aplicaciones valvulas de control

7/29/2019 Aplicaciones Valvulas de Control

1/45

ISA is the international society for measurement and controlISA is the international society for measurement and control

Volume EMC 21.01

Control Valve ApplicationsHerbert L. Miller

Application Categories

Valves in Parallel and Series

Frequent Application Problems

Taken from the Practical Guide Series book: Control Valves


2/45

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


3/45

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 havenot 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

To order: Internet: www.isa.org

Phone: 919/549-8411

Fax: 919/549-8288

Email: [email protected]


4/45


5/45

v

Table of Contents

Chapter 12 CONTROL VALVE APPLICATIONS 1by Herbert L. Miller

The Application Categories, 2

Valves in Parallel, 14

Valves in Series, 23

Frequent Application Problems, 25

References, 37

INDEX 39


6/45


7/45

1

12

Control Valve

Applications

The purpose of this chapter is to present the information that will help processengineers 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 bookControl Valves; selection among materials was discussed in Chapter

11 ofControl Valves. There are exceptions within every category because of the

needs unique to a specific process within an industry. An example of this would bea 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


8/45

Control Valve Applications

2

application. Chapter 6 ofControl Valves discusses sizing; Chapter 22 ofControl

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. Thisvalve 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:

Flow control Level control Pressure control


9/45


3

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 becontinuously 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


10/45


4

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 alsoassures 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 ofControl Valves) and

excessive noise (see Chapter 8 ofControl 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 thisvalve and that is accuracy of feed control. In this application, other

attributes are usually secondary. These are as follows:


11/45


5

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 productin 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 pressuredifferential 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.


12/45


6

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 bythe 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, soselecting 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. Anexample 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.

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.


13/45


7

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, whichrequire 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. Areasonable 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.


14/45


8

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:


15/45


9

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 linearcharacterization 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-safedirection, which is usually fail-open.

4. Stroke speed. The dual functions of these applications dictate the speed ofoperation. 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 forthe 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 mustperform the role of modulating control and then become a block valve when shut. For many valves that areused 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 maynot be bothersome when the valve is new. However, since these valves are normally shut, fluid erosion ofthe 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 meetingthe 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 themaximum differential operating pressure may be used. The permissible leakage under this standard shall beless than 10 cc/h of liquid per inch of diameter of the nominal valve size.


16/45


10

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.

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, orsystem-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 ismeasurably reduced.

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 butassuring 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.


17/45


11

Because most fluidsare erosive, particularlyliquids, a small leak

through the valve willresult in the rapiddeterioration of theseating surfaces.

Figure 12-4. Recirculation/Recycle Application.

This valve will see the highest pressure drop in the entire plant when youbypass 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

In addition to the comments about leakage and control attributes discussedunder 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.


18/45


12

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 unbalancedplug designs with large actuators.

Anti-cavitation/low noise trim

Pipe vibration elimination

Secondary considerations include the following:

Flow characteristics. Usually linear for pumps and characterized forcompressors

Failure mode. Normally open

Stroke speed. 2 to 5 seconds for compressible; up to 25 seconds forliquids

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


19/45


13

opened to 95% of full stroke in only 1.7 seconds. The valve and actuator sizes and

conditions were as follows:

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.

Fluid Mixed Refrigerant

Actuator size 113 in2 730 mm2

Valve stroke 24 in 610 mm

Plug weight 1320 lb 600 kg

Inlet pressure 190 psi 1.31 MPa

Outlet pressure 45 psi 0.31 MPa

Flow rate 1,620,000 lb/hr 735 t/h

Air supply pressure 100 psi 0.69 MPa


20/45


14

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 exampleis 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


21/45

Valves in Parallel

15

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 controlis 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. VRTTrim Valve. (Courtesy of Masoneilan)


22/45


16

Figure 12-8. V-Line Series Noise Attenuator Ball. (Courtesy of Fisher Controls International Inc.)

INSERT PHOTO HERE


23/45

Valves in Parallel

17

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

INSERT PHOTO HERE


24/45


18

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

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


25/45

Valves in Parallel

19

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

INSERT PHOTO HERE


26/45


20

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. Thisincreases 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.


27/45

Valves in Parallel

21

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 andlong-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.

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 Cvvalues of each valve in parallel. That is,

Cv,o = Cv,1 + Cv,2 + + Cv,i (12-1)

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 thefull 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.


28/45


22

where

Cv,o = overall Cv

Cv,i = 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 thecavitation 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.


29/45

Valves in Series

23

Valves in Series

It is always desirable to

use one valve becauseof the controlinstabilities that arise inthese seriesconfigurations.

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 inseries 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.

