Aplicaciones Valvulas de Control

Volume EMC 21.
01
Control Valve Applications

Herbert L. Miller
Application Categories Valves in Parallel and Series Frequent Application Problems
Taken from the Practical Guide Series book: Control Valves
ISA is the international society for measurement and control
Notice The information presented in this publication is for the general education of the reader. Because neither the authors nor the publisher have any control over the use of the information by the reader, both the authors and the publisher disclaim any and all liability of any kind arising out of such use. The reader is expected to exercise sound professional judgment in using any of the information presented in a particular application. Additionally, neither the authors nor the publisher have investigated or considered the effect of any patents on the ability of the reader to use any of the information in a particular application. The reader is responsible for reviewing any possible patents that may affect any particular use of the information presented. Any references to commercial products in the work are cited as examples only. Neither the authors nor the publisher endorse any referenced commercial product. Any trademarks or tradenames referenced belong to the respective owner of the mark or name. Neither the authors nor the publisher make any representation regarding the availability of any referenced commercial product at any time. The manufacturers instructions on use of any commercial product must be followed at all times, even if in conflict with the information in this publication.
Copyright 2000 Instrument Society of America. All rights reserved. Printed in the United States of America. No part of this publication may be reproduced, stored in retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher. ISA 67 Alexander Drive P.O. Box 12277 Research Triangle Park North Carolina 27709
Editors Introduction
This mini-book is available both in downloadable form, as part of the ISA Encyclopedia of Measurement and Control, and bound in a print format. Mini-books are small, unified volumes, from 25 to 100 pages long, drawn from the ISA catalog of reference and technical books. ISA makes mini-books available to readers who need narrowly focused information on particular subjects rather than a broad-ranging text that provides an overview of the entire subject. Each provides the most recent version of the materialin some cases including revisions that have not yet been incorporated in the larger parent volume. Each has been re-indexed and renumbered so it can be used independently of the parent volume. Other mini-books on related subjects are available. The material in this mini-book was drawn from the following ISA titles: Control Valves, a Practical Guide Series book edited by Guy Borden, Jr. and Paul G. Friedmann, Chapter 12. Order Number: 1-55617-565-5 Internet: www.isa.org Phone: 919/549-8411 Fax: 919/549-8288 Email: info@isa.org
To order:
Table of Contents
Chapter 12 CONTROL VALVE APPLICATIONS 1 by Herbert L. Miller The Application Categories, 2 Valves in Parallel, 14 Valves in Series, 23 Frequent Application Problems, 25 References, 37 INDEX 39
12 Control Valve Applications

The purpose of this chapter is to present the information that will help process
engineers specify control valves properly. It will show that not all of the control valves attributes are uniquely required for the process design. There are many common requirements for the categories of applications, regardless of the industry or the process. By recognizing the common requirements and their importance, the designer can specify the most economical design that meets performance needs. Many books on control valves concentrate on the design aspects of the valve, and little information is included on the subject of applications. An exception to this is Reference 1, which provides the practical design input associated with a good valve application. This reference provides many examples of control valve applications and illustrates the importance of knowing the control characteristics dictated by the process. Another source of good application knowledge is the control valve manufacturers. Many have prepared specific application brochures and guidelines for experiences they have frequently encountered. These brochures have a wealth of background information because they integrate the experience of many customers involving the same applications and the experience gained across different industries encountering similar applications. Many practical guidelines regarding selection and installation practices are also provided in Reference 2. In the discussion in this chapter all applications of control valves are classified into four categories. In describing the applications, the attributes that will be emphasized are those important in each of the four categories, independent of the many different application names used from industry to industry. Valve types or materials will not be discussed unless they are key to a successful application. Selection among valve types is covered in Chapter 14 of the Practical Guide Series book Control Valves; selection among materials was discussed in Chapter 11 of Control Valves. There are exceptions within every category because of the needs unique to a specific process within an industry. An example of this would be a pump recirculation valve that would normally fail open to protect the pump but, in some nuclear power plants, fails closed by design. It is assumed in the following discussion that the valves have been sized correctly. A correctly sized valve is a major factor in a successful application. The use of the ISA data sheet, Reference 3, ensures that all the pertinent information is available for the sizing decisions. It is recognized that many companies have their own valve data sheets, but many contain omissions that result in important information not being passed onto the valve designer. The ISA data sheet standard provides a good checklist of key information that may impact the users specific
1
application. Chapter 6 of Control Valves discusses sizing; Chapter 22 of Control Valves describes a valve sizing computer program. In the discussion in this chapter emphasis is placed on the attributes that are unique to the application and not so much on problems that occur due to poor selection. Of course, being aware of these attributes can help assure that you make a proper valve design choice. The application categories are as follows: Process control/feed regulation Continuous letdown Intermittent letdown Recirculation
In addition to these four categories, a separate discussion is provided for valves that are installed in parallel or series applications. These configurations occur in many processes so a specific section is devoted to them. Also discussed in this chapter are common valve application problems caused by improper specification or installation practices. The definitions of terms used in this chapter are discussed in Chapter 3 of Control Valves and are based on the ISA standard S75.05, titled Control Valve Terminology.
The Application Categories

