You are on page 1of 5

The Control Valves

HiddenBottom Line Impact on the


By Bill Fitzgerald and Charles Linden

(Part 2)

Part 1 of this article (VALVE Magazine, Summer 2003) made the case that the control valve, as part of the overall process control infrastructure, is often overlooked when end users consider ways to improve financial performance in their plants. One of the prime reasons for this problem is that control valves are generally selected and maintained as if they were static devices and not part of the highly dynamic process control system.
2
VALVE MAGAZINE

n Part 1, we highlighted the control valves dynamic characteristics, which are generally not well understood and can have a significant impact on the ability of the system to perform properly. In Part 2, we will examine one other characteristic that can be overlooked when evaluating overall return on investment for valve-seat leakage. Well conclude with a discussion on recommended maintenance practices and present several case studies that illustrate how valves operating at less-than-optimum conditions could be costing the plant money.

Control Valve Leakage


On the surface, this whole subject can appear to be pretty simple. However, ensuring that control valves will shut off properly when installed and throughout their service life requires an understanding of the physical principles behind shutoff. First, how much leakage can be permitted for a given control valve application? ANSI specifies six different leakage classes. A Class I valve really is not required to shut off at all. From Class II to Class VI, the allowable leakage decreases, as illustrated in Figure 1. Class II is a very basic valve that doesnt ever really have to shut off. Class IV typically can be achieved with a standard construction control valve. Class V requires some fairly sophisticated changes to the seating profiles if the valve is metal-seated, and Class VI normally requires some type of soft seat (elastomeric seal). While this table covers a large range of possible leakage levels and is a laudable attempt at standardization, it has some idiosyncrasies worth noting. The test medium changes from air to water, and there really is no good way to correlate results from one type of medium to the other. Note also that the Class V test is done at actual service pressure, while the other tests are generally run at 50 psid or less.

C o n t r o l Va l v e s H i d d e n I m p a c t

Figure 1. Leakage specifications

Figure 2. Sectional of typical valve construction

This is important, because the test pressure used can have a relatively serious impact on the valves ability to shut off since the pressure drop across the valve plug and seat results in an unbalance that will change the seating force between these two parts, and the seating force is one of the primary determinants for valve leakage. If you examine a typical valve construction (Figure 2), it is clear why this happens. The pressure drop either lifts the plug, tending to open it, or pushes down on the plug, increasing the seating force and assisting with shutoff. So, while these ANSI guidelines do help to determine relative ability of a given valve design to shut off under somewhat consistent test conditions, they really dont tell the end user how well a valve will shut off in actual service conditions, with the exception of the Class V test where the service medium is a fluid like water. The Class V test is unique in this respect since it is the only test done under actual service pressure drops. What this means is that a valve can pass the test requirement for a given leak class (other than Class V) but may not meet the real requirements for the valve in the field. This is an important point since many end users are under the mistaken assumption that simply specifying an appropriate leak class will give them what they need. In reality, to be sure a valve shuts off in service, test it under actual service conditions, or perform an analysis to show that changes in seating load in service will not adversely affect shutoff. This is not as complicated as it sounds since most reputable manufacturers have developed fairly sophisticated actuator sizing tools that can easily determine the impact the pressure drop can have so that it can be compensated for in service. To summarize, make sure you really understand your valve shutoff needs given the application, and select a valve that will not only give you that during the production test, but also installed in the line. Also, before wasting time servicing a leaking valve, make sure the design is one that will provide the level of shutoff required. Many of these valves are actually

