PID Loop Optimizer

The only complete tool kit for control loop optimization.

Fast, Accurate, and Easy!

• Eliminate guesswork in controller tuning and loop optimization.
• Resolve problems with sensors, filtering, tuning, and control valves.
• Understand process dynamics.

View case studies for PID Loop Optimizer and PlantTriage^® Performance Supervision System.

Consider PlantTriage for Continual Improvement

All the benefits of PID Loop Optimizer are included in PlantTriage. If you are looking at Loop Optimizer, many of our customers find great additional benefit by using PlantTriage Performance Supervision System. PlantTriage gives these additional benefits:

• Continual performance improvement for all the loops, all the time.
• Prioritize work efforts—Which of my 1000 loops should I look at?
• Uncover problems before they cause a shut-down or failure.

What's Included

Supports All DCS Control Algorithms

ABB, Bailey, Foxboro, Fisher, Moore, Honeywell, Yokogawa, etc…    Includes setup wizards that make connecting a snap. View the complete list of supported controllers.

In the rare case that your controller is not listed, give us a call at +1 (262) 369 7711. We can probably add it!

Loop Summary Table Automatically Summarizes Your Tests

The Loop Summary table provides a concise way of quickly recording the results of several loop tests. It tabulates tuning parameters, model parameters, RRT, quality of fit, and relative stability index from several loop tests. From these tests it automatically:

• Selects the most conservative tuning values.
• Provides the average tuning values.
• Indicates relative stability.

The table lists the results from your previous tests on the loop:

• PID values
• Model parameters
• Relative Response Time (RRT)
• Stability Index. Easily compare the relative stability of the PID tuning
• Quality of frequency data fit

Valve Wear Analysis

ExperTune shows how to reduce control valve wear extending valve life and useful valve service time.

Only ExperTune calculates optimal PV filtering and optimal PID values to reduce the valve travel and reversals for less valve maintenance cost. Lets you replace and rebuild valves less often.

Improve and maximize valve life by comparing tuning and PV filtering. Try what-if.

ExperTune compares predictions of valve travel and valve reversals numerically and graphically in the presence of noise for both current and optimized values. Both the Reversal and Travel Indices are directly related to valve wear. The software analyzes optimized values to reduce the indices resulting in longer valve life. Often times the life of the valve can be increased without sacrificing loop performance or robustness.

• Valve Travel Index—a quantitative comparison of how far the valve will move—lets you compare tunings and filters to see the effect on valve wear. Total valve travel for each case also given.
• Valve reversal index—quantitative comparison of the number of times the valve will change direction—compare tunings and filters to see the effect on valve wear. Total valve reversals for each case also given.
• Use the PV filter to minimize valve wear. Choose between first order, second order, Butterworth or averaging filters. Try changing the filter and PID values yourself and see immediately the effect on valve wear and loop performance.
• Simulation with noise. Try what-if to see how changing tunings or filters effects controller output wear.
• For a quantitative check of performance and robustness, the software computes a Performance Increase and Robustness Index. Use the Performance Increase to see if loop responsiveness is compromised by reducing valve wear. Use the Robustness Index to see if the loop sensitivity (to the process changing) is compromised by reducing valve wear.
• For ISO9000 or if problems occur later, ExperTune creates a detailed report of current to new settings for comparison. The report sums up the Tuning parameters, filter, Valve wear analysis, Performance and Robustness indices, and process model. The report also includes comparison graphs including: Robustness plots, simulated loop response, noise response plots. ExperTune creates reports in Microsoft Word. The report's graphs and values are driven by user set bookmarks, and are completely user customizable.
• Any tuning parameter, PV filter size or type are easily tested and compared off-line in simulated time response plots of set point changes, load upsets, or response to noise. These in one window and the robustness plot in another. Change a parameter and the graphs instantly update with both the current and new response plotted in red and blue for easy comparison. Try what-if analysis to get the exact response you want before you download tuning parameters or apply a PV filter.

PV Filter Analysis

Only ExperTune calculates the largest possible filter for your selected tuning category. The filter is as large as possible without degrading performance significantly.