Process Control/Feed Regulation
In this category, a control loop or system has been pressurized and has a valve that is controlling the feed to a process in response to a control signal. The control signal could be based on a need for flow or used to maintain a pressure or temperature to the process or to maintain a fluid level. This setup would typically require continuous valve operation. Frequently, there are parallel control valves to aid in start-up or shutdown when wide rangeability is required. To provide proper valve control trim characterization is usually needed. This valve trim is required for process startup and shutdown, which, of course, cannot be avoided even though it may not be frequent. When the system is operating at low loads both the flow rate and system pressure drop are low, but valve inlet pressure can be high due to pump runback. This results in high pressure drop across the control valve at low flow. At full flow (full load) conditions, the system pressure loss has increased to cause lower pressure drop across the control valve, as shown in Figure 12-1. These conditions point to the need for good valve rangeability, which usually results in reciprocating stem-type valves with equal percentage trim characteristic. With parallel start-up valves and/or low-pressure processes, this regulation may be possible using a rotary-type valve with a modified ball design. Figure 12-1 shows a schematic of the situation covered by this application. Also shown is a frequently encountered pressure-versus-load curve for the application. As seen at the low-load condition, the pressure drop across the valve is very high. This is the difference between the pressure developed by a pump or compressor upstream and the pressure drop through the system. The names assigned to the valves in this application are numerous. Some of them are as follows:
2
Flow control Level control Pressure control
Pressure reduction Regulator (flow) Throttle
Figure 12-1. Process Control/Feed Regulation Application.
Key attributes for this application category are as follows: 1. Accuracy of control. The output of this valve affects almost all of the downstream functions in the process. If a steady feed rate cannot be maintained, then all of the pressures, temperatures, and flows will be continuously changing, sometimes with an increased gain or error signal multiplication. Thus, the quality of the process output, whether it is electric power output, clean gas, paper, chemical product, or other, is impacted by the ability of this valve to maintain an accurate output. In some processes, the variation of an inaccurate valve output may not be noticed at the process output because of a damping due to long residence times. In these cases, however, the continuous variation may lead to long period oscillations and fatigue in process equipment near the valve. Small signal response is often important to optimize efficiency or process
3
output. 2. Rangeability. This is the second most important attribute for this valve because the process must be controlled during start-up and shutdown. Also, during operation, abnormal conditions may frequently exist that create the need to operate at reduced loads. In the electric power generation field, it would be necessary to reduce load depending upon whether the load dispatcher responded to diminished use of electric power. In the chemical process field it may be a breakdown or maintenance of a parallel train in the system that calls for the process to operate at reduced load. To handle the start-up and shutdown conditions, it has been traditional to use two valves in parallel to achieve the needed rangeability. There are, however, valve designs today that can handle this function in one valve body (see the discussion on parallel valves later in this chapter). To achieve good rangeability and linearity of the valve-flow-to-stem position (installed linearity), the valve trim must be characterized. Thus, at the low-load conditions, a valve capacity versus position curve (characterization) will look like that shown in Figure 12-2. This characterization not only provides better rangeability and control but also assures that the valve closure member does not operate near the seat and thus minimizes damage to the critical seating surfaces as a result of excessive fluid velocities. Depending upon the low-load pressure drop, other measures may be necessary to limit high fluid velocity erosion such as the use of multi-stage trim designs. These designs become a consideration when pressure ratios, p1/p2, across the valve exceed three; otherwise pressure drop across the valve could result in cavitation or excessive noise. Cavitation (see Chapter 7 of Control Valves) and excessive noise (see Chapter 8 of Control Valves) can occur at pressure drops as low as 30 psi (0.2 MPa) for some fluid conditions. The characterization and trim-erosion considerations are very important because they contribute significantly to the most vital function of this valve and that is accuracy of feed control. In this application, other attributes are usually secondary. These are as follows:
Figure 12-2. Cv/Flow versus Travel.
3. Failure mode. Fail safe is the preferred failure mode and is usually in place or last position. This means that all actuator types can be used; pneumatic, electric, or hydraulic. There are always exceptions. One is in the nuclear power industry where the normal feedwater regulators are failed closed to eliminate an uncontrolled leak path, while the auxiliary feedwater system takes control. 4. Stroke speed. Stroke speed is generally not a consideration because most processes cannot change load quickly because of stored energy or product in the system. Thus, normal valve speeds are quick enough to respond to the demands imposed. 5. Shutoff. This is usually not a consideration because these valves are seldom shut. So, an ANSI/FCI-70-2 leakage class III or even less is sufficient in most cases.
The most obvious example of a valve in this application would be a boiler feedwater regulator valve where a constant-speed feed pump is used. The pressure condition for the valve inlet (pump output) and valve outlet (system pressure) are shown on Figure 12-1. A representative set of conditions for this application would be a flow rate of 55,000 lb/h (metric 25 t/h) at a pressure differential of 735 psi (5.07 MPa) at start-up. At a full-load flow rate of 990,000 lb/h (450 t/h) the pressure differential is 30 psi (0.20 MPa). Thus the pressure drop across the valve changes by 25 times as flow is varied by 20 times over the load range. The resultant valve rangeability exceeds 100 in that the minimum Cv is 4, and the maximum is 370. However, to provide control at the maximum condition, a Cv of over 400 is required.
A less obvious example of a valve in this application is that of a gas transmission metering valve. The highest flows occurs at the lowest differential pressure, and the lowest flow rates occur at the highest differential pressures. This is caused by variable load resulting from changing weather conditions. As noted by Reference 4, When it is cold outside, consumers require large volumes of gas, which results in great demand placed on the local distribution company, LDC. The LDCs in turn require a greater amount of gas from the pipeline system. This results in small differential pressure between the pipeline and the LDC, with corresponding high volumes of gas. This situation requires a valve with a high capacity. In contrast, when the weather turns warmer, consumer demand diminishes and the LDCs place very little demand on the pipeline. The pipeline system pressure remains high because of the reduced demand. Therefore, high differential pressures occur between the pipeline and LDC with very little flow occurring. This situation requires a valve that can handle low capacity.
A range of conditions for the gas transmission metering valves would be a low flow pressure differential of 900 psi (6.2 MPa) and a high flow pressure differential of 100 psi (0.7 MPa). Another valve selection criteria in this application that may be a consideration is valve-generated noise. With the high compressible fluid pressure differentials, some valve designs could produce noise levels in excess of 100 dBA.
Continuous Letdown
In this application category, the valve is located between two large reservoirs of different pressures where the downstream reservoir could be the atmosphere. In this case, the valve sees a constant pressure drop, so flow control is established by the valve position. There is no need to provide anything other than a linear trim characterization for this purpose unless extended duty at low feed rates is also required by the process. An equal percent closure member will allow travel farther off the seat for extended operation at low flows.
Reliability and ruggedness are the keys to this continuous-duty application, so selecting the proper valve design for the pressure drop condition is important. Table 12-1 provides a guideline for which type of valve to select for specific pressure ratios.
Figure 12-3 shows a schematic of the situation covered by this application. The most usual control condition is that of either upstream or downstream pressure or level control. The feed rate is varied as necessary to maintain pressure or level. An example of upstream pressure control is the letdown from a process, such as a chemical reactor vessel, a gas reservoir, or a reservoir level control. Probably more numerous are the downstream pressure control requirements, such as for steam to an auxiliary turbine or process reactor and for gas flow into a distribution system such as for multiple burners. The control variable can also be flow instead of pressure. This would be the case for burner control valves, spray or mixing valves used for pressure or temperature control, and gas transmission pressure-reducing valves.
Figure 12-3. Continuous and Intermittent Letdown Application.
In specifying conditions for these valve applications, the most common error is omitting the off-load operating conditions. Usually, only one set of operating conditions is provided, that of steady-state full load. The off-load conditions are important because of the complexity of most plants start-up procedures, which require holding at partial loads for extended periods. Thus, the need for good rangeability is usually missed in these applications, resulting in poor valve selection with the corresponding result of poor control of the process start-up or part-load performance. Typical names assigned to these valves are: Attemperation Blowdown Flow control Letdown Level control Pressure control Pressure regulator Reducing Spray
Key attributes for the continuous letdown applications are the following: 1. Accuracy of control. As with the process control valve, this is the most important function for this valve. Maintaining a near constant pressure, flow, or temperature condition is essential to the process product quality and reliability of the equipment affected by the control. A continuously varying output would have many short- and long-term damaging influences. This application is fairly routine in terms of its demand on the control valve. Many different types of valve designs can handle the conditions imposed without detrimental effects. There, of course, can be exceptions where the attributes discussed here may assume more importance than indicated here. 2. Rangeability. This is not generally an important issue because most valve designs provide sufficient rangeability to meet the process needs. A reasonable need is for a fifteen-to-one rangeability. Since the pressure drop across the valve is constant, the trim of the valve is linear, and for most valves the travel position is a reasonable approximation of the load. 3. Failure mode. This depends upon the specific application, and all modes of failures on loss of power to the actuator are used. If a generalization is to be made, it would be to have this valve fail-in-place so that the system is not disturbed by an abrupt change in pressure or flow conditions. Some engineers may prefer to fail the valve closed so that the operator is alerted to the valve failure.
7
4. Stroke speed. As with the process control/feed regulation application, this is usually not a consideration. 5. Shutoff. This is not an important consideration because the valves are rarely closed unless the process is off line and isolated by other system valves.
Intermittent Letdown
As with continuous letdown applications, the valve reduces the pressure between two large reservoirs, and frequently the downstream reservoir is the atmosphere. Obviously, for fluids damaging to the environment and public health, a downstream reservoir would capture the flow for further processing, such as through a flare arrangement, where the fluid is burned before release. The pressure drop could be constant or variable depending upon the specific application. The variable case would occur on a blowdown situation, so the inlet pressure would decrease with time. The valve stroke would be increased if a near constant flow rate was desired to minimize blowdown time. Or, for the case of upstream pressure control, the stroke would be varying depending upon the process input and output of the upstream reservoir. Figure 12-3 is the same for continuous and intermittent letdown applications. As implied by the name, the only difference in the two processes is the time of operation or duty cycle on the valve. Even though the valve is used less frequently, the conditions for the valve represent a tougher service. In this case, the control valve must perform the dual functions of providing control and tight shutoff. The latter usually means a Class V or VI leakage (ANSI/FCI-70-2), but in many cases this is not sufficient. A block valve leakage requirement must be used to provide a sufficient criterion for the permissible leakage. The reason a tight shutoff is needed is that any leakage through this valve means a loss of process fluid that is needed upstream to make the process product or output. Valves in intermittent letdown service primarily perform a bypassing function. The valves are opened to bypass the entire process or parts of the process during a start-up/shutdown function or a safety-relieving function. In some cases, the valve can perform both functions of bypassing for control and safety relief. However, local codes must be checked to see if a dual role is permitted. Typical names assigned to various intermittent letdown valves are shown in the following list. Frequently, the name of the equipment being bypassed or blown down is used in combination with these labels: Antisurge Dump Reject Auxiliary Extraction Relief Blowdown Flare Start-up Bypass Injection Vent(ing) Depressurizing Letdown
The attributes for the intermittent letdown application are as follows:

8
1. Accuracy of control. For most valves in this application this is not an important function because most control valves provide sufficient resolution to meet the requirements. An exception frequently exists when the valves are in a start-up bypass function where a fine, accurate control is needed to maintain flow or upstream pressure conditions over extended periods. The pressure drop across these valves is constant for most of the operating conditions. As a result of this and the accuracy needs, a linear characterization of the trim is normally sufficient. 2. Rangeability. Normal design capability is sufficient. 3. Failure mode. Normally, this valve is in a fail-close configuration, its normal status, so that if power is lost to the valve the process is not disturbed unnecessarily. Occasionally, process system needs will require the valve to fail in its last position.
When the valve is in a safety role, the valve configuration must be in the fail-safe direction, which is usually fail-open.
4. Stroke speed. The dual functions of these applications dictate the speed of operation. When the valve is used in a safety function, the speed needs to be very fast, on the order of one-half to five seconds. The safety function could be to protect personnel, equipment, or the process output. This would only be necessary for the opening direction because, in this mode, it is usually relieving pressure from an over-pressurized upstream reservoir. When the valve is used in the process bypass mode, for start-up and shutdown function, speed is generally not a priority consideration. Speeds can be achieved without requiring special considerations because commercial positioners usually have sufficient capacity for pneumatic actuators. Electric drives, although they tend to be relatively slow, are fast enough. The use of hydraulic actuators for speed purposes would be over design except when the valve is used for safety relief. 5. Shutoff. This is a key and critical function of valves in this application, regardless of whether the valve is performing a bypass or a safety role. Any loss of fluid through this valve reduces the process efficiency. Either pumps or compressors must work harder to overcome the leakage or the maximum loads achievable are reduced. The minimum shutoff requirement would be a Class V leakage for a control valve. However, often a Class VI or even a MSS-SP61* block valve closure is prudent for reliable long-term valve operation. For this reason, many control valve
The MSS-SP61 is a standard issued by the Manufacturers Standardization Society of the Valve and Fittings Industry, Inc. The SP61 standard is titled Pressure Testing of Steel Valves. It covers the requirements for the shell and seat closure pressure testing of steel valves. The standard is not intended to be used for control valves. However, as noted in this chapter, there are many control valve applications in which the valve must perform the role of modulating control and then become a block valve when shut. For many valves that are used in safety relief and plant start-up or shutdown applications, these dual roles are frequent requirements. If only a control valve leakage class is imposed then some leakage is permitted. The amount of leakage may not be bothersome when the valve is new. However, since these valves are normally shut, fluid erosion of the closure members due to the leakage results in a complete failure of the valve to shut off. In the extreme case, the ability to control is also lost. Manufacturers can produce control valve designs capable of meeting the MSS-SP61 standard. The standard calls for a leakage test pressure drop of 110% of the 100F (38C) ANSI class rating pressure. If this high pressure may damage the seat, a possibility in control valve designs, then 110% of the maximum differential operating pressure may be used. The permissible leakage under this standard shall be less than 10 cc/h of liquid per inch of diameter of the nominal valve size.
designs use the fluid pressure to help the closure member maintain a tight seating force. Soft seats are frequently used, but these may not be reliable in the high pressure drop situations, over 1500 psi (10 MPa), that frequently occur in this application category.
A typical example of this application is the steam turbine bypass valve. This valve is installed to bypass the steam around the turbine until pressure and temperature are at appropriate turbine start-up conditions. The valves are closed as the turbine is brought on line. These turbine bypass valves can also be used for pressure relief valves per some code regulations, particularly in Europe. For safety purposes, the American Society of Mechanical Engineers Boiler and Pressure Vessel Code does not permit the use of this bypass control valve as the pressure relief valve. For the turbine bypass function, the most important valve attributes are usually shutoff, control, and opening speed.
Another example involves the turbine drive of a synthetic gas compressor in an ammonia plant. The primary purpose here is not turbine protection but assuring there is high-pressure steam available to a downstream reformer if the gas compressor is tripped. Steam is needed to avoid catalyst deterioration resulting from carbon deposition and to provide time to achieve an orderly shutdown of the reformer. Thus, the trip-open function is very important in this application, requiring that the valves open in one to four seconds. Also, because of the fairly high pressure drop of 1000 psi (6.9 MPa), noise control is also a consideration in valve selection.
Recirculation
This application could be thought of as a subset of intermittent letdown in that the duty cycle is intermittent and the function performed is generally a process bypass situation. The bypass need is usually driven by the start-up, shutdown, or system-upset conditions. However, because the applications listed under this category are usually the most severe within the process system, they need special consideration to assure their reliable and long-term operability. Thus, the recirculation application is reserved for the cases in which either a pump or compressor is bypassed, as shown in Figure 12-4. In this case, the fluid that has been pressurized or compressed is reduced in pressure and returned to the pump or compressor inlet reservoir. The valves experience a wide range of pressures and temperatures, however, within a specific application, rangeability and control are seldom critical. The valve is usually closed, and its performance is judged by how well it shuts off. If shutoff is not maintained, the fluid must be repressurized for use by the process at increased energy expense. There have been cases in which recirculation loss is so high that the total output of the system is measurably reduced.
10
Figure 12-4. Recirculation/Recycle Application.
This valve will see the highest pressure drop in the entire plant when you bypass the main pump or compressor. The cost of fluid leakage through this valve traditionally exceeds the purchase price of a new valve many times over. This loss is seen in reduced plant efficiency, unavailable load and the increased pumping power required to pressurize or compress the fluid stream. It is not an application where the engineer should compromise on valve selection. The engineer must select a valve that will meet the long-term seat integrity requirements to assure leak-free operation. Frequently, this would utilize a design in which the fluid pressure assists in assuring that there are good seating forces between the closure member and the seat ring. Valves in this application go by many names depending upon whether it is flow through a pump or a compressor that is being bypassed. The most common names are as follows: Antisurge Mini-flow Bypass Recirculation Dump Recycle Kickback Return Leak-off Spillback Letdown Surge control
Because most fluids are erosive, particularly liquids, a small leak through the valve will result in the rapid deterioration of the seating surfaces.
In addition to the comments about leakage and control attributes discussed under the intermittent letdown application, there are two other significant considerations concerning the selection of these valves. For liquid recirculation applications, those considerations are cavitation and vibration, and for compressor applications, they are noise and vibration. In both cases, the detrimental affect is caused by the high pressure drop associated with these valves and the accompanying low back pressure. The back pressure is usually near atmospheric but could be a vacuum if the downstream reservoir is a condenser. These applications frequently demand severe service valves that are specifically designed to handle these conditions.
11
In summary, then, the key attributes for valves in this application, assuming proper sizing, of course, are as follows: Tight shutoff. Usually pressure assisted via piloted designs or unbalanced plug designs with large actuators. Anti-cavitation/low noise trim Pipe vibration elimination Flow characteristics. Usually linear for pumps and characterized for compressors Failure mode. Normally open Stroke speed. 2 to 5 seconds for compressible; up to 25 seconds for liquids
Secondary considerations include the following:
Frequently, the speed requirement for the compressor antisurge valve drives the designer to question the capability of pneumatic actuators. High speeds can be achieved with a pneumatic system as is demonstrated by the actual field test results from a compressor recycle valve. As shown by Figure 12-5, that valve
12
opened to 95% of full stroke in only 1.7 seconds. The valve and actuator sizes and conditions were as follows:
Fluid Actuator size Valve stroke Plug weight Inlet pressure Outlet pressure Flow rate Air supply pressure Mixed Refrigerant 113 in2 24 in 1320 lb 190 psi 45 psi 1,620,000 lb/hr 100 psi 730 mm2 610 mm 600 kg 1.31 MPa 0.31 MPa 735 t/h 0.69 MPa
Figure 12-5. Valve Position versus Time Opening.
As this actual valve design demonstrates, a pneumatic actuator can achieve very rapid position changes to provide good surge control of the compressor. The air supply tubing to the actuator and the air source must be large enough to provide the rapid air flow rate to the actuator. Otherwise, speed will be limited. The air supply pressure to achieve this quick opening is dependent upon the valve hydraulic and friction forces. Thus, the air supply pressure may decrease and still allow the stroke time to be achieved, provided the pressure results in enough force to move the valve closure member. Consult the valve manufacturer if you are concerned that the air supply may be restricted.
13
Valves in Parallel
There are three primary reasons for having control valves in parallel. These are illustrated in Figure 12-6. All of the situations illustrated in the figure are driven by the process but are heavily influenced by the capability of the control valve selected.
Figure 12-6. Parallel Valve Application.
The first reason for having control valves in parallel is to have a redundant valve available in the event that problems with the primary valve develop. In this case, it is very important that the process be kept on line and running. The usual reason is the economics of the output product or the high cost of a shutdown. A shutdown may require a lengthy start-up time, or in some cases an unsafe condition could result from the lack of feed control. Certainly, the first consideration by the engineer is to use a highly reliable valve in this situation, a valve that is designed for the service conditions. An example in which these conditions would exist is in coal gasification, where erosive fluids cause valves to wear out quickly. To avoid repeated shutdown, parallel valves are used so that repair can take place while the process remains up and running. Another example is the use of parallel or redundant valves in the nuclear industry, where decay heat from the reactor core must be removed after a reactor is tripped. Thus, regardless of the reliability of the valve, the consequences of even a remote possibility that the primary valve will fail are offset by the availability of the redundant parallel valve system. A second reason for having two valves in parallel is to assure balance in the process equipment. Balance may mean equalizing temperature or chemical concentration, as in mixing situations, or splitting the flow stream for process benefit. As shown in Figure 12-6, an example of thermal balancing would be a
