Powerful Assessments In PlantTriage^®

PlantTriage^® includes over 80 powerful and specific assessments to improve plant performance. Each assessment is a powerful metric to help your business run better. This includes assessments for:

• Plant Operational Assessments
• 6-sigma & Statistical Analysis
• Hi-Level Unit Assessments
• Oscillation Analysis
• Controller Performance
• Control Valves
• Communications & Troubleshooting
• Control System Availability
• User-Defined Assessments

With over 80 assessments for every process variable, PlantTriage is the only way to get specific analysis of operational, equipment, and control performance. Keep in mind that every person will not be interested in every assessment on every variable. PlantTriage scales organizes the information according to both technical and economic factors, and sets priorities specific to your plant, so that you can remain focused on the most important issues.

PlantTriage Assessments

This section contains a summary of many of the assessments offered in PlantTriage. Each assessment offers some insight into the performance of your process, control system, or equipment.

Average Error

The average error is the average absolute value of the error between the process variable and the setpoint. The average error is a value between 0 and 100%. Lower numbers are better.

This calculation is the sum of the absolute errors (in percent) divided by the number of points. Average Error is calculated as long as PlantTriage has the setpoint.

The smaller the IAE the better. IAE is directly proportional to the costs of the loop.

The enforcement of operational targets is good when the average error is low. By achieving low IAE the process is better able to drive towards new targets and achieve them. Average Error, is an inferential quality variable. It may represent how close the plant can push up against constraints.

CO Noise band

This is the statistical analysis of the percent noise in the controller output (CO) during the assessment period. PlantTriage finds the noise band by finding the standard deviation (sigma) of CO during the quietest time.

If the CO noise band is large, then valve travel and valve reversals should be checked to see if there is a problem with the valve or PID tuning. Alternatively, measurement noise could be coming through the controller. If the CO noise band is large, this may be an opportunity to add a PV filter, or perhaps check the derivative value for proper setting.

Controller Health

PlantTriage's Controller Health Assessment provides an overall assessment of the performance of the controller. "Controller Health" is a composite assessment that includes many aspects of controller performance. When you need a single number to summarize the performance of each controller, the controller health assessment will point you in the right direction.

Data Availability

Troubleshoot communications problems down to the loop level. PlantTriage^® will track the percent availability of each process variable signal. This is very useful information for distinguishing between network, controller, or instrument signal failures.

ExperTune Index

The ExperTune Index is a measure of the performance increase possible if the controller is tuned with optimal, robust tuning values. This index indicates where wholesale tuning improvements on large numbers of loops can be made very quickly. No one else in the world has this unique controller performance index. Find out more.

The ExperTune Index is automatically calculated for control loops using Active Model Capture Technology

Filter—Estimated

Proper filter sizing is critical to good control performance. If the filter is too small, the controller and control valve will react to meaningless noise, resulting in increased process variability and shorter valve life.

If the filter is too big, the controller will respond sluggishly to upset conditions. This can cause quality and cost problems. PlantTriage can help you to choose a proper filter.

Harris Index

The Harris Index is defined as the ratio between the error (PV - SP) variance during this assessment over the variance achievable by a minimum variance controller. The larger the value, the poorer the performance of the loop. The Harris Index will vary as a number between 1 and infinity. A value of 1 is perfection or minimum variance control. Larger numbers might be considered worse.

Sluggish tuning has little negative effect on the Harris Index. Sluggish tuning usually causes the Harris Index to get smaller. Cyclic response or overly aggressive tunings will make the Harris Index a large number. A large number for the Harris Index might be considered between 50 and 100. However this may vary per loop and depends on setpoint changes and load upsets occurring. The Harris Index can get very large. Noise amount and pattern can have a great effect on the Harris Index.

