What is the Purpose of a PID Controller’s Integral Term? Why is PI Control So Widely Used in Industry?

    While the Integral Term Introduces a Degree of Complexity the PI Form of the Controller is the Most Widely Used in Industry

    In the realm of process control it makes complete sense that the primary goal is – you guessed it – to control the process.  For decades the PID controller has proven a reliable tool for fulfilling that objective.  Even so, there are various forms of the controller that can be used and each has its unique performance attributes.  In spite of the added complexity associated with the Integral Term the PI controller is the form most widely used in industry.  It delivers improved Set Point Tracking relative to the Proportional-Only Control and is well suited for the disturbances that afflict many industrial process control applications.


    Proportional-Only – or P-Only Control – was covered in a previous post
    . While P-Only Control is easy to implement and offers advantages in certain application, it clearly has limitations. Chief among those limitations is its propensity for Offset – the difference between a control loop’s Set Point and its input (i.e. Error) which often results from a sustained disturbance. Since many industrial applications are both prone to frequent disturbances and require tighter Set Point tracking P-Only Control is often insufficient.

    The Integral Terms continually sums Error, pushing or pulling the process until the Process Variable equals Set Point.

    We found a publication that you might enjoy:

    Why the Model Matters

    The PI form of the controller provides a valuable correction for Offset. Rather than responding to the value of Error at a specific time the Integral term continually sums Error, either adding Error to the Controller Output (CO) when below Set Point or subtracting Error when the CO is above Set Point. The Integral term will only remain constant when the Process Variable (PV) equals Set Point. As such a controller’s integral action can be viewed as accumulating influence over time based on how long and how far away the measured PV is away from Set Point, and it serves to push or pull the PV back in line with Set Point.

    Across the process industries PI Control is the dominant form of the PID in use today. As effective a tool as it may be for countering Offset, PI and the use of Integral still present some challenges:

    Added Complexity

    There are no two ways about it – introducing the Integral term adds to the complexity of control loop tuning. The two terms – Gain and Integral – interact with each other, making it a challenge to arrive at values which are “best” for meeting the loop’s unique control objective. Even experienced practitioners can be fooled into adjusting the wrong term when their analysis is based on visual inspection of trend data.

    Reset Windup

    It’s possible that Integral Error can grow too large and become nonsensical. Consider an Integral Error value that requires a given control loop’s Final Control Element – a valve per se – to open 120%. Such a condition is termed “Windup”. In such a situation the controller cannot regulate the process until the Error changes sign and shrinks sufficiently.

    In spite of the added complexity PI Control is by far the dominant form of the PID used in industry. It is highly effective at correcting for Error associated with Offset and provides superior performance from the standpoint of Set Point tracking.

    Are you ready to bring your plant to the next level?