14
Valves in Parallel
heat exchanger that must have flow through both tube banks to assure that the tubes do not overheat. In this situation, the most common problem is that small signal changes to the valve result in too much flow change, a valve resolution problem. Generally, the resolution problem is related to the valve selection. Too much capacity has been installed, and/or the trim characterization does not provide sufficient gain for good resolution. There are valves available today that can provide extra rangeability and excellent characterization so that good control is achievable. Valves that have an ability to achieve this capability but with different degrees of success are illustrated by Figures 12-7 through 12-12. The figures illustrate various design concepts used to expand the flexibility of the valve to meet a wide range of flow conditions. In many cases, a single manufacturer will supply more than one concept. It might be argued that the rangeability and characterization needs could be achieved by the control system feeding the correct signal to the valve. The problem arises in the valves ability to respond to the small signal change with a correspondingly small flow change. The trim design and characterization are key attributes that influence the flow response to the change in control signal.
Figure 12-7. VRT Trim Valve. (Courtesy of Masoneilan)
15
INSERT PHOTO HERE
Figure 12-8. V-Line Series Noise Attenuator Ball. (Courtesy of Fisher Controls International Inc.)
16
Valves in Parallel
INSERT PHOTO HERE
Figure 12-9. Cascade Trim Valve. (Courtesy of Copes Vulcan, Inc.)
17
Figure 12-10. Mark One Globe Valve. (Courtesy of Valtek International)
Figure 12-11. Q-Ball Valve Schematic. (Courtesy of Neles-Jamesbury Inc.)
18
Valves in Parallel
INSERT PHOTO HERE
Figure 12-12. Drag Trim Valve. (Courtesy of Control Components Inc.)
19
Another example of valves in parallel to maintain balance is illustrated by the use of a three-way valve design to replace two valves in parallel, as illustrated in Figure 12-13. Control of the temperature of the fluid leaving the heat exchanger can be accomplished by bypassing a portion of the heating steam. To be successful, experience has shown that the bypass flow must be less than 25% of the total flow. As the bypass flow increases, the available pressure drop across the heat exchanger decreases with the lower flow through the exchanger. This increases the pressure drop across the exchanger steam control valve branch, which now has to finely control the steam flow at a higher valve pressure drop in order to provide accurate temperature control. The pressure drop across the bypass valve branch has also increased with the lower exchanger flow. If the bypass valve outlet pressure drops below the valve pressure ratio that causes choking, then the ability to control temperature is lost since the bypass flow cannot be varied in the choked condition. That the mixing function of the three-way valve is an equivalent to the parallel valves is also apparent from Figure 12-13. In this situation, the two fluids would enter ports B and C and exit port A in proportion to the position of the valve stroke and porting size used in the design.
Figure 12-13. Three-way Valve Substitute for Parallel Valves.
20
Valves in Parallel
When everything has been done right to achieve good rangeability, it may still be necessary to add a small valve in parallel with a larger capacity valve. This represents the third example of valves in parallel as illustrated in Figure 12-6. The smaller valve can either provide a fine control during the start-up or low load, called sequencing, or it can provide a trimming function for small process changes while operating at a higher load condition. The definition of the control logic will be based upon which need is dictated for the valves. From a valve integrity and long-term reliability standpoint, the control logic designer should work to make sure the valves do not operate for extended periods with the valve closure element very near the seating surface. Travel at a position that causes the pressure drop to occur in the flow path between the closure member and the seating surface will cause erosion of these surfaces. The result is degradation of control, instability, and excessive leakage early in the valves life cycle. When two valves in parallel are sequenced for rangeability, the valve design and control logic selection are critical to a reliable system. The valve trim should not be of an inherent linear characteristic. The reason for this is that there is a discontinuity in the gain of the two valves when switching from one to the other. This is best understood by an example for an application involving continuous letdown and constant pressure drop.
Consider a small valve with a Cv of 10 that is being sequenced with a valve with a maximum Cv of 100. Both valves have linear trim and actuators capable of one percent resolution. At the transition from the small valve to the large valve, the gain, or minimum change in Cv for the small valve is 1% of 10 or a 0.1 Cv change. After transition, the large valve gain is now in effect, and its gain is 1% of 100 or a Cv change of 1.0. Thus, the gain has increased ten times, creating a strong potential for an unstable or an oscillatory control scheme. When sequencing, the parallel valves should have a near equal percentage trim characteristic and be of equal gain on each side of the transition. This is shown in Figure 12-14 for the preceding example of two valves with a Cv of 10 and 100. It is assumed that the large valve has a minimum controllable Cv of 1/30 of the full Cv. For this case, as load is increased and the small valve reaches its maximum Cv of 10, the larger valve is opened to a position of Cv equal to 10, and the small valve is closed. The gains are equal at the transition, and thus the process control resolution is not affected. Similarly, upon reducing load, the large valve controls until the Cv is down to 3.3, after which it closes and the small valve is opened to this capacity. The amount of overlap between valves is dependent upon many factors, which includes stroke time, valve characterization, and process response time.
The transition from one valve to the other is usually quite fast relative to the process, which results in a bumpless transfer. The engineer should give consideration to the transition point so these valves are not continually opening and closing during normal operation. An operation that continually opens and closes will experience a gradual degradation due to premature wear. This higher wear rate can be avoided by performing proper Cv selection of the parallel valves. Reference 5 includes additional information on parallel logic involving control valves. The total capacity of valves in parallel is obtained simply by adding the Cv values of each valve in parallel. That is, Cv,o = Cv,1 + Cv,2 + + Cv,i (12-1)
21
where Cv,o Cv,i = overall Cv = Cv of the ith valve
Reference 6 discusses a fourth reason, controlling cavitation, for installing valves in parallel. This arrangement takes advantage of the change in the cavitation index of a valve when it opens, that is, less cavitation occurs at small openings. This reference provides measured cavitation index values for different valve designs as a function of the valve opening. In this situation, small openings refer to operation within 10 to 20% of the valve full-open position. If extended operation below 10% is needed, then other problems of control develop in that the valve trim parts may erode, vibrate, and create excessive noise. Another problem is the increased cost of a second valve, unless redundancy is a requirement. Reference 6 also discusses how cavitation could be controlled by placing two valves in series, as discussed in the next section. Again, increased cost is a disadvantage as is the potential that the upstream valve will adversely affect the downstream valves performance.
Figure 12-14. Sequencing Two Parallel Valves of = % Trim.
22
Valves in Series
Valves in Series
The use of control valves in series is usually driven by a desire to limit the pressure drop across a single valve. This may have been necessary many years ago. However, there are now valve designs that can absorb very high pressure drop conditions in a single valve without the accompanying problems of cavitation, erosion, vibration and noise. Table 12-1 provides pressure drop guidelines for different valve designs that will aid in the decision regarding whether valves in series is a possible configuration. The cost of the valves, associated controls, installation, and calibrations would then be a second consideration.
Table 12-1. Valve Type Selection Guide.
It is always desirable to use one valve because of the control instabilities that arise in these series configurations.
Valve Type Multi-stage, Multi-path Multi-stage, Single-path Single-seat Globe Double-seat Globe Angle-body Cage Pinch Butterfly, 60% Open Reduced Ball, 80% *p1/p2 = 1/(1 p/p1)
Flow Direction ----------Open Close ------Open Close ----------------
p1/p2 Limit* No limit 5.5 3.3 2.3 3.3 3.0 2.3 2.9 1.5 1.2
p/p1 Limit 1.0 .82 .70 .57 .70 .67 .57 .66 .33 .17
Control instabilities are beyond the purpose of this discussion, but they can occur when the fluid residence time between valves is much faster than the valve response time. As a general guide, the time constant of the valve should be less than the residence time of the fluid between the valves. For example, if the residence time is one second, the valve time constant should be one second or less. The time constant is defined as the time required to complete 63.2% of the total rise or decay of a step change in the signal to the valve. This is illustrated in Figure 12-15. The residence time is calculated by dividing the distance between the two valves by the average fluid velocity in the connecting piping. That is, Residence Time = L/U = AL/w where L U A w = distance between valves = average fluid velocity = piping flow area = mass flow rate = fluid density (12-2)
23
Figure 12-15. Single-Order Response Time.
Instability also results if the pressure drop across one of the valves, particularly the downstream valve, becomes small, less than 20% of the overall pressure drop assigned to the valves. To assure a stable operation from this standpoint, a good rule is to have two-thirds of the overall pressure drop across the first valve and impose 80% of the drop as a top limit. Figure 12-16 shows a case in which control valves would be put in series, that is, when multiple lines feed downstream parallel processes and the designer places an inlet pressure reduction valve upstream of the distribution manifold. Such a case would occur on a gas burner distribution system where the inlet pressure is dropped and then individual valves control the gas flow to each burner.
Figure 12-16. Series Valve Applications.
24
Frequent Application Problems
A gas transmission metering and reducing station is also an example of valves in series. A primary control valve reduces the pressure, and a monitor valve safeguards against the over-pressurization of the downstream line in case of the failure of the primary control valve. The system is designed for the primary valve to fail open and the monitor valve to fail closed. The overall capacity of valves in series if needed can be determined from the following equation: (1/Cv,0)2 = (1/Cv,1)2 + (1/Cv,2)2 + (1/Cv,i)2 (12-3)
In these definitions, the overall pressure drop is used in the expression for Cv,0, and the pressure drop across each of the components is used in the Cv,i definition. This equation must also be used when a valve and another resistance component, such as a diffuser or orifice, is installed in series. (The Cv of an orifice or a diffuser can be approximated by using thirty times the flow area, where the area is expressed in square inches.) Failure to do so can result in an installed Cv less than that required by the process.