Normalized Harris Index

The Normalized Harris Index (NHI) is:

Normalized Harris Index (NHI) = 1 - 1/(Harris Index)

The Normalized Harris Index will vary as a number between 0 and 1. The smaller the value, the better the performance of the loop. A value of 0 is perfection or minimum variance control. This is also called the CLPA or "Closed Loop Performance Assessment."

Large values of the Normalized Harris Index indicate a performance problem. While a low value is theoretically good, in practice low values do not mean the loop is performing well. High values give us information, low values do not.

Normalized Harris values above 0.7 or 0.8 usually indicate that the loop could be performing much better. Values above 0.5 warrant a careful look at the controller.

Loop Diagnosis—Simple

PlantTriage can automatically diagnose instrument, control, and valve problems. It does this using the real-time data and 80+ Assessments available on each loop.

PlantTriage contains an expert system, using engineering rules developed over decades of experience. In a way, it is like a tireless engineer, with over 100 years of experience.

The end result is a specific diagnostic from PlantTriage, telling you exactly where to focus on each loop.

Diagnostic Confidence

Like a human engineer, PlantTriage can draw conclusions by combining data from a variety of metrics. Now, PlantTriage can also assign a degree of confidence in each diagnosis.

Loop Diagnosis—Full

Many control loops have more than one problem. There may be a problem with the controller, and another with the instrument or valve. With PlantTriage's Loop Diagnosis-Full, you get a complete summary of the problems within a control loop.

Number of Mode Changes

This is the number of mode changes in 1 day. If the number of mode changes shifts away from the baseline, is there a reason the controller cannot hold the mode, or is there an operator training issue?

Noise band

This is the statistical analysis of the percent noise during the assessment period. PlantTriage finds the noise band by finding the standard deviation (sigma) of PV during the quietest time.

Normal process noise is minor variations in the process variable not caused by the control system, setpoint changes or load upsets. Normal process noise may come from electrical interference, magnetic fields, turbulence in flow controllers, waves in a tank in level controllers, etc. Generally the larger the noise band the poorer the performance of the controller. A very fast cycle in the controller will appear as noise.

On a bell shaped curve of all the quiet time data points, 3 standard deviations above and 3 standard deviations below the mean (for a total of 6) represent the noise band. 99.7% of all data points in the quiet time fall within the noise band.

PV noise band can be baselined to indicate problems with field transmitters. A shift in the PV noise band indicates a possible problem in the transmitter or measurement system. Possible problems indicated could include a partially plugged orifice plate, transmission lines, fouled sensor, or failure of the transmitter itself.

Opportunity Gap

The Opportunity Gap quantifies the potential for process performance improvement. The greater the Opportunity Gap, the more improvement can be made. When you reduce process variability using PlantTriage^® tools, the Opportunity Gap grows. Pay attention to Opportunity Gap on your most important loops, and adjust process setpoints toward process limits. The payback is immediate, significant, and often directly measurable in economic terms.

Assessments of Oscillation

It's not enough to know only that the loop is oscillating. You'll want to quantify the problem, and get to the root cause. The sections below show you how PlantTriage points you in the right direction to eliminate the root cause of the oscillation.

Oscillating—Tuning

The oscillation is diagnosed to be caused by aggressive tuning. 0% means the oscillation is not from aggressive tuning. Values towards 100% indicate higher confidence that the oscillation is caused from aggressive tuning.

With an "Oscillating—Tuning" diagnosis you should check the tuning of the loop using a process model. Process models are often obtained automatically using Active Model Capture.

All aspects of loop tuning are handled with PlantTriage, and the fully integrated PID Loop Optimizer.

Oscillating—Load

The oscillation is diagnosed to be from a load or load upsets. The higher the percentage, the more confidence in this diagnosis, up to 100% confidence. This diagnosis can occur from interaction with other loops.

To check the interaction with other loops use the PlantTriage Oscillation Periods and Strengths discussed below.