• Choose from first order, second order, Butterworth, or averaging filters. Try changing the filter and PID values yourself and see immediately the effect on valve wear and loop performance.
• Use ExperTune's filter time or try what-if with your own current or new filter times. Simulation and robustness plots update instantly to reflect your filter type and size.
• Use the new PV filter to analyze the effect on valve wear analysis using the new noise simulation.
• See the entire family of PV filters for each tuning category in the PID grid.

Specify your filter address and ExperTune reads the current filter value from your controller. It is displayed next to the current PID values. After ExperTune has found the best filter for your process, click Download and the filter value is written to your controller. Supported in DDE, and OPC communication options.

See the tutorial on PV filtering.

Unfilter Your PV

If your process variable is filtered, the filter may be masking or hiding problems in the loop. Additionally, a filter could make the process much harder to control.

You can direct ExperTune's Advanced Loop Optimizer to automatically unfilter your data. With "Unfilter PV" checked, optimal PID Tuning and filter calculations are based on the true data. Even though you collected filtered data, you can also see the plot of the raw data by checking unfilter.

Two unfiltering options are available allowing you to remove the effects of PV filtering no matter where the filter is in your control system. The options are:

1. Filter time is automatically read and compensated for by PID tuner. In this option the filter time is read from an address in your controller. The data is automatically unfiltered (if you select "Unfilter PV")
2. Or enter the filter time used by your transmitter or control system.

Either option lets you see the result of unfiltering in the time response plot—see the raw, unfiltered data.

Valve Stiction Wizard

Flow Control 2001 Innovation Award winner for Valve Stiction Wizard

A sticking valve caused by friction within the valve can seriously degrade control loop performance. This phenomenon is called valve stiction. The more your valve suffers from stiction, the more your loop will oscillate around the setpoint.

An ExperTune wizard guides you through determining the amount of stiction in your control loop. Anything more than 0.5% stiction and correcting the situation may dramatically reduce cycling and improve loop performance. The wizard will come up with a stiction value even in loops where noise is excessive.

Performance Summary Window

Tuning is a compromise between performance, robustness and valve wear. You want to increase loop performance, but increasing performance can also reduce robustness and valve wear. ExperTune shows you all the performance indicators in one concise window. So you can view overall trade-offs at a glance.

The Performance Summary Window provides a concise listing of the:

• Performance Increase
• Robustness Increase
• Valve Travel Index
• Valve Reversal Index
• Relative Response Time (RRT) with Current and New Tunings

Performance Evaluation

In one glance you can now easily compare and evaluate the trade-offs in your control loop. (Advanced Loop Optimizer only) Quickly compare:

• Performance
• Robustness See the tutorial on Robustness Plots.
• Valve wear (travel and reversals)
• Relative response time (RRT)

Time Line Analysis

Use the Time Line Analysis to help understand the relationships between the time of each component in the control loop. For example, the sample interval should always be shorter than the dead time in the loop. If the sample interval is close to the dead time then shortening the sample interval will improve the performance of the loop.

After ExperTune software makes all the necessary conversions, the following time elements of your control loop are plotted on the time line:

ExperTune evaluates the elements on the time line and reports to you. Green bars indicate OK. Yellow is caution and Red indicates you can make improvements that will significantly improve performance.

The following items are evaluated and checked.

• Sample Interval compared to Dead Time
• PV Filter compared to Controller Derivative Time
• Controller Derivative compared to Controller Integral Time
• PV Filter compared to Dead Time

As with all of ExperTune's analysis, with one click ExperTune creates a full report in Microsoft Word including the Time Line plot, all the items on it, and the comparison ratios. The report documents your loop for future records.

Dial in the robustness you want

It just couldn't be easier to get the best performance trade-off. Start with ExperTune's unique Robustness Plot that shows how sensitive your loop is to the process changing. Drag the plot with the mouse to set the sensitivity you want. The required PID values and all the ExperTune analysis windows, including the Performance Summary, update—instantly! See the tutorial on Robustness Plots.

