1 April 2006

When all else fails

A kinetic energy criterion for control valves via trim retrofit is the final solution

By Herbert Miller, Laurence Stratton, and Mark Hollerbach

Crunch the numbers on 470 control valves retrofitted in the field, with new valve trim, and the user sees that new trim controls the fluid exit energy level.

FAST FORWARD

  • Here is a review of the many causes for a user to take the difficult step of retrofitting a control valve in the field.
  • A significant observation is very high kinetic energy valve designs fail repeatedly.
  • Reduce the fluid energy level exiting the valve trim to acceptable levels works on all valves.
 

With that, the user is back in control of the process, and that's important.

The retrofits took place because the valves were not performing their intended control function. The database shows that after retrofit and using trim that limits the kinetic energy below 70 psi (480 kPa), a valve application results that meets or exceeds the user's expectation.

The population of retrofitted valves covers a wide range of sizes, pressure classes, design types, applications, and original suppliers.

The inertia associated with a change of a control valve in the field is quite significant, and a user will need a strong motivation in order to take this step.

The problems and causes that have pushed the user to implement a retrofit of a valve are many. In almost every case, there have been lots of trials in the field with some unique work-a-rounds and fixes for the problems.

A retrofit is not a trivial undertaking. The user has a significant problem he or she has not been able to solve.

The valves in this study all failed and all resisted any fixes. Retrofitting them with a trim to control the fluid exit kinetic energy turned them into successes.

Kinetic design criterion

The kinetic energy density combines the influence of density with velocity of the jet exiting a valve trim. The term density makes sense because the energy level is per unit volume.

The term density is in the text here whenever kinetic energy is the reference. We define kinetic energy as this:

fails1

The velocity in this expression is the average trim outlet-jet-fluid velocity and Greek letter rho (ρ) is the density of the fluid at the exit.

The fluid kinetic energy, at the trim exit, criterion for control valve applications in continuous duty applications is 70 psi (480 kPa) or less. Levels of energy are in the same units as pressure, and the equation for kinetic energy density is the expression for dynamic pressure.

The application of the kinetic energy criterion is in addition to the traditional design process. That is to say, all decisions regarding materials, capacity sizing, body, trim and actuator selection, adjustments for erosion, cavitation, and/or noise take place and then a check of the trim design is made to be sure the kinetic energy level meets the design criterion.

The kinetic energy criterion (70psi) is an additional consideration and not a substitute for using all prior control-valve design knowledge and practices. The experience as shown by the results presented below will provide the best assurance available that the resulting control valve application will perform to expectations.

Desired tolerances and interfaces

A retrofit is a replacement of a valve's main flow control trim component using the original valve body as installed without alteration. The trim component that is the control component usually goes by the name cage or in the case of a top guided trim set; it is the seat ring and the plug (closure member) parts that form the throttling orifice. The sole purpose of changing the flow control trim component is to minimize the fluid exit kinetic energy. The goal is to reduce the energy level to 70 psi (480 kPa) and lower if space permits.

Other valve internal parts such as the seat ring, plug assembly, guide bushings, and soft goods (seals) change out for the purpose of maintaining desired tolerances and interfacing with the existing valve body. It is an unusual case when body machining takes place. There are occasions when the body is in such bad condition it must be restored by repair and machining. Most of the time sanding and cleaning is enough to assure a proper interface and sealing surface. The actuator may be changed if the needed trim will require a longer stroke and/or to maintain proper seating loads for leakage control.

One of the most important steps in the process is the communication with the customer. It is important to know what problem is causing the need to retrofit the current valve so focus is on the root cause and the original design problem does not recur. This is a design by application, and there can be many questions. There may be as many as 22 conditions of process information needed to make sure an application is properly accommodated (some may be better know by the valve manufacturer than the users). The key design issues will originate from the root cause of the dissatisfaction and failure of the current installation. After this, steps transpire to determine dimensions of the existing valve body either by implication from replacement parts or actual measurements. Some field machining of supplied parts may be necessary to assure a good fit of the energy control trim.

Once the proper process conditions have been determined and agreed to with the user, sizing happens per the ISA/IEC Standard. The required capacity remains constant when selecting the proper number of discreet pressure reduction stages to achieve a 70-psi (480 kPa) fluid kinetic energy density criterion for each condition. If gallery space in the existing valve body permits, additional stages go in to enhance flow control thereby minimizing the trim exit kinetic energy. The trim inlet kinetic energy stays low as well by this focus on the outlet criterion.

