Tuning PID loops for level control

One-in-four control loops are regulating level, but techniques for tuning PID controllers in these integrating processes are not widely understood.

10/30/2014


Figure 1: A typical level control loop. Here the level measurement is sent to the controller, which regulates a valve on the tank’s outlet. Courtesy: DataforthSince the first two PID controller tuning methods were published in 1942 by J. G. Ziegler and N. B. Nichols, more than 100 additional tuning rules have been developed for self-regulating control loops (e.g., flow, temperature, pressure). In contrast, fewer than 10 tuning methods have been developed for integrating (e.g., level) process types, though roughly one-in-four industrial PID loops controls liquid level.

The original Ziegler-Nichols tuning methods aimed for a super-fast response capability, which was achieved at the expense of control loop stability. However, a slight modification of these tuning rules improves loop stability while still maintaining a fast response to setpoint changes and disturbances. As most process experts will agree, stability is generally more important than speed.

Applicable process types

This modified Ziegler-Nichols tuning method is intended for use with integrating processes, and level control loops (Figure 1) are the most common example.

Figure 2: An integrating process is stable only when the controller is at the specific equilibrium point. In this case, the controller output needs to be 40%. Courtesy: DataforthUnlike a self-regulating process, an integrating process will stabilize at only one controller output, which has to be at the point of equilibrium. If the controller output is set to a different value, the process will increase or decrease indefinitely at a steady slope (Figure 2).

Note: This tuning method provides a fast response to disturbances in level and is therefore not suitable for tuning surge tank level control loops.

The modified Ziegler-Nichols tuning rules presented here are designed for use on a non-interactive controller algorithm with its integral time set in minutes. Dataforth's MAQ20 industrial data acquisition and control system uses this approach as do other controllers from a variety of manufacturers.

Procedure

To apply these tuning rules to an integrating process, follow these steps. The process variable and controller output must be time-trended so that measurements can be taken from them, as illustrated in Figure 3.

Step 1. Do a step test

a) Make sure, as far as possible, that the uncontrolled flow in and out of the vessel is as constant as possible.

b) Put the controller in manual control mode.

c) Wait for a steady slope in the level. If the level is very volatile, wait long enough to be able to confidently draw a straight line though the general slope of the level.

d) Make a step change in the controller output. Try to make the step change 5% to 10% in size, if the process can tolerate it.

e) Wait for the level to change its slope into a new direction. If the level is volatile, wait long enough to be able to confidently draw a straight line though the general slope of the level.

f) Restore the level to an acceptable operating point and place the controller back into automatic control mode.

Step 2. Determine process characteristics

Figure 3: Measurements used for tuning a level control loop. These are what you’ll need to do the calculations. Courtesy: DataforthBased on the example shown in Figure 3:

a) Draw a line (Slope 1) through the initial slope, and extend it to the right as shown in Figure 3.

b) Draw a line (Slope 2) through the final slope, and extend it to the left to intersect Slope 1.

c) Measure the time between the beginning of the change in controller output and the intersection between Slope 1 and Slope 2. This is the process dead time (td), the first parameter required for tuning the controller.

d) If td was measured in seconds, divide it by 60 to convert it to minutes. As mentioned earlier, the calculations here are based on the integral time in minutes, so all time measurements should be in minutes.

e) Pick any two points (PV1 and PV2) on Slope 1, located conveniently far from each other to make accurate measurements.

f) Pick any two points (PV3 and PV4) on Slope 2, located conveniently far from each other to make accurate measurements.

g) Calculate the difference in the two slopes (DS) as follows:

DS = (PV4 - PV3) / T2 - (PV2 - PV1) / T1

Note: If T1 and T2 measurements were made in seconds, divide them by 60 to convert them to minutes.

h) If the PV is not ranged 0%-100%, convert DS to a percentage of the range as follows:

DS% = 100 × DS / (PV range max - PV range min)

i) Calculate the process integration rate (ri), which is the second parameter needed for tuning the controller:

ri = DS [in %] / dCO [in %]

Step 3. Repeat

Perform steps 1 and 2 at least three more times to obtain good average values for the process characteristics td and ri.

Step 4. Calculate tuning constants

Using the equations below, calculate your tuning constants. Both PI and PID calculations are provided since some users will select the former based on the slow-moving nature of many level applications. 

For PI Control                       

Controller Gain, Kc =  0.45 / (ri × td)             

Integral Time, Ti =     6.67 × td                     

Derivative Time, Td = 0                                 

For PID Control

Controller Gain, Kc = 0.75 / (ri × td)

Integral Time, Ti =     5 × td 

Derivative Time, Td = 0.4 × td

Note that these tuning equations look different from the commonly published Ziegler-Nichols equations. The first reason is that Kc has been reduced and Ti increased by a factor of two, to make the loop more stable and less oscillatory. The second reason is that the Ziegler-Nichols equations for PID control target an interactive controller algorithm, while this approach is designed for a non-interactive algorithm such as is used in the Dataforth MAQ20 and others. (If you are using a different controller, make sure you find out which approach it uses.) The PID equations above have been adjusted to compensate for the difference.