Linearize Your Process For Optimal Performance At All Production Rates

Does your process oscillate near one end of the range and respond sluggishly at the other? ExperTune's characterizer, helps you linearize your process, so you get uniform performance across the entire range—run your production at its optimum rather than de-tuning for oscillations.

Use ExperTune's characterizer to linearize:

• flow loops
• jacket temperature in chemical reactors
• slave loops in cascades
• any loop where the setpoint changes
• Characterization tech sheet
• White paper by Greg Shinskey
• Characterizers for Control Loops

Universal Linearizer for pH Loops

Linearization for any pH loop resulting in pH control with little or no cycling. Better pH control improves product quality, reduces reagent costs and helps to keep pH within regulation limits.

pH loops are inherently non-linear and cycle around the setpoint while taking too long to recover from large load swings. ExperTune calculates the characterizer that should be applied to both the measurement and set point of the pH loop. The linearized pH measurement can then be accurately controlled by a traditional PID controller.

The ExperTune pH linearizer Wizard takes you though the following steps:

• From titration data, ExperTune plots the pH curve and converts it to a characterizer.
• Modify the curve (characterizer design) by adding or dragging and dropping points.
• The pH linearizer works with any controller that allows characterization or user programming.
• For easy application of the characterizer within your controller, X-Y coordinates, BASIC, FORTRAN, or C code, are all automatically generated.
• See the white paper by Greg Shinskey on linearizing pH Loops

Force Process Modeling Selections

ExperTune's PID Tuner identifies a process model using advanced frequency response methods. ExperTune software develops a high-fidelity model from a variety of data. Here are examples of data sets ExperTune software can use to identify the model:

• Open loop step
• Closed loop step (controller in automatic)
• Open loop pulse
• Closed loop pulse (controller in automatic)
• Pseudo random data
• Open loop doublet pulse

See a presentation on "Plant Data to Use For Modeling and PID Tuning."

ExperTune automatically chooses the best model structure for your process. Either:

• Dead time with gain
• First order with dead time and gain
• Second order with dead time and gain
• Second order over-damped with dead time (Imaginary roots) and gain
• Integrator with dead time
• Double Integrator with dead time
• Integrator with first order and dead time
• Inverse response process or shrink-swell: Integrator with first order, lead time, and dead time

With ExperTune Advanced Loop Optimizer version, you can force ExperTune to use a model structure that will result in one of these models.

White paper: Modeling in ExperTune's PID Tuner/Analyzer and PlantTriage

Statistical Analysis and Histogram

View statistics before and after. See how the standard deviation, variability, and valve wear have improved.

Analyze the grouping of your PV, CO, and Errors. The histogram shows the statistical spread of your data. From noisy data determine if there is an offset one side or the other. For example, an inverse bell curve indicates a probable valve stiction problem. Well tuned loops have a tall thin bell curve. Poorly tuned loops have a short fat bell curve.

You can select the IAE (integrated absolute error) and valve wear indicators to select per hour, day, week, month, or year.

The IAE is a measure of performance. It is the area on the time graph between the setpoint and the process variable. Smaller IAEs are better since it means you were running closer to setpoint for that time. IAE is useful since in some plants it can be related closer to dollars and economics of the plant than any other performance indicator.

In many loops, you want to operate as close as possible to specification without straying off. For example, you need to add an expensive ingredient called MTBE to a gasoline so that the final concentration of MTBE is 2%. Adding more than 2% MTBE is giving it away.

But you need to add enough MTBE to reach 2%. So, typically you would have a set point of slightly higher than 2% MTBE. The better the tuning and lower the IAE, the closer to 2% you are able to set your loop and the less MTBE you will use.

The histogram to the right was taken from normal operating data and indicates a stiction cycle in the loop. The two humps, circled in red, show that the PV was usually on either side of the setpoint. If your loop exhibits this type of histogram, you should perform a stiction check on the loop to verify the loop contains stiction.

