Three Recent Developments that Should Make PID Optimisation a Priority
Whether you tune underperforming PID controllers manually or with the help of software, one thing that both approaches assume is: You know which loop is the bad actor!
A recent post of mine focused on loop tuning best-practices and it outlined a simple method for modelling and tuning PID control loops. What that and other posts failed to mention was that underperforming loops are often hard to isolate.
While not many plants have PID counts that number in the 1,000s, most production sites have several hundred loops which can make it difficult to single out those bad actors. To make matters worse the interaction between different loops, particularly in complex production environments, can confuse things.
On countless occasions I’ve witnessed engineers correcting loops that simply exhibited the symptoms of poor control while overlooking the loop that were the actual root-causes. As shared in my Systems Thinker post I too have been guilty of that crime…before I leveraged the right tools and implemented a better procedure.
Relatively recent changes in the landscape for industrial automation technologies have made it easier to single out underperforming control loops. If you’re actively engaged in discussions related to IIoT and Industry 4.0, then these changes may be familiar. Regardless, here are a few innovations that have made it easier to identify control loop performance issues and to isolate the associated root-causes.
- Data availability
The mindset of most plant managers has changed. It wasn’t too long ago when they viewed data either as a luxury or solely as the means of satisfying a regulatory requirement. Today it’s seen differently. Access to good data is an imperative. Of course, it doesn’t hurt that the cost of sensors has plummeted. Since 1992 the average cost of a sensor has dropped over 95%.
Other innovations such as data historian and processing technologies have also made it easier and cost effective to both capture and capitalise on a plant’s abundant process data.
Performance details for each of 100s or even 1,000s of loops and how the associated dynamics are evolving are now readily accessible to most process manufacturers.
- Advanced analytics
Just like hardware there have been important advances in software as well. For years maintenance and asset reliability technology vendors have offered software application that perform advanced diagnostics and proactively look for trends of excessive wear and tear.
More recently process control technology vendors have introduced solutions that assess the performance of PID loops on a plant-wide scale. Using an array of KPIs control loop performance monitoring (CLPM) tools proactively identify issues including controller tuning, loop interaction, and even mechanical issues.
Helpful KPIs and advanced forensic tools make it easy for process manufacturers to correct for performance issues before production throughput and efficiency are negatively impacted. Stiction (or sticky friction) is one of the metrics that can viewed on the KPI dashboard.
- Known value
Manufacturers have been known to walk away from innovative technologies when the associated benefits weren’t quantifiable. Fortunately for vendors of CLPM solutions a study published in 2001 by the UK’s Energy Efficiency Best Practice Program documented a wide range of benefits linked to tuning PIDs routinely. It documented increases in throughput (2-5%) and yield (5-10%) along with decreases in energy consumption (5-15%) and quality issues (25-50%).
Since being published more and more case studies have popped up and they’ve provided more evidence of economic gains. These gains are more relevant today as manufacturers are coming under increased output, energy and margin pressures that have changed their perspective from “Make at All Costs” to “Make with Maximum Efficiency”
While the benefits of well-tuned control loops has always been assumed, a growing body of end-user case studies provides evidence of significant financial value.Most who follow advances in automation technologies would agree that a sea change in the pursuit of optimisation is underway. Data and associated processing power is now quite affordable. Advanced analytic tools are making it easy to uncover underperforming assets.
What’s more there’s growing evidence of their financial impact. If you consider that the PID controller provides the foundation for a production facility’s control infrastructure, then it’s easy to understand why optimisation should be a priority. Case closed!