In a few cases, the retrofit is a replacement of a top guided or a thin wall cage that does not allow enough pressure drop stages to work. There is a trade off between the outside diameter of the trim and the need to maintain enough space between the trim and the body wall. If this space is not sufficient then a judgment as to whether the energy level is close enough to 70 psi (480 kPa) for the application or a significant enough reduction in the exit energy level has occurred to be assured of a good control valve application takes place. In a number of cases, the retrofit approach is abandoned and a complete valve replacement is used. This judgment comes into play for about 10 to 15% of the retrofit cases and uses the experience gained from designing by application and the excellent performance history of almost 500 retrofitted valves.

In all cases, the energy control trim used is the multi-stage, multi-path tortuous path trim. This trim allows a large number of pressure drop stages to install in a relatively thin radial wall. Each right angle turn acts as a speed bump, reducing the fluid velocity, so that by selecting the correct number of stages, the desired kinetic energy level at the exit of the flow path happens.

Because the flow per channel is less, enough channels work such that the valve capacity transpires. This leads to longer strokes in many cases than would be used for a less tortuous path trim design. Because of the longer stroke, the diaphragm actuators frequently swap out as part of the retrofitting activity with a piston style actuator. The piston actuator also allows higher actuation pressure that provides better resolution, stiffness, and shutoff seating loads.

fails2

Root causes for a retrofit

During the communication phase of the retrofit activity, the user provides what he or she feels is the cause of the problem with the original installation. Usually there are multiple causes, and so multiple reasons go to the record book. There is no attempt to edit or narrow the input, as an understanding of the problem is the most important issue.

As an example, a user may say the problems are poor control and cavitation. In the discussion, it is apparent the poor control occurs during the start up of the process. This would suggest erosion due to cavitation is the real root cause. The erosion of the internal parts does not allow sufficient resolution during the initial travel of the plug to give good start up control, and operators notice the lack of control. In this case, the users' input of poor control and cavitation go to record, even though the true problem is erosion of the trim due to cavitation or excessive fluid kinetic energy.

Note: The leading complaint for the liquid valves is controllability followed by erosion, leakage, vibration, and cavitation. All of these relate directly to the kinetic energy levels in the valve. A big surprise is the number of complaints of stem separation or breakage, six instances, or 3% of the complaints, but it affected 41 valves, 7%. Five of the six complaints regarding a stem problem also listed either vibration or cavitation as a cause for the retrofit. Noise for liquid valves is not a frequent complaint.

Of the 155 liquid application complaints, 145 of them are clearly associated with trim fluid design issues such as fluid energy, sizing, and capacity. The 10 complaints not associated with trim design included high maintenance, lack of support from the original vendor, and linkage breakage. Apart from the lack of support, these relate to excessive fluid energy levels. None of the reasons for a lack of vendor support is acceptable. Reasons for a lack of support could include poor communication channels between the user and the vendor's technical experts, a lack of understanding about the problem or the application, and insufficient resources, to name a few.

Note: The leading complaint for the gas valves is noise followed by vibration and leakage. Controllability is a distant fifth, which likely reflects the reduced potential for erosive damage with a gas. Again, stem separation or breakage came up three times as a complaint with an impact on 3% of the valves. Stem problems always went with vibration as an accompanying cause.

Of the 100 gas application complaints, 85 of them are clearly associated with trim fluid design issues such as fluid energy, sizing, and capacity. The remaining 15 complaints could also relate to fluid energy issues, but it is less obvious. Vendor support, again, reflects a frustration by the user with the supplier's inability to solve the problems with the valve application.

Meeting the energy criterion 70 psi

There are a few cases where the original trim met the energy criterion. Still the valves were retrofit. Except for nine designs, these valves had other flow conditions specified for the application in which the energy exceeded 70 psi (480 kPa). In reviewing the nine designs, the applications represent traditionally difficult applications in which there are problems with cavitation (pump recirculation), vibration (nuclear plant service water), control (auxiliary steam supply and deaerator level), and noise (compressor recycle and a gas valve with sonic exit conditions.) The most probable cause for a failure in these cases is the normal practice of designing to flow condition information only and not considering or understanding the application.

Other excuses could be the original supplier was not conservative enough in their design practice and decided to proceed with the risk and/or the user was proceeding with what they thought was the lowest cost option.