Multi-Variable Loop Analysis

Add as many extra variables as you want on the faceplate. Lets you monitor other variables on the same faceplate trend to see how other process variables affect the loop. These can be configured for modeling and tuning of interactions or other loops. Defining extra loops helps you work with these systems:

• Monitor, tune and analyze multiple loops from one tag
• Add as many extra trends as you want. Lets you monitor other variables on the same faceplate trend to see how other process variables effect the loop.
• Create extra loops from the extra trends for extra modeling, analysis, simulation and tuning
• Multi-variable: Model and decouple multi-variable systems 2x2, 3x3, 2x3, etc.
• Feedforward: Model load processes for feedforward compensation
• Cascade loops and RRT analysis (see next section)
• Interacting loops and RRT analysis (see next section)

Tune Cascade Loops

Use Relative Response Time to tune interacting systems

Only in ExperTune, this is a new cutting edge tool for decoupling interacting loops. The Relative Response Time (RRT) makes it easy to set interacting and cascade loops to work together. The RRT in each interacting or cascade should be a factor of 3 away from the others. Change RRT by adjusting the safety factor or lambda time. The Relative Response Time is the period at the peak amplitude ratio in the closed loop frequency response (to load upsets).

• Cascade: Use ExperTune PID Analyzer to model and tune cascade systems simultaneously with one test. The master (outer) loop should have a Relative Response Time that is 3 times slower (larger) than the slave (inner) loop.
• Interacting Loops: Use cross-correlation to see how one loop effects another. To prevent interaction, adjust the safety factor or lambda value (set point speed) in the interacting loops so that the Relative Response Time is different by a factor of 3.

Feedforward Compensator Design and Modeling

ExperTune has a unique offering allowing you to build a robust feedforward system. First use ExperTune's Advanced PID Loop Optimizer to model both the feedback and load or feedforward system. Then use Loop Simulator to design the compensator and simulate time responses with and without the compensator in place.

See white paper on how to build a feedforward compensator.

Power Spectral Density

Power Spectral Density helps find hidden oscillation problems and identify cyclic interactions between control loops.

• Helps find the root cause of a plant-wide or unit-wide oscillation
• Shows if an oscillation is hidden in the noise.
• Gain From Using One of Process Control's Emerging Tools: Power Spectrum (280K MS Word Document)
• Full story about how ExperTune Power Spectral Density solved a $1 million/day cycling problem. (PDF file, Intech August 1998)

• Lists the 4 largest peak frequencies and their percent energy in the spectrum. The energy tells you how much this peak is contributing to the variability of your product. ExperTune now uses clusters of peaks reducing inaccuracies from leakage and sampling.
• Display frequency and percent energy on peaks in the zoom area only or the entire spectrum
• Optionally use the same chopping and windowing as PlantTriage. This option further decreases inherent errors in the spectrum.

Inverse Response (Shrink-Swell) Process Analysis, Modeling, and Tuning

Drum level control and bottoms control in distillation columns are examples of the rare loops that exhibit an "inverse response". The PV initially goes in the "wrong direction". These are typically very challenging loops to model and tune. Now simply click to indicate you want an inverse response loop, collect some data, and ExperTune gives you a complete analysis including tuning, modeling of the process, simulation, robustness, valve wear analysis, etc.

ExperTune automatically identifies the process model including process lead, lag, integrator, dead time, and gain.

Auto and Cross-Correlation

Correlation analysis helps you see how one loop affects another.

Set Point Filter

By using a setpoint filter, you can get good response to both set point changes and process upsets. Tune the loop for great load rejection and set the filter to eliminate overshoot on set point changes. Select the set point filter option and ExperTune will apply a filter to the setpoint of the simulation of the loop with the New PID settings.

The lag time is automatically set by ExperTune to be approximately the integral time (in seconds). Adjust the lead/lag ratio and instantly see the simulated response. To implement the filter in your system, use your systems lead/lag block or the equations given to program the filter.

Lambda Tuning Categories

Uses traditional Lambda tuning rules. Choose between lag rule or integrating rule.

ExperTune finds lambda tuning from a setpoint change
Other software packages using lambda tuning require you to tune from manual step changes in the controller output (open loop). ExperTune lets you tune and analyze from this data or from a manual pulse, or setpoint (closed loop) data. ExperTune's only requirement on the data is to start settled, do a quick change and end settled (with no load upsets).

