Introduction

Process engineers are searching for more efficient ways to improve process quality and increase plant throughput. Plant control valves are being singled out as one of the most critical components in accomplishing this. In fact, some experts have stated that the undesirable behavior of control valves is the single biggest contributor to poor control loop performance and the reduction of product uniformity. As a result, it has become increasingly critical that process engineers take full advantage of today’s computer systems and instrumentation to better understand modulating valve performance in their processes.

 Knowing valve position is fundamental to understanding modulating control valve performance. A common misconception is that valve position may be determined by observing the signal sent to the valve positioner or by monitoring fluid flow. Only by using a valve position transmitter will operating personnel be able to truly measure actual valve travel. However, once actual position is recorded this variable may be correlated with input signal and fluid flow levels to determine valve performance characteristics at any point in time. Armed with this information, maintenance engineers will have solid data to determine valve design suitability for process control, optimum valve replacement and maintenance schedules. The result will be improved valve selection, reduced unneeded maintenance and improved overall process performance.

 This article discusses some of the specific control valve performance parameters that may be accurately measured with the use of a valve position transmitter.

 Valve Backlash/Stiction

Backlash and stiction are caused by mechanical tolerances and friction in the valve positioner and control valve assembly. A number of factors may contribute to backlash/stiction such as valve design, shut-off requirements, actuator sizing, valve wear and age. When backlash/stiction becomes excessive, process performance may deteriorate to a point where it becomes necessary to maintain or replace the control valve. Measurement of stiction may also indicate acceptability of a particular valve design for a particular process. Without valve travel measurement it becomes extremely difficult to recognize and isolate all but the most severe backlash/stiction problems.

 One method to measure backlash/stiction is to put the control loop on manual and initiate small control input bumps(typically less than 1%) to the control valve. The valve’s position transmitter will feedback actual valve stem travel allowing the instrument engineer to determine the level of input change required before the valve will move and when it does so, how far it will travel. It should be noted that position transmitter accuracy should be well within 1% of travel in order to gauge the valve travel properly.

 Speed of Response and Overshoot

As in the case of backlash/stiction a number of factors contribute to the speed of response and overshoot including actuator sizing and positioner gain. Should these characteristics not be within desired levels process quality and overall system performance will suffer.

 A position transmitter is a critical component in measuring speed of response and overshoot. Feedback measurements are taken similarly to the backlash/stiction except in larger travel steps to more accurately observe travel speed and overshoot occurring with greater travel momentum. Varying step changes may be made to record differing speeds and overshoots under varying travel conditions.

 Valve Travel vs Flow

Over time valve wear may affect the flow rate with a given valve position when other parameters stay the same. By determining valve position with given flow values maintenance engineers will get clues on valve wear or be alerted to changes in the fluid stream which may significantly affect process performance. For example, if valve opening increases from 40% to 50% upstream from a heat exchanger while maintaining a constant flow value it may be an indication that scale is developing in the exchanger core. Or in the case of a highly erosive catalyst, a reduction in valve travel for a given flow value may indicate that the valve plug and trim is being worn away.

 Typically valve performance curves are taken to develop baseline characteristics in a new installation. After a period of time actual performance values may be compared to baseline values and determinations made as to whether deviations are significant to warrant closer scrutiny.

 Valve Shut-off

Where the control valve doubles as a shut-off valve the flow meter reading alone may not accurately determine complete valve closure. In this case, knowing the valve position may be necessary not only for valve performance diagnostics but also be needed to confirm full valve closure. In some processes the flow meter is used as the only element in determining shut-off even in cases where the flow meter cut-out point is extremely high. Flow meter cut out point is the flow value at which zero flow is transmitted. In some flow meters this cut out point may be 3% to 5% of the full flow value. If this is the case the flow meter may be indicating zero flow while up to 5% of the full flow value continues flowing through the pipeline. By using a position transmitter valve position status and the flow meter status may be used by operators to be fully confident that valve shut-off has occurred.