Oscillating - Valve

If the oscillation is diagnosed to be caused from a valve stiction or hysteresis then this assessment will have a value above 0%. 0% means the oscillation is not from your valve. Values towards 100% indicate higher confidence that the oscillation is caused from the loop's valve.

A value above 50% warrants a stiction and hysteresis check be done on the valve:

1. From PlantTriage, view the trend of the loop
2. Click the PID Tuner icon and select the Faceplate window
3. Follow instruction in PID Tuner for the stiction and hysteresis checks

PlantTriage Oscillation Periods and Strengths

Each loop has a dominant oscillation period. This is Oscillation Period 1 and is calculated from the power spectral density for each loop. If there is a cyclic oscillation in your plant, viewing a list of loops with the oscillation periods as a column can help you quickly group all the loops that exhibit this oscillation - even when it is "hidden" in other information. This will help you quickly pinpoint the cause of the oscillation - usually the most upstream loop. If the cyclic loops are interacting loops then first check to see if the oscillation diagnosis, discussed above is "Oscillating—Valve" for any of the loops. If it is then this is the first loop suspect to check for stiction or hysteresis causing a cycle.

The sample report above shows the details of each oscillation, including the periods, strengths, and the likely cause of each oscillation.

For every loop, during every assessment, PlantTriage calculates the relative power at all oscillation periods. The three periods with the most power are recorded. The highest three periods are the most important oscillations of the process variable signal. The three oscillation periods and their corresponding amplitudes are used as performance criteria.

The Oscillation period is in units of seconds. Oscillation strength is in percent of total energy in the spectrum (except the DC component).

Total energy = sum of all amplitudes minus the DC component

Oscillation Strength = Energy at this frequency / Total Energy * 100%

Oscillation Significance

A very powerful PlantTriage Assessment is "Oscillation Significance". This assessment compares oscillations against the other sources of variation in a control loop (noise, SP changes, load upsets), to determine which oscillations are the most significant issues in the plant.

Use the Oscillation Significance assessment to prioritize your efforts.

Oscillation Significance

Focus your efforts on the most important oscillations using the new Oscillation Significance assessment. This assessment uses assessments of variance and oscillation confidence to provide higher-level insight into oscillation problems.

More info about PlantTriage Oscillation Detection

Output At Limit

Output At Limit is the percentage of time during an assessment that the output was saturated at a limit.

Output At Limit indicates that the control valve is up against a hard constraint and automatic control has ceased. This may be indicative of a valve incorrectly sized or an unachievable process target sometimes referred to as a hard constraint. Additional process problems that could be indicated with "Output At Limit" include insufficient pump head, poor heat transfer, damaged or missing valve trim, etc.

High "Output at limit" values could indicate a bottleneck in the process. This could occur after increased plant throughput and quickly pinpoints the bottleneck of future increases.

Output Standard Deviation

The "Output Standard Dev" is the standard deviation of the controller output.

For many processes it can be an indication of an opportunity to achieve the same performance with less valve movement, reducing maintenance costs. It also is an indicator of valve movement or movement of inner controls. Lower values are generally better.

PV Availability

The "PV Availability" is the percentage of time that the PV was available for analysis during the assessment period. It is a measure of the quality of connection for data flowing into PlantTriage.

Real-Time Savings Advisor^™

Know exactly how much money can be saved by following PlantTriage advice. With Real-Time Savings Advice, you can be confident that operating changes are being made for the right reasons, and with the right priority.

Operations can now understand the exact amount of daily savings that can be achieved by using real-time performance assessments like the Opportunity Gap. The PlantTriage Real-Time Savings Advisor^™ helps you to prioritize your actions.

Evaluate how fully you are utilizing your control system. Service factor combines assessments of Time in Normal, Time at Limit, and others, to determine of your control system is being used to its fullest potential.

Robustness

Robustness measures how sensitive the process is to changes. Cycling, oscillations and swings, can result. If the robustness number is low the controller needs to be adjusted for higher robustness. Another unique controller performance index from ExperTune. More about Robustness.