Step 5. Enter the values

Figure 4: Here is the desired result—a level control loop, tuned using this modified Ziegler-Nichols method, responding to a setpoint change. Courtesy: DataforthKey your calculated values into the controller, making sure the algorithm is set to non-interactive, and put the controller back into automatic mode.

Step 6. Test and tune your work

Change the setpoint to test the new values and see how it responds. It might still need some additional fine-tuning to look like Figure 4. For integrating processes, Kc and Ti need to be adjusted simultaneously and in opposite directions. For example, to slow down the control loop, use Kc / 2 and Ti × 2.

With just a few modifications to the original Ziegler-Nichols tuning approach, these rules can be used to tune level control loops for both stability and fast response to setpoint changes and disturbances.

Lee Payne is CEO of Dataforth.

Additional reading: J.G. Ziegler and N.B. Nichols, Optimum settings for automatic controllers, Transactions of the ASME, 64, pp. 759-768, 1942.

ONLINE

To learn more about PID, look for additional application notes at Dataforth website: www.dataforth.com 

Read more about process control strategy with the related stories below.

Key concepts:

  • Modified Ziegler-Nichols tuning rules are effective for tuning level control loops.
  • They are designed for use with integrating processes and on non-interactive controller algorithms.
  • These rules provide loop stability as well as fast response to setpoint changes and disturbances.


K. , Non-US/Not Applicable, India, 11/06/14 11:57 AM:

There's always a mountain of a difference between theory and practice. Though there are numerous methods and thumb rules in tuning, the fine tuning we do in the end for precise control makes the final values deviate much more from the theoretical values obtained and it finally ends up to trial and error method.
MICHAEL , LA, United States, 11/19/14 04:26 PM:

Since a level is known as an integrating entity, use a controller which tries to control with gain only (as far as possible); use very long integral times. We use a controller which uses CL to try to keep the PV within adjustable deadbands, and only then adds itergral action when it is absolutely needed. The integral action is also different for each deadband. This type of control is especially useful when you have a process with levels in series (such as a high press separator feeding a medium press separator which then feeds a low press separator, etc). I have also found that using this type of approach yields a great amount of stability to our (VERY) highly heat integrated xylene unit complex.
Anonymous , 12/12/14 11:10 AM:

It is a long time since I worked in process control, but I used to avoid using deriveative time in level control. Any feedback?
Anonymous , 12/12/14 12:11 PM:

I tuned level loops for years. If the primary purpose of the vessel is just surge, then using a proportional only control works great. I typically set up a level control so that at 20% level the output to the valve would be 0% or 100% depending on if it was an inlet or outlet and at 80% it would be 100% or 0%. Any integral time at all could introduce oscillations in the process. With these settings the vessel would never fill up or go empty unless there was an upset the exceeded the capacity of the valve. It is important to note that in most processes, the level is the least important variable that needs to be controlled.
ROBERT , MI, United States, 12/14/14 01:04 PM:

Procedure needs help with motor drive with variable WR2
The Engineers' Choice Awards highlight some of the best new control, instrumentation and automation products as chosen by...
The System Integrator Giants program lists the top 100 system integrators among companies listed in CFE Media's Global System Integrator Database.
The Engineering Leaders Under 40 program identifies and gives recognition to young engineers who...
This eGuide illustrates solutions, applications and benefits of machine vision systems.
Learn how to increase device reliability in harsh environments and decrease unplanned system downtime.
This eGuide contains a series of articles and videos that considers theoretical and practical; immediate needs and a look into the future.
Sensor-to-cloud interoperability; PID and digital control efficiency; Alarm management system design; Automotive industry advances
Make Big Data and Industrial Internet of Things work for you, 2017 Engineers' Choice Finalists, Avoid control design pitfalls, Managing IIoT processes
Engineering Leaders Under 40; System integration improving packaging operation; Process sensing; PID velocity; Cybersecurity and functional safety
This article collection contains several articles on the Industrial Internet of Things (IIoT) and how it is transforming manufacturing.

Find and connect with the most suitable service provider for your unique application. Start searching the Global System Integrator Database Now!

SCADA at the junction, Managing risk through maintenance, Moving at the speed of data
Flexible offshore fire protection; Big Data's impact on operations; Bridging the skills gap; Identifying security risks
The digital oilfield: Utilizing Big Data can yield big savings; Virtualization a real solution; Tracking SIS performance
click me