There are a number of cases in which the energy level exceeded the criterion even after retrofit. These represent the cases in which the criterion was impossible because of a gallery space limitation in the valve body. However, a review of the applications and the significant reduction in trim exit energy for each retrofitted case resulted in a judgment to proceed. In all of these cases, the reduction in the energy level was more than eight times and frequently much greater. Overall, the average ratio is 21 for the original trim energy level to that of the energy control trim.

The fact that there are a few cases in which the kinetic energy exceeded the criterion of 70 psi (480 kPa) would suggest it is not a hard rule. There is some validity to this as demonstrated by the data; however, during the original valve procurement the incremental cost in achieving the criterion is small in comparison to the risk. The data from "Kinetic energy - all cases" shows 90% of the flow cases (original trim energy greater than the criterion) resulted in field failures when ignoring the criterion.

The maximum energy level supplied with an original trim design was 3280 psi (22.6 MPa), which is almost 50 times the criterion. The average was 480 psi (3.3 MPa) or about seven times the criterion. This is a normal expectation from a supplier when there is no consideration of the trim exit energy. With these high fluid dynamic pressures, it is not surprising that a lot of damage can take place in a control valve piping system.

For the energy control trim, the average energy level was 44 psi (300 kPa) with a maximum of 340 psi (2.3 MPa) for one case. Only 3% of the data exceeded 140 psi (960 kPa), which is twice the criterion. Therefore, there is not a lot of data that would support an increase in the criterion or a push to accept a higher risk by exceeding the criterion.

One question that may arise is whether this database of retrofitted valves is representative of all control valve applications. We divided the valves into three sections representative of the service from "mild" to "robust" to "severe" service conditions.

The "mild" service upper boundary is representative of an ASME Class 300 Pound valve with an inlet pressure on the order of 700 psi (4825 kPa). The "robust" service upper boundary is representative of an ASME Class 600 Pound valve with an inlet pressure on the order of 1400 psi (9.65 MPa). The "severe" service inlet pressure has no limit, however the maximum inlet pressure encountered here was 5840 psi (40.3 MPa).

With these three-market application designations, the number of data points in each segment was 105, 106, and 218 for the "mild," "robust," and "severe" service categories, respectively.

These numbers are a statistically significant representation of the entire application population. They add up to the 429 flow conditions in the total database.

fails3

Final telling observations

We have looked at the many causes for a user to take the difficult step of retrofitting a control valve in the field. This step only happens after many trials to fix a problem valve.

The retrofit of almost 500 valves has shown the power of assuring that the design meets a minimum trim exit fluid kinetic energy density criterion. This energy control feature plays after all other known valve design criteria have implemented.

The data presented shows when the kinetic energy level is not reviewed, very significant jet energy results. This energy is available to enforce and amplify any damaging impact of the trim exit jets and feed the turbulence that can negatively influence the piping system.

One of the most significant observations is the very high kinetic energy levels existing in the original valve designs. These designs failed repeatedly.

These failures became successful applications, for a wide range of original suppliers, applications, and conditions that encompass the entire control valve spectrum. The only change was to reduce the fluid energy level exiting the valve trim to acceptable levels.

About the Authors

Herbert Miller, P.E. (hlmiller37@cox.net) is a senior member of ISA and serves on the Control Valve Standards Committee SP75 and subcommittees on noise and cavitation. He has two degrees in mechanical engineering and has written 56 articles on heat transfer and fluid flow technology. Laurence Stratton (lrs@ccivalve.com) is manager of technical services and automation at Control Components Inc. He is a member of the ISA SP75.07 Control Valve Standards, Noise Subcommittee. Mark Hollerbach (mah@ccivalve.com ) is manager of design engineering at Control Components Inc. He has degrees in physics and mechanical engineering.

RESOURCES

Control Valves By Guy Borden, Jr., and Paul G. Friedmann
www.isa.org/controlvalvehandbook

Establishing Control Valve Trim Flow Velocity By Gerald Liu, Herb Miller, and Laurence Stratton
www.isa.org/link/estabcontrol

ISA-SP75, Control Valve Standards
www.isa.org/isasp75

The Case for a Kinetic Energy Criterion in Control Valves - Part 1 By Herbert Miller, Laurence Stratton, and Mark Hollerbach
www.isa.org/link/KECriterion