Vegas.
“There is no substitute for process knowledge, and certainly none for common sense,” Greg Shinskey once said. This principle became the most valuable lesson in process control that Sigifredo learned from his mentor.
Links:
> “PID-deadtime control of distributed processes” paper by F.G. Shinskey
No comments:
Post a Comment