It is worthwhile to look at some of the common problems experienced when using control valves. These problems tend to be independent of the applications but can be aggravated by the unique needs of each installation. The root cause of these common problems varies from a lack of understanding of valve design and the selection of the wrong valve type (usually because of efforts to reduce initial capital cost) to poor calibration and maintenance.
Controlling Pressure Drop

The first problem to be discussed arises from the purposes of a control valve, which are to do the following: 1. Convert energy by reducing the fluid pressure, 2. Handle deviations from ideal operation, 3. Handle the influence of unavoidable process changes, 4. Permit the smooth transition from one load condition to another. To be able to achieve these objectives, there must be some pressure loss at the valve, and this results in an increase in fluid kinetic energy within the valve. Guidelines that have developed through many years of successful applications indicate that at least 10% of the system pressure drop should be available across the control valve to provide some control. To achieve good control, a value of 30% is desired. These guides are continually challenged, as in the case of the first application category discussed in this chapter, process control/feed regulation. In this application the most efficient plant operation would be to control the full-load condition without any energy absorption, an impractical goal. Thus, special attention must be devoted to the sizing pressure drop conditions to assure that the valve can modulate the feed flow and meet the control objectives.
Excessive Fluid Velocities

Throughout this discussion of control valve applications, we have referred to the problems of erosion, vibration, noise, cavitation, and flow instabilities. All of these problems can be eliminated by considering the control of fluid velocities
25
during the flow and the accompanying pressure drop through the valve. Excessive fluid velocities cause widely varying local pressures because of the conversion of static pressure head to kinetic energy. This results in extreme turbulence and boundary layer separations as the fluid is forced to make directional changes in the flow path through the valve. The high fluid velocities and locally varying pressure differences cause the following: Uneven forces on moving parts leading to vibrations that cause excessive wear, breaking parts, unscrewing bolts, and noise and flow rate oscillations. Shock waves that lead to screech and high noise-level work environments. Even pipe breakage can occur, as noted in Reference 7, when the noise is in excess of 110 dBA. Cavitation and the rapid erosive wear of metallic surfaces near the location of the bubble collapse. Flashing of the fluid and the accompanying high-velocity liquid drops that erode metal surfaces. Erosion of metallic parts when the fluid has entrained hard solids such as sand, pipe scale, weld slag, and catalyst.
It should be apparent that good long-term control cannot be achieved when all or even one of these problems are present. Guidelines on what velocities are acceptable have been developed over many years of experience in a wide range of applications. These are expressed as a limit to the fluid kinetic energy exiting from the valve trim, as discussed in Reference 8. For the kinetic energy evaluation, the location in the valve that is of greatest concern is just downstream of where the fluid is throttled or controlled. At this location, the flow area is the smallest, and the fluid velocity and kinetic energy are the highest. The parts of the valve responsible for controlling and seating are often located at this point and are therefore subjected to the highest energy fluid. Figure 12-17 shows the throttle area for various kinds of valve trim. For a topguided globe valve, the trim outlet flow area is the annulus area between the plug and seat. In a cage-guided valve, the trim outlet flow area is the exposed area of the windows in the cage. For a multi-path cage, the trim outlet flow area is the total area of all the exposed flow paths. For multi-stage trims, the flow area from stage to stage must not increase too rapidly or else the throttling will take place across the first stages, and the later stages will be ineffective (see Figure 12-17e). Butterfly and ball valves usually meet the presented criteria for kinetic energy. The pressure drop across these types of valves is not large enough to accelerate the flow to a high velocity level. Thus, a much lower value of energy is realized. In a valve, the disk or plug moves to increase or decrease the area through which flow can pass. For a given set of conditions, a fixed area of the trim is open to flow. Under any significant pressure drop conditions, this area will be considerably less than the inlet or outlet area of the valve. As a result, the fluid passing this point will have much higher velocities and kinetic energy levels than in other valve locations. The only way to increase this flow area without increasing the flow rate is to increase the resistance of the throttling flow path. The flow conditions defines how far the valve is open, and the valves trim design (flow path resistance) defines how much flow area exists at the trim outlet. Once this area is defined, the continuity equation can be used to calculate the velocity of the fluid at the outlet of the trim.
26
Figure 12-17. Throttling Exit Area (Ao) Examples for Typical Valve Trim Design.
27
w V o = ------------------M1 o Ao
(12-4)
The fluid density and velocity are used to establish the fluids kinetic energy: 1 o Vo KE = -- ----------2 M2
2
(12-5)
Values for the constants M1, and M2 are shown in Table 12-2.* For gas or steam, the fluid velocity at the trim outlet may be sonic. If it is, the density of the fluid at the trim outlet must be higher than the outlet density (o) in order to pass the given mass flow rate, w. This higher density can be estimated using Equation 12-4 by substituting the fluids sonic velocity, c, for the outlet velocity, Vo, and solving for density. This density and sonic velocity are then used in Equation 12-5 to find the kinetic energy. By including the effect of density in the criteria a single kinetic energy guideline can be defined for all fluids. Thus, the velocity for high-density fluids, such as liquids, would be much lower than for gases. Table 12-3 shows criteria for a valve trims outlet kinetic energy. The valve trim should be selected to keep the kinetic energy below these levels.
Table 12-2. Numerical Constants for Velocity and Kinetic Energy Equations.
Constant M M1 M2 25 1.0 4636.8 1000 w lb/h kg/s -
Units Used in Equations
o
lb/ft3 kg/m3 lb/ft3 kg/m3
Ao in2 m2 -
Vo ft/s m/s ft/s m/s
KE psi kPa
For most conditions, an acceptance criterion of 70 psi (480 kPa) for the trim outlet kinetic energy will lead to a trouble-free valve. In some applications, where the service is intermittent (the valve is closed more than 95% of the time) and the fluid is clean (no cavitation, flashing, or entrained solids), the acceptance criteria can be increased but should never be higher than 150 psi (1030 kPa).
From the general form of the energy equation, potential energy is expressed in terms of a column height of the fluid. Similarly, kinetic energy is expressed in this discussion as a head. The form that is commonly referred to is the velocity head. The units of pressure are traditionally used for the velocity head expression, which can also be converted to a height of the fluid to be consistent with the potential energy term. For objects with a mass the kinetic energy is expressed as: KE = (mass)(V2/2) Similarly, the velocity head is an expression of the kinetic energy of the fluid, although in the form that is relative to a unit volume of the flowing medium. Thus, KE = (mass/volume)(V2/2) = V2/2 The gravitational constant is usually shown in the denominator of the velocity head expression. This is required in order to convert from mass units to force units. The gravitational constant has been included in the constant M2.
28
Table 12-3. Trim Outlet Kinetic Energy Criteria.
Service Conditions
Kinetic Energy Criteria psi kPa 480 275 75 70 40 11
Equivalent Water Velocity ft/s 100 75 40 m/s 30 23 12
Continuous Service, Singlephase Fluids Cavitating and Multi-phase Fluid Outlet Vibration-sensitive System
In flashing service, liquid droplets are carried by their vapor at much higher velocities than would be the case for a single-phase liquid flow. To eliminate the risk of erosion in this situation, the acceptance criteria for flashing or potentially cavitating service should be lowered to 40 psi (275 kPa). The same criteria exists for liquids carrying entrained solids. Special applications may require even more stringent kinetic energy criteria. For example, pressure letdown valves used in pump test loops must be vibrationfree so that a proper evaluation of the pump can be made. These valves are designed with trims that reduce the kinetic energy to less than 11 psi (75 kPa). Gas or steam valves with very low noise requirements may also result in extra-low trim outlet kinetic energy requirements. These kinetic energy criteria are additional sizing considerations that assure the reliable long-term operation of the control valve. A decision to ignore these rules may result in lower procurement cost but cause high operation and maintenance costs. In some cases, the ability of the valve to perform its control function may be jeopardized. References 9 and 10 present examples in which feedwater regulators could not perform the intended control function. Control was achieved when a trim was used that halved the fluid exit velocity of the originally installed cage trim and local smoothing of the flow path was incorporated. The calculation of trim velocities and kinetic energy will require a knowledge of the cross-sectional area for the flow channel of interest as well as the number of ports. Approximate calculations can be made with some information about the valve design, but it is best to work with the manufacturer to obtain more accurate results. Another approach would be to use Equations 12-5 and 12-4 to calculate the Vo and Ao, respectively, by fixing the kinetic energy at the Table 12-3 values. The area calculated is the area needed, at the valve travel, that passes the flow input to Equation 12-4. Then Equation 12-5 is used to calculate the fluid density so it can be used in Equation 12-4 with the velocity at sonic. This calculated area can then be compared with the flow area for the valve being proposed by the valve supplier. The actual area should always be greater than the calculated value in order to meet the kinetic energy selection criteria. A larger area would also be needed if cavitation or noise were major considerations in the application.
Remember to check the calculated velocity to be sure that it does not exceed the sonic velocity of the fluid. If the velocity is greater than sonic it must be set equal to sonic velocity.
Calibrating It Right
A third problem involves achieving full-seat loading to maintain tight shutoff when the valve is closed. A prevalent practice is to calibrate the valve, or the bench set when a positioner is not used, so that the closure member (e.g., plug, diaphragm, disk) is just positioned at the seat instead of also assuring that the closure member is fully loaded against its seat.
29