performing just as designed, and no amount of maintenance is going to improve performance. You may simply need to change to a different type of valve trim that is better suited to the leakage class you need. Assuming you have properly selected the valve to begin with, how do you ensure that the leakage level will not increase as the valve sees service? Here are some basic rules to observe that will maximize the chances of proper shutoff over the service life of a typical control valve: Make sure there is good contact between the seating surfaces on the plug and the seat ring, a necessary condition for good shut-off performance. If the actuator and valve are not properly connected, aligned, and adjusted, these two surfaces may never come in proper contact, ensuring poor performance. The issue of axial alignment between the valve and actuator stems deserves special mention. Take care to ensure the two stems are aligned properly when the actuator is assembled on the valve body assembly. More than about 1/16 inch of misalignment on a small valve or 1/8 inch on a larger model could result in difficulties getting the valve to shut off (and will also hurt packing performance). If excessive misalignment is present, corrective action must be taken to bring the two stems closer together. Assuming good contact between the mating surfaces, you need to ensure there is sufficient seat load to bring them together and hold them tightly under service conditions. The required seating load is normally a function of the shutoff class and the size of the port, and is usually given as lbs/lineal inch of seat for a particular valve. The higher the load per lineal inch, the better the shutoff. The actuator load also has to compensate for whatever unbalance is present due to the pressure drop across the valve. This will ensure the parts stay together and maintain the specified load. The mating surfaces must also be machined and prepared to ensure good solid contact all the way around. Normally, line
FALL 2003

C o n t r o l Va l v e s H i d d e n I m p a c t

Figure 3. Seat load diagnostic scan

Figure 4. Expanded seat load diagnostic scan

contact is the best as this provides the highest contact stress on the parts, which ensures they conform to one another. This typically is achieved by mismatched angles on the two faces or having one face cut at an angle and the other on a radius. Note that lapping of seats is a common practice in the industry, but some plants rely too much on lapping to bring seats into contact. Seats should be re-cut first to ensure good contact all the way around, and then lapping should be done to improve the surface finish of the mating surfaces. Lapping will not correct major problems with quality of the seating surfaces. It is a finishing step. And finally, there should be no debris caught between the seat or damage to the seat that could result in poor localized contact between the two pieces. If these conditions are met, the valve shutoff performance should be satisfactory. And to keep it functioning over time, do everything possible to ensure that these conditions dont change. That means a regular check of seat and actuator loading and, if possible, a seating signature evaluation. Fortunately, the advent of air-operated valve diagnostics has made checking seat condition and loading very easy. Simply run a diagnostic scan, either with a portable testing device or a valve-mounted smart positioner, to determine whether there is good seat condition and the proper load. The output for one of these tests is shown in Figures 3 and 4, where we have plotted actuator pressure vs. travel for the valve to show how much actuator load is available to load the seat at the closed end. Figure 4 shows the expanded view of the seating profile, which provides clues about the seating condition of the mating surfaces. As long as these profiles look good, the valve should function as designed. One other precaution deserves mention here. This situation can significantly impact valve shut-off, and it is of particular concern in power plants, where high-temperature graphite packing is employed. Many valves as shipped work fine. Over the years, the valve packing may end up getting adjusted to the point where, with high-friction graphite, it begins to add sig4
VALVE MAGAZINE

nificantly to the overall friction in the valve assembly. If the friction gets too high, it can begin to detract from the force available to seat the valve and provide proper shut-off. This effect is sometimes lost on the end user that tries to correct a packing problem by over-torquing the packing and creates a seat load and leakage problem in the process. Packing is another source of significant leakage in process plants and is particularly bothersome as it results in a release into the atmosphere as opposed to the controlled leakage going past the valve seat. The sealing surface between the stem and the packing needs to be carefully maintained. The better the finish, the better the initial seal and the longer the life. Take care not to damage this surface using things like emery cloth to polish it. A detailed discussion of packing performance is outside the scope of this article but, in summary, the end user should view packing as a precision subcomponent in a control valve assembly, which needs to be tightly controlled to avoid leakage.