 Position Transmitter Selection

After it’s apparent that a position transmitter is a necessity for the modulating control valve, you must decide which type of device to specify. All transmitters are not created equal so it is important to scrutinize this part of the control valve system if you expect to receive the reliability and performance critical to the diagnostics you will be performing. Before reviewing the possibilities there are a few parameters that are important to note. The most critical typically include (1) Vibration at the valve, (2) ambient and temperature swings (3) accuracy and (4) type of environment (hazardous area). Once you have these parameters in hand you can more wisely select from the common options listed below:

1.  Potentiometer driven transmitter. 

These are most common and possess a variety of differing characteristics. They give near instantaneous feedback and offer application flexibility. Be sure to evaluate tolerance to vibration and the type of linkage used to drive the potentiometer from actuator stem to potentiometer element.

• Tolerance to vibration: Many potentiometers are intolerant of vibration exheeding 20 Hertz. Typically potentiometers with cycle life with less the 10 million rotations are susceptible to wear out in three to five months due to "dither" casused by vibration. When the potentiometer wiper rubs back and forth at high frequency a wear spot may be created through the resistance substrate where continuity is affected. To tolerate high frequency dither special hardened surface potentiometers may be selected which have cycle lives exceeding 50 million rotations.

• Temperature stability: (variation in output signal due to changing temperature) Typically the output signal varies less than 0.25% over a 60%C range. Variations are primarily due to electronic circuitry since resistance changes are inconsequential with potentiometer elements operating from 0% to 60%C. Be aware that some potentiometer elements have a substrate lubricant that may limit the potentiometer to temperatures above -10% C. When the lubricant becomes Too viscous due to low temperatures resistance values may begin to change.

• Resolution: (change in input position before a change in output occurs) Cermet or long life industrial grade potentiometers have infinite resolution.

• Linearity error: (ability to track a straight line output function) Potentiometers range in maximum linearity error from 2% to less than 0.1%. Usually longer life more vibration tolerant potentiometers will also have better linearity characteristics. In many cases potentiometer performance specifications are listed without consideration for the type of linkage and/or gearing and the rotational motion which will be used in your application. Care must also be taken to interpolate the potentiometer specs over the effective range being used in the application. For example a maximum linearity error of 1% of full scale may actually be over 90% of travel. Most potentiometers have a full scale of 340% and therefore this would be an error of 3.4% (340% x .01 = 3.4%). Using 90% of actual travel the error would be 3.8% (3.4%/90%). So be sure to adjust these performance specs to your application.

• Repeatability: (ability to follow the same output pattern consisitently) Most industrial grade cermet potentiometers have repeatability within 0.25% or better.

• Hysteresis: (variation in output when increasing vs decreasing scale)This value is typically the same as the repeatability in potentiometer elements and is usually not significant.  

When adapting to linear applications lever arm linkage systems are often times used with potentiometer based transmitters. The geometry of the linkage systems will affect linearity error dramatically. (See diagram) Other error factors such as repeatability and hysteresis error may also be effected depending on the linkage play. However, in rotary applications with zero backlash couplings throughout the automated valve system linearity, repeatability and hysteresis error may remain unaffected.

 • Signal response time: (time lapse between input movement and output signal change) Potentiometer elements are passive devices requiring negligible energy to operate thus have instantaneous response.

• Cost:  Potentiometer based transmitters are typically economical and increase in cost significantly depending on the quality of the potentiometer. High vibration tolerant, high accuracy potentiometer based transmitters with overall error of less than 0.5% in explosion proof enclosures are typically $600 to $800. With maximum in the 1.0% range and less susceptible to vibration the costs run from $400 to $600.

2.   LVDTs and RVDTs. 

Linear and Rotary Variable Differential Transformers possess good accuracy and are not affected by vibration. However they tend to be very expensive and are typically not flexible in adapting one model to varying applications.

• Tolerance to vibration: Since the RVDT and LVDT do not have contacting surfaces there is little if any wear due to vibration.

• Temperature stability: Typically the output signal varies less than 0.25% over a 60%C range.

• Resolution: Infinite resolution

• Linearity error: Maximum error typically ranges from 1.0% to 3.0% and is affected by the amount of the effective range used on the differential transformer

• Repeatability: Typically error due to repeatability is negligible running less than 0.25% for most commonly available devices.

• Hysteresis: This value may be significant with the maximum errors hovering in the 1.0% to 2.0% range.

