2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

Lookout360
ReMix
Wiki ►► Communities
Control ► Control
Loop Tuning Theory
- Mipac ►
Intranet Lookout Newsfeed OneDrive Sites Kolyo Roydev

BROWSE PAGE FOLLOW EDIT

ReMix Communities Control Theory

Control Loop Tuning

Preface
Loop-Tuning – Art or Science

Tuning feedback control loops is something which we must all face from time to time. A key MIPAC activity involves
commissioning plant control systems either DCS or PLC. While the most common loops encountered during
commissioning are bread-and-butter flow and level loops occasionally more difficult tuning problems comes along.
For example how would be you go about tuning a pH loop, thickener bed pressure loop or a heat exchanger
temperature loop?

The ‘As-Found’ tuning method (described in the article) is a useful approach in these cases and should be in the
toolbox of all MIPAC commissioning engineers. The ‘As-Found’ tuning method is a semi-formal method developed
by Harold Wade. It was developed for use in the process industry and is typically used only for PI controllers.
However, this procedure lends itself exceptionally well to the HVAC control technician and engineer as well. The
primary advantages of this procedure are its simplicity, the relatively short time it takes to tune the loop, and the fact
that there is minimal disruption to the operating process.

Even with these limitations the ‘As-Found’ method is probably worth a try if only to get a set of starter tuning
parameters. As Found Method Here

External Information
Wikipedia article on the PID controller
Wikipedia article on the PID controller, subsection on loop tuning

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 1/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

Wikipedia article on Control Theory

Main Article
Control Loop Tuning
Once a process and associated control components are installed, the control loop must be calibrated to perform in
the manner intended by the designer. This is referred to as control loop tuning. All material prior to this section is
intended to prepare you for the tuning process.

Before you set out to tune a control loop, you need to ensure the following.

All sensor/transmitters must be calibrated to manufacturer’s specifications

The end device must be properly sized for the process. This implies all parts of the system must be linearized to
minimize variation in process gain over the normal operating range of the process
Valves and dampers must be properly serviced to minimize sticking friction
All linkages must be properly configured to eliminate play

If you do not take the time to ensure all elements of the process and control loop are properly configured, the time
taken to tune the loop will likely be wasted.

Controller Structures and Process Models

The ideal controller equation for a PID controller was noted in the last section. However, there are a number of
ways any given controller manufacturer may implement this algorithm. According to the book “Handbook of PI and
PID Controller Tuning Rules”, there are five different PI controller structures and 26 different PID controller
structures. Various manufacturers developed these structures for different applications.

Not only are there different mathematical implementations of the ideal controller equation, but also, each process
being controlled may have a different mathematical model. Even if the end result of two different processes is
similar, there is likely to be different values of dead time and time delay. These similar process differences may
require different tuning parameters, even if identical controllers are used, so as to allow the control loop to respond
as desired.

When one considers the number of controller structures available, coupled with the large number of process models
in existence, one can surmise that different tuning rules may be necessary for each unique application. In fact,
dozens of tuning rules have evolved over the years to handle such combinations of different controllers and
different processes. Some of these tuning rules are quite formal and highly mathematical in nature. Others are
strictly trial-and-error and should generally be avoided. Between these extremes are semi-formal methods that
combine a structured trial-and-error approach with minimal mathematic calculation. These semi-formal methods
tend to work well in the field. They are a good compromise between the more sophisticated mathematical methods
and the time required to perform the less accurate trial-and-error methods.

For our purposes, we will briefly outline a common trial-and-error method, but will concentrate on the Zeigler-
Nichols continuous cycling method, two approaches to a process reaction method, and the IAE (Integral of Absolute
Error) method. We will also address a structured trial-and-error method for PI control loops known as the As-Found
method.

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 2/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

One should also realize each tuning method will likely produce a different set of tuning parameters. Even different
persons using the same tuning method on the same loop may arrive at different parameters. This is simply because
two-mode and three-mode controllers have numerous response curves occurring at different oscillatory
frequencies. What we wish to do is determine a set of tuning parameters that is optimal for the process at hand.