Unfortunately, it is not obvious to the inexperienced engineer that there must be a good seat load between the two mating seating surfaces. If there is not a sufficient load, then fluid leakage will rapidly erode the surfaces, the erosion time depending upon the valve pressure drop and fluid. If proper seat loads are not maintained, then it is impossible to maintain design leakage rates.
The situation is analogous to the manual control of the home water faucet. If the faucet is turned to just stop the water flow, it will not be long before it will start to drip. As a result of this experience, we subconsciously apply an extra torque to load the faucet plug against its seat to assure there will be no distracting or wasteful leakage. Full loading of a push-down-to-close sliding stem valve is only assured when the conditions within the actuator are as shown in Table 12-4. Single-acting actuators are most frequently a diaphragm design. With this design, the spring is either reducing the seating load or providing the entire seating force. The doubleacting actuator is typically a piston design. With the piston design, the supply pressure is not limited as with the diaphragm design, and full-supply pressure is available to achieve higher seating forces. The higher pressures in the piston design have the additional benefit of increased stiffness and better control resolution.
Table 12-4. Actuator Seating Load Source.
Actuator Type Single-acting, Spring to Open
Conditions for Full Seat Load Supply Pressure on Top of Diaphragm or Piston Pushing toward the Seat Spring Pushing away from Seat Spring Pushing toward Seat 0 psig under Diaphragm or Piston Supply Pressure on Top of Piston Pushing toward the Seat 0 psig under Piston
Single-acting, Spring to Close Double-acting
As Baumann pointed out in his article on power signals (Reference 11), many designers think of the 4-to-20-mA signal as an information signal instead of a power signal. This signal, in the case of a control valve, is not only dictating the required position of the closure member, it also drives the operating power to position and seat the valve. When the valve is calibrated to be just closed at exactly 4 mA, the extra power built into the valve design to assure seating load is not applied. This seating load would only be applied when the control signal drops below 4 mA, a signal that is usually not built into the control system. Thus, the goal is to calibrate so that the valve positions and then seats with full loading when it is closed.
CALIBRATION WITH POSITIONER
For small signal changes on double-acting actuators, positioners are typically designed to maintain two-thirds to three-quarters of the supply pressure on the side of the piston in the direction of stem movement. In single-acting actuators, this bias pressure is replaced by a spring. In either case, the only way to ensure that maximum actuator load is provided to the valve seat when the valve is closed is through careful calibration.
30
To ensure that the actuator fully loads the valve closure member against its seat, the calibrator has to purposely create a sufficient error between the valve position feedback to the positioner and the positioner input signal in order to cause the positioner to try to correct its position in the closing direction. To accomplish this, the positioner is typically calibrated to have the valve closure member reach the seat within the signal ranges shown in Table 12-5. The difference between these values and the true endpoint values of 3.0 psi, 15.0 psi (0.02 MPa, 0.10 MPa), 4.0 mA, or 20.0 mA is sufficient to create the error required within the positioner to cause it to provide maximum load in the seating direction.
Table 12-5. Calibration for Seat Load.
Positioner Signal Range 3.0 to 15.0 psi (0.02 to 0.1 MPa)
Valve Action on Increasing Signal Opens Closes
Positioner Signal when Closure Member Just Contacts Seat >3.1 but <3.6 psi (>0.021 but <0.025 MPa) >14.4 but <14.9 psi (>0.100 but <0.103 MPa) >4.2 but <4.8 mA >19.2 but <19.8 mA
4.0 to 20.0 mA
Opens Closes
CALIBRATION WITHOUT POSITIONERS
The problem of sufficient seat loading is particularly acute in throttling control valves that dont use a positioner. In these cases, the signal pressure to the actuator is also the operating pressure for the actuator. Even with the older standard signal ranges of 3 to 27 or 6 to 30 psi (0.02 to 0.19 or 0.04 to 0.21 MPa), resulting seat loads are considerably less than those available using a positioner, where up to full-supply pressure can be utilized to provide actuator thrust. Valve throttling applications without positioner control utilize a spring to oppose the operating pressure. They are generally of the spring and diaphragm style because of the low pressures involved and have large diaphragm areas relative to piston actuators. The user has very little flexibility in altering the calibration built into the factory selection of the actuator and spring. The spring force can be adjusted to increase the seating force, but if changed too much the valve may not fully open. Also, adjusting the valve when the valve is not pressurized may result in incorrect loads. This is because each valve design has its unique hydraulic and frictional forces upon which the actuator was selected and of which the user may not be aware. The calibration of the actuator takes into consideration two factors. The first is the actual signal pressure range. For example, if the designer counted on a 3-to-15 psi (0.02 to 0.10 MPa) signal to the actuator being able to actually range from 0 to 18 psi (0 to 0.12 MPa), the additional 3 psi (0.02 MPa) at the closed end of the stroke was assumed to provide seat load. If the valve is then calibrated in service to have a minimum of 3 psi (0.02 MPa) and a maximum of 15 psi (0.10 MPa) applied to the diaphragm, seat load will be reduced by 3 psi (0.02 MPa) times the diaphragm area, which is commonly several hundred pounds (kilograms). Reference 12 addresses this situation in the form of a standard that calls for an extended signal range. The second consideration in calibrating the actuator are the forces that the actuator will have to work with and against. These forces are hydraulic unbalance,
31
friction, and seat load requirements. For example, if an actuator is intended by the user to be operated with a true 3-to-15 psi (0.02 to 0.10 MPa) signal, with no under- or over-ranging, the designer will have to build these loads, including seat load, into the actuator spring force or diaphragm force. This is commonly accomplished by selecting hardware that satisfies a bench set pressure range that may be quite different from the operating signal range of 3 to 15 psi (0.02 to 0.10 MPa). This bench set pressure range is intended to allow the actuator to compensate for the forces just noted when the valve is in service so that it will operate properly (e.g., provide the design seat load) within a 3-to-15 psi (0.02 to 0.10 MPa) signal. To compensate for the requirement that there be an infinite number of springs to fit individual situations, spring adjusting screws are normally incorporated into the actuator design to achieve accurate spring forces. Thus, because of the relatively small available forces, when the technician sizes and selects actuators for throttling applications without a positioner it is critical that the technician calibrating the valve in the field know the operating pressure assumptions that were used to select the actuator and to calibrate the actuator accordingly. It should also be apparent that if the valve is moved to another application or fluid pressure conditions change in the existing installation, the technician will probably require new springs and a new calibration, or bench set, to assure the highest seating load.
MORE CALIBRATION COMMENTS
The calibration methods described in the preceding section provide excellent performance in almost every control system. In a control loop, the signal to the valve is created from a measurement in the process. The measurement is compared against a desired value, and a correction signal is then fed back to the valve. The valve responds to this signal to correct by changing position. Thus, in this basic function of the control valve there is no need for the travel stroke to match the input signal. So the calibration described in the last section will assure the design seat load when the valve is closed. The calibration described in the last section will provide a sufficient indication of valve position using the input signal to the valve. The position indication seldom must be highly accurate because it only provides relative feedback on the operation of the valve. If highly accurate position indication is required, a separate position transmitter can be added to provide this information. In the event that the calibration methods described in the last section are not acceptable, there are electrical and pneumatic devices that can then be added to the control schematic to cause a fully loaded seat interface when the signal reaches the closed position. These devices could add to the initial cost of the control valve, especially for small valves. However, this cost is much less than the maintenance cost associated with a valve that seals poorly. An example would be for a 3-to-15 psi (0.02 to 0.10 MPa) pneumatic signal, increasing signal to close, to add a high-gain relay so that at 15 psi (0.10 MPa) the power to the actuator would be the maximum supply pressure. But even in this case it would be advisable to set the relay at some signal level below 15 psi (0.10 MPa) to allow a margin for calibration or signal error. The control schematic for this example is shown in Figure 12-18 in which the full seating force is applied at 13.8 psi (0.095 MPa), 1.2 psi (0.008 MPa) below the full closed signal. Some manufacturers current-to-pneumatic (I/P) converters have a snap-shut signal wherein the 3-to-15 psi (0.02 to 0.10 MPa) output signal drops to zero as the 4-to-20 mA command drops below 4.2 mA.
32
Figure 12-18. Control Schematic to Modulate between 10% and 100% Open.
The schematic shown in Figure 12-18 is for a 3-to-15 psi (0.02-0.10 MPa) modulation signal with increasing signal to close the valve, using a double-acting piston-type actuator. The actuator is fitted with a spring for fail-open on loss of the signal. A regulator between the supply and snap-acting relay is set at 20 psi (0.138 MPa). The snap-acting relay is set at 13.8 psi (0.095 MPa), which represents the signal for the 10% open position. When the signal is modulating between 3 and 13.8 psi (0.02 and 0.095 MPa), the high selector and snap-acting relays are in the positions shown. Thus, the positioner receives the modulating signal and moves the valve actuator in accordance with the demanded position. When the signal exceeds 13.8 psi (0.095 MPa), the snap-acting relay changes position from venting to straight-through. The 20 psi (0.138 MPa) signal from the regulator then changes the high selector relay position and feeds this pressure to the positioner. The 20 psi (0.138 MPa) signal into the positioner causes the valve to go fully closed with the pressure above the piston at the supply pressure and the pressure below the piston at atmospheric. Thus, the maximum actuator force is directed to sealing the valve closure member. Other schematics can be created to assure full seating load for decreasing signal to close and for single-acting actuators.
Operating Close to the Seat

