How Does Loop Monitoring Help with Root-Cause Analysis?

For most individuals, a house is the most expensive asset that they’ll ever purchase. And while living there a homeowner can expect a significant problem or two to arise whether that problem involves a leaky roof, termite damage, or some other costly issue. A downside of such problems is that the damage is typically and irrevocably done before the homeowner encounters the first visible signs that the issue even existed.

Manufacturing plants are similar albeit significantly larger, significantly more complicated, and more thoroughly analyzed – significantly so.  And even though most are fully staffed around the clock, problems do arise.  Things do go wrong.  In general the technologies and staff involved in maintaining control and production are focused on attributes that are essential to the plant’s financial viability.  Like a homeowner who pays the mortgage on time and tends to the garbage on a daily basis, production staff tend to focus on inputs and outputs.  They keep an eye on alarms.  And that means they’re probably overlooking other important details.

While a growing community of manufacturers supplement their production control systems with Control Loop Performance Monitoring (CLPM) technology, more companies should realize what they’re missing in terms of valuable, plant-wide performance diagnostics. Here are three (3) ways in which the root-cause analysis capabilities within a CLPM solution provide value:

Blind Man’s Bluff

Negative performance trends are routinely undetected by the human eye mostly due to the nature of the data itself.  It’s not unusual for an operator to overlook a gradual increase in oscillation or to loose count of the Set Point changes that occurred during a given shift.  On the one hand, shifts in the process data are often hard to detect as the data can be highly erratic and multiple trends are often layered on top of each other.  On the other hand, the number of Set Point changes means very little to an operator.  CLPM tools are designed to capture those subtle, prolonged shifts in performance as well as event-based data that other production staff need to track.

The Doctor Is In

By addressing the symptoms rather than the root-causes, performance issues will most likely persist or potentially worsen.  Due to the highly interactive nature of most production processes it is easy to mistake a symptom for a root-cause.  Interestingly the identity of most problems is usually hidden in plain sight – right there in the data.  CLPM tools utilize features like Power Spectrum and Cross-Correlation to clarify the relationships that exist between and among PID control loops.  What’s more those same tools can simplify the isolation of an issue’s root-cause.  No more fixing things that aren’t broken.

Charting a Course

Just because a practitioner is aware of a problem doesn’t mean that the solution is clear.  Not every issue can be solved by tuning.  Irregular valve behavior can be the result of any number of different things.  And countless performance issues are the result of poor architecture.  Again, many controller performance monitoring technologies can help with both diagnosis and recommendations for corrective action.  They provide a standardized framework for assessing issues and developing  appropriate course of action so that production is quickly returned to optimal levels.