Trial and error - OR - As Found Method

The following method is strictly a trial-and-error method. The primary advantage of this method is that it is easy to
learn. The process is as follows.

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 3/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 4/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

Closed Loop Method – Ziegler-Nichols

The engineering team of Ziegler and Nichols developed a modification of the continuous cycling method in the early
1940’s. This method results in a system that is tuned for quarter-wave damping.

Quarter-wave damping is illustrated in Figure 1. It is considered a compromise between excessive oscillation and
excessive settling time.

Quarter-wave damping refers to a tuning method in which control loop oscillations damp in such a fashion that the
ratio of overshoot of any given wave is ¼ that of the previous wave. One must understand not all processes should
be tuned to quarter-wave damping. Although this is the goal of most of the tuning methods presented in this paper,
there are other tuning methods that produce other results for other purposes.

To employ the Ziegler-Nichols approach to loop tuning based on continuous cycling, follow the first three steps
listed above under the trial-and-error method. The gain at which continuous cycling occurs is known as ultimate
gain (Ku). Determine the period of the oscillations as shown in Figure 2. Using the equations summarized in Table
1, calculate the appropriate controller tuning parameters necessary to achieve quarter-wave damping. Input the
calculated values of gain, integral time constant, and derivative time constant into the controller. Test and adjust as
necessary.
https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 5/13
2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

Process Reaction Methods

Process reaction methods of loop tuning are accomplished with an open loop. In other words, the feedback portion
of the loop is broken. When a step disturbance is injected to the system, the system will respond in accordance with
the time elements associated with the loop and reach a new point of equilibrium. An analysis of this response
allows one to determine the appropriate controller tuning parameters to achieve quarter amplitude damping. The
following steps describe the data collection phase of the open loop analysis.

Place the controller in manual mode. Make sure the process is lined out. Make a step change in the controller
output. The step change should be as small as possible, but should not exceed a 10% change. This change will
force the process to change in a fashion similar to Figure 3.

If the controller cannot be placed in manual, you must disconnect the sensor input. Connect an artificial signal to the
controller input terminals. Bump the controller output as described above.

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 6/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

Calculate a process gain, a process dead time, and a process time lag as indicated in Figure 3. Make sure you
draw the tangent line to the steepest part of the reaction curve. Use the relations in Table 2 or in Table 3 to
determine the appropriate controller tuning parameters.

Enter the tuning parameters, place the controller back in ‘Auto’, and perform a closed loop test. The test should
result in quarter-wave damping.

Tweak the settings as necessary.

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 7/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

Although there are a number of process reaction methods available, the Ziegler-Nichols method and the Cohen-
Coon method are debatably the most common. The goal of both methods is quarter-wave damping. The primary
difference is the Cohen-Coon relationships represent a refinement of the Ziegler-Nichols process reaction method.

Advantages

There is no need to perform multiple trials. Generally, a single trial is all that is necessary to obtain proper
controller settings.
Controller settings are relatively easy to calculate
You do not operate at the stability limit of the process
There is less chance of saturating control loop components
Determination of process dead time and process time lag often provides insight into the process not otherwise
obtained with a closed-loop test

Disadvantages

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 8/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

Since the test is conducted without controller feedback, any significant change in process load may provide
erroneous results
If the process is noisy, it may be difficult to determine which part of the reaction curve has the steepest slope so
as to construct your tangent line. In such cases, multiple trials may be necessary.
These relations are considered valid only if .

Chien Method

All of the aforementioned methods result in quarter-amplitude damping. Although this is purported to be a valid
compromise between stability and speed of response, it is an aggressive tuning method not suitable for all
applications. When one has an application in which quarter amplitude damping is inappropriate, a different method
may be more suitable. One such method is proposed by Chien. This method is an open loop method in which the
goal is either 0% overshoot or 20% overshoot. To apply the method, perform an open loop test as described above.
Determine the dead time and time lag. The Chien method is valid only if . Once you determine

the process characteristics fall within this range, you can determine the controller settings as outlined in Table 4.