An application problem that is frequently encountered involves the extended duration of operating the valve with its closure member close to the valve seat (approximately 5% or less). Such operation causes seat deterioration. Under this low-flow condition, the flow path is the small opening between the closure
33
member and seating surfaces. The fluid velocities can be extreme, resulting in erosion, noise, vibration, and, for liquids, cavitation damage. In some cases, erosion is so severe that the damage to the valve at low lift increases flow area, and the valve cannot control a needed minimum flow. To eliminate this problem, the minimum controllable flow must be determined from the process data and valve designs selected that have the needed rangeability for the application. The control schematic shown in Figure 12-18 can also be used to avoid operation with the closure member close to the seat. When the valve stroke reaches a preset position, the snap-acting relay changes positioner signal causing the valve to fully close and apply the full seating force.
Poor Assembly
Another contributor to valve problems is poor assembly in the field. It is critical that the manufacturers assembly instructions be followed to assure that trim parts are aligned well and, as discussed earlier, calibrated properly. The mechanical assembly of a valve may seem quite straightforward; however, the assembler may not be aware of the special design features included to assure a good functional control valve and its dependence upon proper assembly procedures. The issue of the alignment of parts is particularly critical in many applications. If alignment is not maintained within tolerances then the seating line or contact area between the closure member and the seat ring may not be as anticipated. Significant leakage will occur with the long-term erosion of these sealing surfaces. The erosion of parts also reduces the valves ability to provide good control. Another adverse effect of poor alignment is that the stem may be scratched and galled because of the close tolerances within the spacers and guides used in the packing box. If the stem becomes scored then the packing will have a dramatically reduced life as packing material is removed from the packing box during modulation. In general, the higher the fluid pressures, the tighter the spacer tolerances. The tight tolerances are needed to minimize the extrusion of the packing material from the box. Poor alignment results in high friction for the moving parts. This is not always apparent at the time of assembly. Friction can become a major problem during the operation of the valve as the fluid system reacts to the excessive movement of the valve disk or plug in response to the higher friction. This excessive movement will cause the valve to continually search for the position that corresponds to the system control point. Poor alignment generally results from improper torquing of the packing box and the valve bonnet bolting. The latter is the body/bonnet closure joint that seals the internal valve pressure from the ambient. Both of these bolted joints must be tightened by bringing the mating components together while maintaining them as close to their final and fitted relationship as possible. While parts will have tightfitting location registers, improper torquing will push the mating parts to the extreme of the tolerances, thus reducing design margins. Correct tightening of the bolts is achieved by using a crisscross tightening pattern and working up to the final torque value in small steps. If too large a torquing step is used, the parts become skewed and remain so through the final torque sequence. The skewed posture loads the internal mating parts non-uniformly and causes a misalignment. Tightening the packing box is a critical operation in the assembly of most valves. Regardless of the packing design supplied by the valve manufacturer, the instructions should be rigidly followed to assure alignment and to avoid excessive packing friction. Tightening is most critical in the high-temperature graphite
34
packing designs. This procedure generally involves a sequence of torquing levels, with the valve stem moved three or more times at each torque level until the maximum sealability torque is achieved. The torque is then backed off to a maximum operability torque level for optimum control. The torque levels should then be verified periodically during the start-up and operation of the valve to assure the optimum control performance. The maximum sealability and operability levels will vary depending upon the pressure rating of the valves. Do not reuse old seals and gaskets when reassembling a control valve. The seals and associated springs will take a permanent set in the application, whether elastomeric or graphite. They may not have the strength to initiate a proper seal on reuse. Elastomeric materials may be near the end of their lives and fail because of brittleness or poor ductility. Small nicks in the seals may also contribute to leak initiation and the subsequent erosion of the seal in the application. The replacement seals should be of the original design because special compositions or winding processes may have been specified by the manufacturer to assure proper function and long life for the specific application. The small cost of using a new and correct set of these soft goods is minor in comparison to the cost of excessive leakage or poor control function. In some cases, the reuse of seals and gaskets can even result in a shutdown of the process. Depending upon the design, there are times when it is necessary to connect the actuation system to the valve in a way that assures that an actuator stop does not limit the seating load. In some cases, this could even hold the closure member off of the sealing surface. The assembler must make sure that the seating load is not limited. The manufacturer must provide instructions that will assure that this attachment process can be done correctly. Following the proper assembly procedure assures that the design features of the valve are not compromised or negated. Thus, the reliability and performance of the valve will fulfill expectations.
To assure that the final torque level results in a properly bolted joint all mating parts including the studs and nuts, must be well lubricated. It is not uncommon to lose between 20 and 40% of the torque value to friction between these parts.
Actuator Supports
Occasionally it is necessary to provide added support for a valve actuator. This may be the case if the actuator is quite heavy, such as for many electrical and electro-hydraulic designs, or when the valve is installed in a position in which the actuator is off the vertical centerline. In these cases, it is very critical that the support be flexible enough to allow for the thermal growth of the piping system and that the installation does not impose large lateral forces. Such forces could cause higher friction on the valves moving parts. The ideal support system is simply a pulley with a dead-weight counterbalancing the weight of the actuator, as shown in Figure 12-19. Other support systems require enough design effort to be assured that the actuator is not constrained or deflections imposed that add undesirable side loads on the actuator. One criterion that should never be used when designing a support system is to attempt to limit vibration and in particular to limit a pipeline vibration problem. The former almost always results in the fatigue failure of a restrained component and the latter will result in short-term destruction of the valve-actuator assembly. The vibration issue must be resolved by using damping devices mounted directly on the pipe or by eliminating the root cause of the vibration.
Trying to stop pipeline vibration by holding the actuator is analogous to trying to stop a train with your car.
35
Figure 12-19. Supporting a Heavy Actuator.
System Debris and Flushing

