Our company has spent a lot of time and money to improve control loop performance in the field. We now have reliable controllers that will allow us to move our setpoints closer to manufacturing limits. Unfortunately, just because we CAN run our plants nearer the constraints doesn’t mean that every operator actually WILL. Even the implementations of expensive Advanced Process Control projects rely on the board operator enabling that control scheme. With all of the information that is brought into the control room through the Distributed Control Systems (DCS), the operator can get information overload. This can result in the operator running the plant at a comfortable level that allows some room for excursions. That way, if the DCS operator initially misses a change in one variable, they have a little room for maneuvering before causing a plant upset. Unfortunately, this is not the most profitable way to run a plant. To avoid falling into these “comfort zones”, we needed a method to monitor how closely to the plant’s ideal target we were running. This method needed to be easy to follow, inexpensive, and allow for target changes. We decided to use the two most basic tools of the PI System: PI ProcessBook and PI DataLink. The program we implemented is called the Critical Variables Program.
Step 1: The production units established a list of critical variables to use as key indices for the plant. For each variable, the list included a TARGET range (green), HI/LO range (yellow), and HIHI/LOLO range (red). Some of the variables were critical to quality and efficiency while others were critical to mechanical aspects of the plant (prevent plugging, corrosion, etc.). Step 2: PI ProcessBook displays using multi-state color changes were created to show an instantaneous picture of where the plant was running. Ideally, all parameters would be green. Any parameters that were yellow or red indicated that some action was needed to bring these parameters back into the green target range. Step 3: Daily PI DataLink reports using the time filter function were created to indicate how closely the plant ran to the targets. The percent of time that each variable spent in the green, yellow, and red ranges was listed. This allowed the supervisors and management to monitor where the plant was run on off-shifts. By focusing on the key indices of the plant, the Critical Variable Program gave the operators a concise map for running the plant and an immediate visual notification when they were “off-target”. Also, the supervisor follow-up on any parameters that were not maintained within target ranges re-enforced the importance of these targets to the operators. Results: In its first year, the Critical Variables Program in one of the production units resulted in a steam savings of over $1.5M and increased the rework capacity of that plant. It is now being implemented in several other plants onsite. No additional investments were required and no special programming skills were used - just the basics of PI ProcessBook and PI DataLink.