Why Does Control Loop Performance Change Over Time? What is Time Variant Behavior?

Changes To Regulatory Control Loop Performance Are Occurring Everyday…Right Under Your Nose!

You visit a doctor annually for a physical examination.  Your car gets an oil change every 3,000 miles.  Your company assesses profit versus loss every 3 months.  The list of routine “check ups” goes on and on.  Why?  Because such things are important and they change over time.  It’s valuable both to appraise their relative health and to determine what adjustments may be necessary.  As the foundation for regulatory control at most production facilities, PID control loops are no different.

Staying on top of control loop performance is a demanding, oftentimes thankless job. That is particularly true since the typical process manufacturer utilizes 100s if not 1000s of PID controllers and the job of monitoring and tuning control loops is usually just one of many other responsibilities that go unnoticed until something goes wrong. Regardless of the time involved and the competing demands of the job, there are obvious benefits to maintaining a healthy production environment. The most obvious benefit is that things do indeed change every day and those time-based changes can put production at risk if they’re left unchecked.

A few things to consider about processes and how they change with time:

  • Surfaces Foul or Corrode

As a process’ environment changes it can be expected that the associated dynamics will follow suit. Each individual change in isolation may be insignificant and it may not warrant immediate attention. In the aggregate, however, the changes to a process’ environment over time can be material and undermine control.

  • Mechanical Elements Wear Out

Mechanical elements like pumps and motors perform differently when they’re first installed versus when they’ve been in service for years. Over time an instrument’s tight, accurate responses to each change in Controller Output ultimately give way to increasingly sluggish, inexact responses. Steady erosion in performance should be expected and actively monitored.

  • Feedstock Quality Varies

More often than not the raw material used in most production processes is a commodity and it can originate from almost anywhere. Although produced to specification the feedstock’s content and corresponding quality can vary, resulting in subtle differences in process performance. The impact of those differences can be expected with every new arrival at the loading dock.

  • Environmental Conditions Change

Differences in environmental conditions can affect a production environment and its processes. Whether variation in an environment’s humidity that affects the stickiness of instrumentation or differences in heat that influence temperature control, the process’ dynamics may differ. More often than not these influences are seasonal in nature and can be anticipated.

Staying on top of time variant behavior and its impact on regulatory loops is a necessary aspect of production control and optimization. As the expression goes: What gets measured, gets managed. At a minimum practitioners should include regular checks of their facility’s control loop performance as part of a scheduled maintenance program. As covered previously, technologies that actively monitor control loop performance and identify changes that may negatively affect production are available.  Either of these options (the manual and the automated) facilitate regular appraisals so that adjustments can be made before those little changes become big problems.

These resources offer related content:

shutterstock_1938620782 (1)

PI Asset Framework: A CLPM History Lesson

When does History matter? Well, in 2017 OSIsoft introduced Control Station as a valued member of the Partner EcoSphere. At the time Control Station’s PlantESP was the first control loop performance monitoring (CLPM) solution to integrate seamlessly with the PI System using the PI SDK....
http://controlstation.com/wp-content/uploads/2019/07/cropped-Yoko-WP-3.png

Six Steps to Improved PID Tuning

Click Here for the White Paper
Iconic-Mark_Inverted-Color

Using the Oscillation Metric to Uncover Tuning Issues

  The Oscillation Metric Can Identify a PID Controller that Requires Tuning, But More Often Than Not it Provides an Indication of Loop Interaction or Mechanical Issues. There’s a wealth of information available in most every data historian. The data can be used to evaluate...

Still looking for more?

Now that you’ve gotten the basics, connect with our team to learn how our people, processes and technologies can help you optimize.