The following discussion is part of an occasional series, "Ask the Automation Pros," authored by Greg McMillan, industry consultant, author of numerous process control books, and 2010 ISA Life Achievement Award recipient. Program administrators will collect submitted questions and solicit responses from automation professionals. Past Q&A videos are available on the ISA YouTube channel. View the playlist here. You can read all the posts from this series here.
Looking for additional career guidance, or to offer support to those new to automation? Sign up for the ISA Mentor Program.
All of the tension regarding AI and the future reminded me of my first ISA book, from back in the late ‘70s and ‘80s, called Modernizing Control Systems. At the time, the influx of on-process computing and broad communication power had illuminated a future much different than our present. I was doing a lot of work for a major oil company in those days when they asked me to take my education and experience (from the ten years since my BSEE) to think about where this was going to take the industry and how that change might affect various groups within the company.
“Right now,” feels a lot like, “back in those days." In 2021, I was doing a monthly blog for ISA on a variety of subjects, and someone suggested discussing the change described in Modernizing Control Systems with the contemporary conditions and direction of the same groups, factors, and issues. “A 1978 Perspective on Control Systems Future” was published in ISA Interchange on 28 September 2021. It juxtaposed the view from 1978 with then-current perspective.
Before we get into a broader and more difficult topic of the future of modernization, let us start with a view of the present best practices by the question: When are we modernizing control systems?
My first brush with the modernization of a control system occurred in 1984 when I was recruited to work on the replacement of an analog control system with a modern DCS. Although the analog system was a capable product, reliability, and support had become an issue. Since the legacy system had been blamed in part for a total plant shutdown and delayed recovery, the decision to modernize the controls was not a difficult one to make. Modernization of a control system is not a small undertaking nor is it without significant cost, making an informed decision essential.
Where the legacy system is in its lifecycle, is a factor in making a decision regarding modernization. Near the end of product life, vendor support and the availability of system components may not be sufficient to maintain the system in a serviceable state, and may directly impact plant availability and production. Other reasons for considering modernization include:
Functional limitations—where the opportunity to improve process control (and plant profitability) is restricted by the ability of the installed system to incorporate advanced control applications, or when the process or operating strategy is changed and the ability to add functionality is restricted.
For safety shutdown and control systems, the legacy installation may not be able to meet corporate or legislative requirements for safety integrity, and the practical targets for system availability.
In instances where there are plant expansions, issues of interoperability with systems installed with the expansion may suggest a modernization for operational reasons. A plant expansion project may also suggest a modernization of the systems controlling existing plants to reduce the number of disparate systems to make technical support easier. Human Factors consideration may also suggest a modernization of the legacy system so that HMIs are consistent in “look and feel” to reduce operator error in emergent situations.
Not to be overlooked in justifying a modernization is the opportunity to include a “digital twin” which may not be available for the legacy system. This can benefit operator training, the training of in-house technical support, and not least, the proving of control strategies.
Once the project reaches completion, we typically experience heightened reliability and safety, alongside improved tools.
Even if the primary aim wasn't performance enhancement, it should be an integral part of the project. Usually, we achieve better interoperability and introduce additional tools and functions, such as MPC, fuzzy logic, and AI tools.
In practice, performance issues often arise when dealing with diverse algorithms and functions. Consequently, it's imperative to include a thorough review of these aspects in the project, even addressing seemingly simple elements like PID algorithm conversion. Consider instances where an old legacy system might have utilized parallel PID, while the new system offers Series or Ideal algorithms. Ensuring a smooth transition involves the essential task of converting previous parameters to align with the new setup. Additionally, configuring feedforward or override schemes may differ, requiring careful adjustment.
Lastly, the significance of interlocking and alarming should not be overlooked, as these aspects can pose challenges post-modernization of a control system.