Sensor Health

PlantTriage's Sensor Health Assessment provides an overall assessment of the performance of each instrument or sensor. "Sensor Health" looks at instrument failure modes such as spiking, noise, and flat-lining. By looking at this single assessment, you can get an understanding of the overall health of each instrument.

Service Factor

Evaluate how fully you are utilizing your control system. Service factor combines assessments of Time in Normal, Time at Limit, and others, to determine if your control system is being used to its fullest potential.

Settling Time

The settling time is the closed loop response time to an impulse. It is based off the impulse response of an auto-regressive (AR) closed loop model. For the purpose of settling time, PlantTriage defines the response as settled when it is within 6% of the initial value.

Shifts in the settling time can indicate a change in process dynamics warranting further investigation of why the shift occurred.

Number of SP Changes

This is the number of setpoint changes. It is normalized over 1 day. So if you had 5 setpoint changes in an assessment with an assessment time of 4 hours, the Number of Setpoint Changes would be 30.

If the value of this assessment increases over time, then this may indicate operations is having difficulty with this controller. Either the target is wrong or this controller or upstream controller has an unachievable target.

Setpoint Crossings

The Setpoint Crossings is the number of times the PV crosses the Setpoint in 1 day.

Setpoint Crossings, is applicable to quality controls. In many systems, whenever the process variable crosses the set point, off-spec product is produced. By developing an index of Set Point crossings and examining the causes, less off-spec product is produced and quality specifications can be crowded. Common examples include temperature, and moisture control.

Stiction

Another PlantTriage Assessment quantifies the effect of stiction on control loop performance. With "Sources of Variation", you can determine how much process variation can be attributed directly to valve stiction.

Often, plant personnel are asked to "tune the loop" to stop cycling. If the cycling comes from stiction, then you can't "tune" the problem. Focus your maintenance and process improvement efforts on fixing the right things, for the right reasons.

Time in Normal

The Time in Normal is the percent amount of time the controller spends in Normal mode. Typically normal would be Automatic, but it could be any mode specified.

In general, the more time in normal mode, the better. If the controller is not in normal mode, then the effectiveness of this controller and the process is likely compromised. A simple campaign to raise the "Time in Normal" of a population of controllers will result in significant performance improvement. A larger population of controls will be in a higher mode, consistent with the designers intention, thereby raising adherence to process objectives.

Valve Health

PlantTriage's Valve Health Assessment provides an overall assessment of the performance of each control valve. "Valve Health" incorporates the results of several aspects of control valve performance, to give you the best summary assessment. This is a powerful way to evaluate the overall performance of control valves.

Valve Travel

This is the total distance the controller output moves during the assessment period normalized to be on a 24 hour basis.

It is found by summing the absolute value of the controller output change between each sample. This number is then divided by controller output span and normalized to be on a 24 hour basis.

Valve travel is a direct indication of valve wear. For many processes it can be an indication of an opportunity to achieve the same or higher performance with less valve movement, reducing maintenance costs. ExperTune's experience with thousands of clients shows that a reduction in valve travel of a factor of 5, roughly doubles valve life. Benefits include extended equipment life, reduced cost and higher performance from the controls and process.

Valve Reversals

Valve Reversals is the number of times the controller output changes direction above a threshold value. This threshold is automatically determined by PlantTriage based on the process noise. Valve Reversals is normalized to be on a 24 hour basis. More reversals result in more valve wear. Benefits are similar to valve travel in that valve reversals can be reduced while achieving higher performance and reduced maintenance costs.

Variability

Variability is the relative value of Variance in the PV during the assessment time. It is expressed as a percentage of the mean and so allows comparison between the level of variability in different processes.

variability= 2*(Standard Deviation)*100 / Abs(Mean)

where

Abs = absolute value

Standard Deviation = sqrt(Variance)

Variance = sum(Mean-x(i))² / (npts-1)

Since variability is created by dividing by the mean it usually is a more useful comparison between loops than the standard deviation. For example, a 1 GPM standard deviation on a mean flow of 2 GPM is much different than a 1 GPM standard deviation on a mean flow of 20 GPM. So the variability can provide a more useful comparison between loops than the standard deviation.