During new construction or the major repair of existing equipment there is always debris generated by the processes of cutting, machining, grinding, welding, erection, and installation. This occurs in spite of efforts to make sure the fluid side of the system is clear of all trash. Thus, it is critical that the system be flushed and the internals of valves and other critical equipment be protected by removing debris or by providing upstream strainers during initial operation to minimize damage caused by the debris. The effort to assure that a system is cleaned by flushing it requires significant planning; however, the investment pays itself back many times by extending the life of the many components in the system. To assist in this planning process, the reader is referred to References 13 through 16. A measure of the significance of this issue is provided in Reference 15, which noted that nine tons of iron oxide were removed from the power plant piping. The Installation and Location section of Chapter 18 of Control Valves includes other examples of debris problems. The influence of debris in a system can have a very detrimental impact on the control valves. In the references cited in this chapter it is recommended that the trim be removed before the flushing operations and that some sacrificial trim be substituted into one or more of the valves. In some cases, the entire valve may be removed and a pipe spool bolted into the line. If the debris is not flushed from the system, the valve trim will be damaged, regardless of the type of valve design involved. In all valve designs, there are parts that are tightly toleranced, particularly the tolerances that control the gaps between moving parts. It is in these small gaps that debris can become lodged, severely scratching the mating surfaces when the valve is modulated. The moving surfaces, such as those on the plug, also have an associated seal. Modulating the scratched surface through these
36
References
seals causes the seal to fail, and then additional damage follows as a result of erosion and cavitation. Vibration can also occur if the seal has been damaged to the point that it does not provide enough support or transient hydraulic loading occurs. Almost everything that can be carried by construction personnel has been found in plant piping systems during initial operations. Most of these items, such as wrenches and other tools, will get lodged in a turn or crevice and result in only minor damage. The most damaging and populous debris is the pipe scale that has formed during construction as well as fabrication debris, such as weld slag, weld rod, and weld splatter. The pipe scale is thin and brittle so it breaks into a fine dust. This dust will usually move through the system, even flowing through the smaller passages of tortuous path trim valves. The influence of this pipe scale will show up on valve parts as eroded surfaces, which can have a very rough finish and can be mistaken for cavitation damage. Weld slag is one of the most damaging types of debris in a piping system. Its glassy structure makes it brittle so that it breaks up and eventually moves through the system. However, these particles tend to be much larger and tougher than the pipe scale. Weld splatter are small beads of weld or base metal that cling to the pipe wall until they are loosened by the flowing medium or thermal cycling caused by the start-up and shutdown operations. The beads are very hard as a result of the thermal quenching that is inherent in the welding process. These beads of metal then get carried downstream and lodge between close-fitting parts, causing friction and galling. The influence of the weld slag and splatter is seen as the gouges and deep, sharp-edged scratches that are caused by the particles getting wedged between moving parts. Even hardened and overlaid surfaces are easily scratched by this material, frequently to the point where parts must be replaced to restore the valves capabilities. Weld rod in a system will get caught between moving parts and cause high friction that in some cases cannot be overcome by the actuator. The flux on the rod will break off and pass through as a dust, and only the ductile wire will be found when the valve is disassembled. Moving parts will be scratched, but the grooves generally have rounded edges, as opposed to the sharp-edged grooves caused by slag and weld splatter.
References
1. Driskell, L., Control Valve Selection and Sizing, Unit 20 (Research Triangle Park, N.C.: ISA, 1983), pp. 391-414. 2. Hutchison, J. W., editor, ISA Handbook of Control Valves, 2nd ed. (Research Triangle Park, N.C.: ISA, 1976), chapters 12 and 13. 3. ISA Form S20.50 (taken from ISA-S20-1981), Specification Forms for Process Measurement and Control Instruments: Primary Elements and Control Valves (Research Triangle Park, N.C.: ISA, 1981). 4. Application Bulletin: AB 401 (Worcester, Mass.: Neles-Jamesbury, 1992). 5. Liptak, B. G., Control Valves in Optimized Systems, Chemical Engineering, September 5, 1983, pp. 104-10. 6. Tullis, J. Paul, Hydraulics of Pipelines: Pumps, Valves, Cavitation, Transients (New York: John Wiley & Sons, 1989), pp. 153-57.
37
7. Fagerlund, A. C., Recommended Maximum Valve Noise Levels, Paper No. 86-2862 (Research Triangle Park, N.C.: ISA, 1986). 8. Miller, H. L., and L. R. Stratton, Fluid Kinetic Energy as a Selection Criteria for Control Valves, American Society of Mechanical Engineers (ASME), Fluids Engineering Division, summer meeting, June 22-26, 1997. 9. Parker, A. Q., K. J. Avery, H. L. Miller, and C. G. Sterud, Retrofitting Feedwater Regulating Valves at North Anna, Power Engineering 98, No. 11 (1994). 10. Donavin, P. R., J. J. Patel, H. L. Miller, and C. G. Sterud, Feedpump Discharge Line Vibration Solved at Byron Station, Unit 2, American Power Conference, April 1993. 11. Baumann, H. D., The Need for Pneumatic Power Signals for Control Valves, InTech, January 1988, pp. 37-41. 12. FCI-87-2-1990 (R1994), Power Signal Standard for Spring-Diaphragm Actuated Control Valves (Cleveland, Ohio: Fluid Controls Institute, Inc., 1994). 13. Powerplant Piping Systems, Power 129 (May 1985): S28-S30. 14. Heneage, J. F., How to Steam-Flush Nuclear Main-Steam Lines before Service, Power 128 (October 1984): 99-100. 15. Booth, C. R., C. E. Guitierrez, C. A. Moore, and K. C. McCubbins, Prestartup Chemical Cleaning of Nuclear Plant Piping, Power 127 (August 1983): 51-53. 16. American National Standard Institute, Cleaning of Fluid Systems and Associated Components during Construction Phase of Nuclear Power Plants, ANSI N45.2.2, American Society of Mechanical Engineers (ASME), New York, 1973.
About the Author

Herbert L. Miller is Vice President of Operations for CCI (formerly Control Components Inc.). He holds a Bachelors of Science and a Masters of Science in Mechanical Engineering from Ohio Northern University and Northwestern University, respectively. He has authored over 35 articles on heat transfer and fluid flow technology. Technical articles on control valves have included many applications dealing with corrections to improperly applied valve designs. The applications have been concentrated in the severe and tough service installations of the electrical power generation and oil and gas handling industries. He has been a senior member of the ISA and its technical committee on control valves since 1984 and has been active on the noise and cavitation subcommittees. He has been a member of the American Society of Mechanical Engineers since 1963 with leadership in Sectional and Regional Administrative roles as well as technical committees dealing with materials and qualification of mechanical and electrical equipment in nuclear power plants.
Index
A
actuator supports, 35 assembly, 34
C
calibration, 30
D
debris, 36
F
feed regulation, 2 fluid velocity, 25 flushing, 36
L
letdown, 6, 8
P
parallel valves in, 14 pressure drop, 25 process control, 2
R
recirculation, 10
S
series valves in, 23
39

Aplicaciones Valvulas de Control

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Aplicaciones Valvulas de Control

Hochgeladen von

Copyright:

Verfügbare Formate

Volume EMC 21.

Control Valve Applications

Application Categories Valves in Parallel and Series Frequent Application Problems

Taken from the Practical Guide Series book: Control Valves

ISA is the international society for measurement and control

12 Control Valve Applications

Control Valve Applications

The Application Categories

Flow control Level control Pressure control

The Application Categories

Pressure reduction Regulator (flow) Throttle

Figure 12-1. Process Control/Feed Regulation Application.

Control Valve Applications

The Application Categories

Figure 12-2. Cv/Flow versus Travel.

Control Valve Applications

The Application Categories

Figure 12-3. Continuous and Intermittent Letdown Application.

Control Valve Applications

The attributes for the intermittent letdown application are as follows:

The Application Categories

Control Valve Applications

The Application Categories

Figure 12-4. Recirculation/Recycle Application.

Control Valve Applications

Secondary considerations include the following:

The Application Categories

Figure 12-5. Valve Position versus Time Opening.

Control Valve Applications

Figure 12-6. Parallel Valve Application.

Figure 12-7. VRT Trim Valve. (Courtesy of Masoneilan)

Control Valve Applications

INSERT PHOTO HERE

INSERT PHOTO HERE

Figure 12-9. Cascade Trim Valve. (Courtesy of Copes Vulcan, Inc.)

Control Valve Applications

Figure 12-10. Mark One Globe Valve. (Courtesy of Valtek International)

Figure 12-11. Q-Ball Valve Schematic. (Courtesy of Neles-Jamesbury Inc.)

INSERT PHOTO HERE

Figure 12-12. Drag Trim Valve. (Courtesy of Control Components Inc.)

Control Valve Applications

Figure 12-13. Three-way Valve Substitute for Parallel Valves.

Control Valve Applications

where Cv,o Cv,i = overall Cv = Cv of the ith valve

Figure 12-14. Sequencing Two Parallel Valves of = % Trim.

Flow Direction ----------Open Close ------Open Close ----------------

Control Valve Applications

Figure 12-15. Single-Order Response Time.

Figure 12-16. Series Valve Applications.

Frequent Application Problems

Frequent Application Problems

Controlling Pressure Drop

Excessive Fluid Velocities

Control Valve Applications

Frequent Application Problems

Control Valve Applications

Constant M M1 M2 25 1.0 4636.8 1000 w lb/h kg/s -

Units Used in Equations

Vo ft/s m/s ft/s m/s

Frequent Application Problems

Table 12-3. Trim Outlet Kinetic Energy Criteria.

Kinetic Energy Criteria psi kPa 480 275 75 70 40 11

Equivalent Water Velocity ft/s 100 75 40 m/s 30 23 12