Control instabilities are beyond the purpose of this discussion, but they canoccur 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 (12-2)

where

L = distance between valvesU = average fluid velocity

A = piping flow area

w = mass flow rate

= fluid density

Valve Type Flow Direction p1/p2 Limit* p/p1 Limit

Multi-stage, Multi-path ------ No limit 1.0

Multi-stage, Single-path ------ 5.5 .82

Single-seat Globe Open 3.3 .70

Close 2.3 .57

Double-seat Globe ------- 3.3 .70

Angle-body Cage Open 3.0 .67

Close 2.3 .57

Pinch ------ 2.9 .66

Butterfly, 60% Open ------ 1.5 .33

Reduced Ball, 80% ------ 1.2 .17

*p1/p2 = 1/(1 p/p1)


30/45


24

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.


31/45


25

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 diffusercan 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 C v less than

that required by the process.

Frequent Application ProblemsIt 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. Inthis 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


32/45


26

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 excessivewear, breaking parts, unscrewing bolts, and noise and flow rateoscillations.

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 isin excess of 110 dBA.

Cavitation and the rapid erosive wear of metallic surfaces near thelocation of the bubble collapse.

Flashing of the fluid and the accompanying high-velocity liquid dropsthat erode metal surfaces.

Erosion of metallic parts when the fluid has entrained hard solids such assand, 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 top-

guided 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 opento 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.


33/45


27

Figure 12-17. Throttling Exit Area (Ao) Examples for Typical Valve Trim Design.


34/45


28

(12-4)

The fluid density and velocity are used to establish the fluids kinetic energy:

(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) inorder 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.

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 ofthe fluid. Similarly, kinetic energy is expressed in this discussion as a head. The form that is commonlyreferred 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 inthe constant M2.

Constant Units Used in Equations

M w o Ao Vo KE

M1 25 lb/h lb/ft3 in2 ft/s -

1.0 kg/s kg/m3 m2 m/s -

M2 4636.8 - lb/ft3 - ft/s psi

1000 - kg/m3 - m/s kPa

Vow

M1oAo-------------------=

KE1

2---oVo

2

M2------------=


35/45


29

Table 12-3. Trim Outlet Kinetic Energy Criteria.

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 vibration-free 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.

Remember to check thecalculated velocity tobe sure that it does notexceed the sonicvelocity of the fluid. Ifthe velocity is greater

than sonic it must beset equal to sonicvelocity.

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 andAo, 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.

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.

Service ConditionsKinetic Energy

CriteriaEquivalent Water

Velocity

psi kPa ft/s m/s

Continuous Service, Single-

phase Fluids

70 480 100 30

Cavitating and Multi-phaseFluid Outlet

40 275 75 23

Vibration-sensitive System 11 75 40 12


36/45


30

Unfortunately, it is not obvious to the inexperienced engineer that there must be agood seat load between the two mating seating surfaces. If there is not a sufficientload, 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 isimpossible 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 startto 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 double-

acting actuator is typically a piston design. With the piston design, the supplypressure 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.

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.

Actuator Type Conditions for Full Seat Load

Single-acting, Spring to Open Supply Pressure on Top of Diaphragmor Piston Pushing toward the Seat

Spring Pushing away from Seat

Single-acting, Spring to Close Spring Pushing toward Seat

0 psig under Diaphragm or Piston

Double-acting Supply Pressure on Top of Piston

Pushing toward the Seat0 psig under Piston


37/45


31

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.

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 diaphragmstyle 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 to18 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,

Positioner SignalRange

Valve Action onIncreasing Signal

Positioner Signal when ClosureMember Just Contacts Seat

3.0 to 15.0 psi(0.02 to 0.1 MPa)

Opens >3.1 but 0.021 but 14.4 but 0.100 but 4.2 but 19.2 but


38/45


32

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 tocompensate 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.


39/45


33

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 thesignal. 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


40/45


34

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 avoidoperation with the closure member close to the seat. When the valve stroke

reaches a preset position, the snap-acting relay changes positioner signal causingthe 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 theoperation 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 tight-

fitting location registers, improper torquing will push the mating parts to theextreme 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


41/45


35

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.

To assure that the finaltorque level results in aproperly bolted joint allmating parts includingthe studs and nuts,must be well lubricated.It is not uncommon tolose between 20 and40% of the torque valueto friction betweenthese parts.

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.

Actuator Supports

Occasionally it is necessary to provide added support for a valve actuator. Thismay 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 theactuator 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.

Trying to stop pipelinevibration by holding theactuator is analogousto trying to stop a trainwith your car.

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.


42/45


36

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 beflushed 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 ofControl 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 thetrim be removed before the flushing operations and that some sacrificial trim be

subs

aplicaciones valvulas de control

Documents