• Signal response time: (time lapse between input movement and output signal change) Power requirements are significant to RVDT and LVDTs so if they have to be run on a maximum of 4 milliamps response times may be slowed to 20 milliseconds in order to conserve power.

• Cost . As part of a 4 to 20 mA loop or when signal conditioning is required, in an enclosure suitable for hazardous environment RVDT and LVDT transmitters range from $800 to $1600.

3.  Optical encoder driven transmitter

These devices sense position optically through a slotted disk. Their outputs are not affected by vibration or temperature within a specified range. However this disk is typically made of glass so excessive vibration or shock is a concern. They also tend to be expensive and power hungry with the result that they do not give instantaneous output values.

• Tolerance to vibration: Encoders also do not have contacting surfaces so there is little if any wear due to vibration.

• Temperature stability: Output does not vary with temperature because encoders are digital devices. However, the encoder uses an LED shining through an optic disk which may be affected by cold temperatures causing humidity to deflect the sensor light. As a result cold temperature ratings are limited to -10%C.

• Resolution: (change in input position before a change in output occurs) Combined accuracy of the device (repeatability, hysteresis, and linearity error is negligible)is completely dependent on the resolution. Encoders may be purchased with a variety of resolutions typical from 256 increments over 360% to over 4096 increments over 360%.

• Signal response time: Power requirements dictate the speed of response for encoders. High resolution absolute encoders (one unique position for each bit) may require over 50mA for continuous operation so the devices must have the output slowed dramatically to operate below the 4mA threshold in a 4 to 20 mA loop. Incremental encoders (require power at all times to maintain position reference) require less than for 4 mA for continuous operation and therefore may be operated with near instantaneous output.

 Both rotary and linear encoders are available. However, rotary are most common and have the same linkage limitations as mentioned with potentiometer based transmitters.

• Cost:  Encoder based transmitters vary in cost based on signal output required, resolution and whether the instrument is an absolute or digital encoder. Low resolution incremental encoders in a hazardous enclosure may cost approximately $400 to $600 with abolute high resolution models exceeding $2000.

 4. Hall effect transmitters

 Transmitters based on this system utilize a magnetic coupling and accuracy can vary considerably depending on device manufacturer.

 • Tolerance to vibration: Hall effect transmitters do not have contacting surfaces so there is little if any wear due to vibration.

• Temperature stability: Typically the output signal can vary  from 1% to 3% over a 60%C range.

• Resolution: Infinite

• Linearity error: Maximum error typically ranges from 1.0% to 5.0% and is affected by sensing element linearity as well as the target magnet arrangement

• Repeatability: Typically error due to repeatability is negligible running less than 0.25% for most commonly available devices.

Some positioners have begun to integrate position transmitters internally into their enclosures. This is usually acceptable for intrinsically safe applications in hazardous areas or for general purpose environments. The preference for many in hazardous areas is a separate enclosure which may be adapted directly to the valve system or may be "piggy backed" onto the positioner.

 StoneL’s HP-7 position transmitter incorporates a potentiometer with a maximum linearity error of 0.35% and minimum life expectancy of over 50 million cycles. The potentiometer element is constructed with a hardened conductive surface and wiper system which is designed to withstand high frequency "dither" that is experienced on many modulating control valves. The HP-7 may be adapted to linear travel stems with precision ball joint connections or retrofitted directly to positioners with a zero-backlash drive system. Adapting systems are readily available to Neles-Jamesbury, Fisher, Bailey, Moore and PMV.

 Conclusion

Now that processing plants are taking advantage of computer control systems to optimize operations, it becomes intuitively obvious that the next step is to properly instrument the modulating control valves to better understand their performance. StoneL position transmitters are now available at prices running in the $400 to $600 range which can be specified with new control valves and easily retrofitted onto existing control valves. When preventative maintenance of the valve system runs anywhere from $1,000 to $10,000 per point and downtime due to control valve failure can exceed 10 times those amounts it becomes essential to specify a position transmitter as part of the modulating control valve system.

 References:

1. Entech Control Engineering "Control Valve Dynamic Specification" (Version 2.1, 7/94)

2. Instrument Society of America, Hughs, T.A. "Measurement & Control Basics" (1988)

3. StoneL Corp. "Valve Communication and Control Sensor Issue" (No. 1)