Seat Leakage
The issue of seat leakage is important as excessive leakage in a valve can result in these two negative issues cropping up, both of which impact the bottom line: One factor is the loss of fluid downstream and the negative impact it can have on operational performance. For example, many critical steam valves in a power plant end up leaking either to the atmosphere or to the condenser so the generated steam never gets a chance to get to the turbine. Studies have shown that a typical power plant can lose 1-2% of its capacity due to control valve leakage if the characteristics just described are not well understood. Another issue is that leakage generally results in trim damage to both the seating surfaces on the plug and the seat ring. This translates into additional costs to tear apart the valve for repairs and can increase the amount of money spent on replacement trim. Long story shortleaking valves can cost a typical process plant $1 million to $2 million a year in lost capacity and excess operations and maintenance (O&M) expenses.

C o n t r o l Va l v e s H i d d e n I m p a c t

Figure 5. Plant performance optimization

Proactive Control Valve Maintenance

So far, weve discussed how you need to make sure you consider more than just static performance when selecting control valves. But even if you do your homework and have selected the best valve from an overall financial performance standpoint, how can you be sure the valve will sustain that performance over the life of the plant? This is where your approach to maintenance comes in. A full spectrum of maintenance approaches can be applied to control valves. As illustrated in Figure 5, they range from totally reactive and run to failure, to a full-blown predictive maintenance program. Most customers these days understand the value of the more proactive approaches to maintenance and apply them wherever they can. Most studies conclude that predictive maintenance is up to five times less expensive than scheduled maintenance, and 10 times less expensive than corrective or reactive maintenance. Unfortunately, until recently it has been tough to adopt a proactive approach to control valve maintenance for a couple of reasons. First, it can be difficult to establish performance criteria. For instance, how do you determine when a valve is not functioning properly and needs maintenance? Control valves are relatively robust mechanisms and can continue to function and respond to an input signal even though their dynamic performance has already degraded to the point where it is impacting the bottom line. But, if past practice focused on the static performance of the valveas weve argued earlier in this articlea valve might be judged to be performing well enough even after it has failed from a dynamic standpoint. The result is that no clear definition of failure typically has been established. Second, there has been no easy way to determine the current Real Life Experience operating state of the valve in question. Nearly all predictive As you can see, the game has changed. Focusing on dynamic permaintenance programs are based on being able to determine the formance of a process control system by optimizing control valve current performance levels of a given piece of equipment and performance can have a very positive impact on your O&M compare it to a baseline, typically referred to as condition-based budget and your throughput. Utilizing modern digital equipment Circle Reader Response #12
FALL 2003

maintenance. As deviations grow between the acceptable baseline and the current performance, they can be tracked and maintenance planned for a period before the deviations become significant. With control valves it has been difficult to exactly determine the current performance because there was no easy way to monitor and report on the condition. As a result, most maintenance has been either unplanned or tied to some calendar date in the hopes that regular service would keep most problems from becoming too severe. In most cases, these approaches cost the end users dearly since failed valves not only tore themselves up but resulted in less-than-optimum process control performance. All that has changed with the advent of digitally based field devices that feature two-way communication between the device in the line and the control room. The end user can now monitor a piece of equipment (such as a control valve) online through a digital positioner mounted on the valve and determine the overall operating performance without going to the field or breaking into the line. Once monitored performance levels drop below a predetermined threshold level, the diagnostic device can even provide clues as to the root cause by checking a dozen or so of the underlying operating characteristics of the valve. As can be seen in Figure 6, once a problem is detected, an alarm button is triggered and then built-in software can explain the problem with recommended corrective action. This new technology addresses both of the problems just highlighted in that it permits a full-blown condition assessment to occur on critical valve characteristics and compares them to well-established norms based on equipment testing. So we end up knowing very well what we should expect from our control valves in terms of performance, and we also have a painless way of remotely monitoring the equipment so we can determine when a maintenance intervention might be required. We can finally predict when maintenance problems might occur and employ just-in-time proactive techniques that allow us to schedule maintenance so it has the minimum financial impact on the operation of the plant while ensuring the performance of our control valves are up to standards. Taken to the ultimate level, we can then categorize control valves by criticality rating and custom design the maintenance and inventory programs around the rating, focusing maintenance activities and spare parts inventories on those valves that have the greatest impact on operations. We can also use our predictive capabilities to cut down on inventories since we no longer have to maintain safety stock in case of a reactive failureanother significant benefit to the bottom line.