The calculations shown in Table 4 are for a non-interacting controller in which the proportional, integral and
derivative logic is arranged in parallel then added to provide a controller output. Most controllers use some form of
interacting model in which the proportional, integral, and derivative logic is placed in a series arrangement. For the
case of an interacting controller, Chien was modified by O’Dwyer to provide tuning parameters for a non-interacting
PID controller. As with Chien, O’Dwyer is valid only if .

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 9/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

IAE Tuning Method

The Integral of Absolute Error (IAE) tuning method is a rather sophisticated method that attempts to provide tuning
parameters that are unique for a set point disturbance vs. a load disturbance. When a disturbance is applied to a
process, the control loop responds and attempts to compensate for that disturbance. Until the system is brought
back under control, there exists an error, at any instant of time, between the control point and the set point. The IAE
method attempts to minimize the sum of errors over any specified period of time. Mathematically, this method is
defined as:

Since a set point disturbance is quite different from a load disturbance, Tables 5 and 6 provide different parameters
for each condition. As with the previous methods, the IAE method is applicable only when .

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 10/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

‘As-Found’ Tuning

The ‘As-Found’ tuning method is a semi-formal method developed by Harold Wade. It was developed for use in the
process industry and is typically used only for PI controllers. However, this procedure lends itself exceptionally well
to the HVAC control technician and engineer as well. The primary advantages of this procedure are its simplicity,
the relatively short time it takes to tune the loop, and the fact that there is minimal disruption to the operating
process.

The following description of the process, and the flow chart used to graphically depict the process, is taken from a
computer simulation laboratory exercise developed by Harold Wade. This procedure has been published in ‘InTech’,
the member journal for the Instrumentation, Systems, and Automation Society (ISA).

As excerpted from a student lab exercise written by Harold

Wade
Quite often a control systems engineer or instrumentation technician is called upon to improve the behavior of a
loop that is currently in operation, but without resorting to either the open loop or closed loop testing methods.
Assuming that the loop is not behaving acceptably at present, and that process and equipment problems (e.g.,
sticking valve) have been eliminated, then most persons resort to “trial and error” tuning. For novice tuners, this is
often simply an exploratory procedure; “How about changing this knob in this direction and see what happens”.

This laboratory exercise presents a method for directed trial and error tuning, where each tuning parameter change
is made for a deliberate reason. The objective is to go from the current unacceptable behavior to acceptable
behavior as efficiently as possible; i.e., in the fewest number of tuning parameter changes.

This method is based upon the premise that if a PI controller, controlling a self-regulating process, is well tuned
(that is, exhibiting a slightly underdamped oscillation with a quarter wave decay), then there will be a predictable
relationship between the period of oscillation (P) and the integral time (TI). This relationship (stated in three different
ways) is:

This premise leads to the following rule-based procedure:

1. If the loop is not oscillating, increase the gain, say by 25 to 50%.
2. If the loop is oscillating then:
1. If the Period is between 1.5 and 2.0 times the integral time (or the period-to-reset ratio is between 1.5 and
2.0), then either increase or decrease the gain as required to obtain the desired decay ratio (such as
quarter wave damping)
2. If the Period is greater than 2.0 times the integral time (or the period-to-reset ratio is greater than 2.0), then
choose a new integral time according to the criterion: TD3
3. If the Period is less than 1.5 times the integral time (or the period-to-reset ratio is less than 1.5), then

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 11/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

1. If the decay ratio is greater than 1/4, then decrease the gain, say by 25 to 50%, depending upon how
much the decay ratio exceeds 1/4.
2. If the decay ratio is less than, or approximately equal to, 1/4, then choose a new integral time, using the
criterion given in 2.2.
3. After each adjustment, make a slight set point change to test the response to the latest combination of tuning
parameters.

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 12/13

2/21/2020 Control Theory - Control Loop Tuning - Mipac Intranet

The essence of the rule-based procedure listed above is shown in flow chart form:

https://remix.mipac.com.au/Communities/ControlTheory/Wiki/Control Loop Tuning - Mipac Intranet.aspx 13/13