Variance, Normalized

Variance measures the spread or dispersion of the process variable during the assessment time. "Variance, normalized" is dimensionless by being normalized with 100 divided by the span of the PV, squared:

                       sum(Mean-x(i))² * (100)²
Variance = ------------------------------------------------
                       (npts-1) * (span of pv)²

6-Sigma & Statistical Assessments

PlantTriage^® is a great tool to support 6-sigma efforts. You get real-time statistical information, as well as Pp, Ppk, and On-Spec/Off-Spec info.

Process Capability - Pp and Ppk

Real-time process capability assessments let you know how the process is performing. When there is a shift in process capability, you can find out about it immediately, using PlantTriage's Automated Alerts.

Statistical Tools

PlantTriage provides real-time statistical analysis of the process, including:

• Average of SP
• Standard deviation of PV
• Standard deviation of SP
• Standard deviation of Error

With an integrated historian, PlantTriage can help you to track & trend process performance over time.

Assessments Based on Industry Benchmarks

Compare your plant to others in your industry. Determine if your plant is performing better or worse than your competition. PlantTriage now follows established industry guidelines for reporting of loop performance, allowing direct comparisons between plants.

PlantTriage supports real-time loop performance metrics established by the ARC Benchmarking Consortium.

Unit Assessments

Tracking plant performance at a higher level helps to relate performance improvements directly to the bottom line. In PlantTriage, you have the ability to track and assess the following high-level performance measures for every unit operation in your plant:

Throughput (Production Rate)

Track overall production rate for each process unit. This information can then be trended with other key assessments within PlantTriage. Determine the root cause of production losses, and find ways to increase production rate.

Quality

Process quality is often tightly linked top process control performance. With the unit quality assessment, you can track your important on-line quality information, such as percent of time off-spec. With the tools in PlantTriage, you can identify the upstream root-cause parameters that have a direct effect on overall quality.

More info on using PlantTriage for Quality.

Material Cost

Track your overall material cost per ton. Plotting this alongside other PlantTriage assessments to determine the "sweet spot" for operating your process.

Energy Cost

Many PlantTriage tools help to drive a reduction in energy costs. By tracking unit-level energy costs inside PlantTriage, you will be able to show a direct link between process and control improvements and energy savings.

Reliability/Up Time

You can also track Reliability information for each unit. Track and trend this information along with assessments like stiction, valve travel, and sensor spiking, to get a better understanding of root cause issues.

Create Your Own Assessments

Create your own customized assessments for loops. You could use this for:

• Positioner Feedback
• Alarming Information
• Loop-by-loop Material Costs
• Operator tracking
• Diagnostic Information from Smart Instruments
• Any Customized, Proprietary Assessment You Want

Create free-form assessments for every loop. You have complete control over the calculations. The full power of the PlantTriage Assessment Engine is at your disposal. Your custom assessments will be historized, trended, and reported using all the standard built-in features of PlantTriage.

For example:

• They can be baselined and contribute to loop health
• They are historized and can be viewed in trends
• They can be included in browser interface loop reports

Complete Customization Possible

If you want to do some specialized or proprietary analysis, the PlantTriage Equation Builder can help you to calculate it. Your calculations can then be analyzed, tracked, and reported in all the standard PlantTriage tools. You can even use your custom assessments in TreeMaps, trends, and dashboards.

Get a Demonstration of PlantTriage

There is so much to see in PlantTriage. Get a private guided tour. Demo can be accomplished over the web, and takes about 1 hour.

Request a demonstration on-line or Contact ExperTune at +1 (262) 369 7711.