C o n t r o l Va l v e s H i d d e n I m p a c t

Figure 6. Sustaining performance with digital valve intelligence

to monitor your valves and establishing a state-of-the-art maintenance program can further boost your bottom line. Well complete this article by reviewing a few case studies: A coated-paper machine was having production and quality problems. A dynamic audit of the process control system identified stability and control problems in the headbox flow control system. A faster, more stable valve was installed and new loop tuning applied; sheet quality went up by an order of magnitude; and overall production rates improved by 1.4% due to reduced times to change grades. Bottom-line annual impact: $1.1 million. A Canadian coal-fired plant was running 19 MWs under its rated capacity and could not be automatically dispatched because its ramp rates were too slow. A loop audit found 24 of 35 loops were suboptimized due to aging and obsolete control valve assemblies. Corrective action included installation of an upgrade to the DCS system and the installation of 24 new control valve assemblies with modern digital positioners. The investment: $900,000. The payback: $1.4 million a year, primarily due to increased capacity related to smoother operation of critical loops. A crude unit had a main feed valve deadband (see Part 1 of this article) of 6%. This resulted in fluctuations of +/-12 psi to the de-salter units in the pre-heat train. The setpoint had to be lowered due to the fluctuations. A faster, more stable valve was installed with a digital positioner, and fluctuations dropped to +/-1 psi, allowing the loops setpoint to be increased. This higher setpoint resulted in an increase in flow of 2,000 barrels/day, an increase in revenue for the plant of $1.9 million a year.

In another crude unit, fired heaters were employed. The fuel gas valve was found to be unresponsive, and this resulted in fluctuations in temperature of +/-9 F and an average fuel consumption of 18.7 MMSCFD. After the valve performance was optimized, the fuel consumption dropped to 17.5 MMSCFD and the temperature fluctuations dropped to +/-2 F. The fuel savings alone amounted to $500,000 a year. A nuclear plant in the Midwest is looking at a 12% uprate in power output from its plant, which will greatly improve overall plant financial performance. A lot of attention was being paid to the major plant components like the turbines and the steam generators, but as plant personnel worked on determining the implications of the power uprate on some of the subsystems within the bowels of the plant, they began to find a number of critical valves that were becoming system bottlenecks. The uprate was driving higher pressures, temperatures, and flows throughout the plant, and in some cases the old valves could not handle these new operating conditions. As a result, plant personnel are now drilling down into their control valve infrastructure to ensure that these components will not keep them from obtaining operational goals. By converting to new equipment to handle the more stringent operating conditions, they are using the occasion to upgrade to new technologies that will enable them to take advantage of the operational advantages mentioned earlier in this article. A 12% uprate in a typical nuclear plant could mean as much as $24 million in increased annual revenue for the plant.

Conclusion
As you can see from the previous examples, there is gold to be mined if one takes the proper approach to the care and feeding of the control valve installed base. New technology makes this approach easy to adopt and the paybacks are tremendous. Progressive process companies looking for the next opportunity to improve performance need look no farther than their plants lowly control valves. Every one is a profit opportunity! VM Bill Fitzgerald is Vice President, Steam Conditioning and Nuclear, Emerson Process Management (Fisher Controls), Marshalltown, IA; 641.754.3865; www.Fisher.com; Bill.Fitzgerald@emersonprocess.com. Charles Linden is Engineering Technical Specialist, Omaha Public Power District, Fort Calhoun, NE; 402.533.6930; www.oppd.com; clinden@oppd.com.

VALVE MAGAZINE

You might also like