Advanced Reporting

Sample Loop Report in MS Word

Create a detailed analysis report in the same standard format (Microsoft Word) for every loop in your plant. If problems occur later, you have a detailed report for comparison.

After tuning a PID loop, one mouse click creates a detailed loop report with all the analysis graphs and tuning values. You can also select additional analysis options added. For example, the power spectral plots pinpoint the cause of cyclic disturbances. The hysteresis graph helps tell you if your valve is wearing out or sized incorrectly. The robustness plot tells you how sensitive your loop is to process changes with whatever tuning you have.

All these plots and companion analysis information appear within the Word report document. You can completely customize and configure reports. In the Word template, choose the location and size of added analysis pictures or text. It automatically generates a summary and conclusion based on information entered.

The template provided with ExperTune allows you to add all the items below. However the template is completely customizable. You can add to existing loop report or start new one. When you first add an item to a report ExperTune starts Word if it is not already running. Reporting includes detailed analysis of Hysteresis and Linearization.

Reports include:

• Tuning section — one mouse click. The tuning section is included with standard version also.
  • Current and new tuning values
  • Loop Notes
  • Time data for graph
  • PID Tuning Grid
  • Modeler Window
  • Robustness Plot
  • Frequency Response Plot
  • Simulation to Set Point
  • Simulation to Load Upset
• Characterization section includes data for characterization, characterizer window, range of process gain and recommendation for linearization.
• Hysteresis section includes data collected for hysteresis check, lists valve gain, hysteresis, noise, and recommended ranges.
• Select action to take for hysteresis, linearization, asymmetry, and tuning from a drop down list. Colors changes to indicate severity.
• Before and after, showing closed loop, open loop, and statistical analysis.
• Automatically generates a summary and conclusion based on information you've entered
• Completely customize and configure report templates any way you like.
• Resize, move, or add graphic placeholders.
• Have a report template for each category of loop.
• Add any text or links in your templates.
• Modify completed report any way you like.
• You can have several reports open at the same time.

Tune and Analyze Data from Your Plant Historian

Use the historical plant data from your plant historian for optimization, analysis, and PID Tuning. PID Advanced Loop Optimizer, includes HDA, or Historical Data Access allowing you to view, analyze, and tune data from any HDA source. (HDA is a standard from the OPC foundation that allows tools like ExperTune's to easily access the historian's data.)

Allows you to:

• Use historized data for PID Tuning and modeling. If operations has made set point changes or changes in manual previously, you can use this for test data. You will not have to bump the process.
• Comparison to previous times. For example, you may want to compare to 6 months ago. Run statistical analysis, and modeling to see how the plant has changed.

Use historized data from:

• OSI PI Historian
• Honeywell PHD
• AspenTech InfoPlus.21
• Invensys AIM*Historian
• Any historian supporting OPC HDA

More features in Advanced PID Loop Optimizer

• ExperTune includes an extensive list of setup wizards to get you connected to your controller.
• Manual hysteresis option to manually choose step times. Drag lines onto areas just before the CO change. Allows hysteresis checking with poor or noisy data.
• TimePlot: Manually set min and max PV and CO. Allows for easy before and after comparisons. These min and max settings are saved to the archive—so next time you view timeplot, your min and max are the same as the ones you've set. Auto-scaling is still the default
• Separate entry for controller sample interval. If your controller is overloaded and unable to run the PID fast enough to get optimum response, simply enter your controller's actual PID Controller Sample Interval. ExperTune will take the slow PID Controller Sample Interval into account and recommend the best settings for your slow controller.
• Specify the maximum data compression (up to 32,768 points)

PID Loop Analyzer for Non-DCS Systems

The loop optimization tools shown above on this page are available with the Advanced PID Loop Optimizer only. The Advanced PID Loop Optimizer is required for DCS systems.

How To Connect To Your Controller

Compare Advanced PID Loop Optimizer to the Loop Analyzer for Non-DCS Systems

ExperTune's Advanced PID Loop Optimizer includes all of the tools of ExperTune's Loop Analyzer for non-DCS Systems.