Control Valve Selection

All control systems manipulate some end device. This end device has some impact, either
direct or indirect, on the value of the controlled variable. For example:
At a particular point in a duct system, the static pressure is held constant. A control
system may do so by sending a signal to a variable speed drive, which in turn varies the
rotational speed of the fan. As fan speed varies, so does the static pressure in the duct
system.
In a catalyst regeneration system, a control system must maintain the temperature of
the hot air leaving the regenerator. This may be accomplished by controlling the
volume flow rate of the air/fuel mixture entering a boiler. The boiler will increase or
decrease the output temperature of hot air to the regenerator. This, in turn, impacts
the temperature of the hot air leaving the regenerator.
In a shell and tube heat exchanger, the water temperature leaving the exchanger on
the tube side may be controlled by manipulating the flow rate of the hot fluid entering
the shell side. This flow rate is manipulated with a liquid control valve.
There are a number of other end devices used for any number of purposes. These
include air dampers, SCRs (Silicon Controlled Rectifiers), SSRs (Solid State Relays), etc.
When dealing with fluid flow and heat transfer systems, the end device in question is a
control valve. Although control valves come in a wide range of configurations, the
parameters used to select a valve are essentially the same. These parameters include:
Static Pressure Rating
Close-off pressure
Valve Action (Normally Open vs. Normally Closed)
Valve Coefficient (C
v
)
Valve Authority
Valve Rangeability
Valve Cavitation Coefficient
Static Pressure Rating
The static pressure rating of a valve is simply the maximum static pressure the valve is
designed to handle when placed in service. The rating is stated in manufacturers'
literature. Obviously, the rating of the valve must be greater than the maximum static
pressure attained in the system as determined when the system is designed. Although
proper selection is important for safe and reliable operation of the system, it does not
impact the controllability of the system.
Close-off Pressure
Close-off pressure is how much differential pressure the valve / actuator combination can
overcome in order to seat the valve and stop flow completely. As such, this criterion has
a direct impact on sizing the valve actuator. The determination of the necessary close-off
pressure is a function of the hydraulic design of the system. Once the designer
responsible for the design determines this requirement, the control designer can use this
information to properly select the actuator. Actuator sizing can become quite involved
and is beyond the scope of this course. Fortunately, actuator selection is typically done
by the valve manufacturer.
Typically, actuators are either electronic or pneumatic. In some industries, actuators may
also be hydraulic or vacuum operated. Most industrial and commercial processes use
pneumatic and electronic actuators, so we will focus our discussion on these. Pneumatic
actuators are very popular due to their relatively low cost, high power output, and
reliability. They are available in a myriad of pressure ranges and are easy to sequence.
Recently, electronic actuators have become quite popular. Technological advances have
reduced their cost, increased their power output, and improved their reliability. The
major difference between pneumatic and electronic actuators is the speed of operation.
The effect of speed of operation is discussed under the topic of system time lags and
dead time.
Valve Action
Valve action defines whether the valve is full open when the control signal to the actuator
is low, or if is full closed when the control signal to the actuator is low. The selection of a
normally closed vs. a normally open valve is based on a desired fail-safe condition. Fail-
safe refers to that condition in which you wish the valve to be in the case of a failure of
the control system. For example, assume an office suite is heated by air flowing over a
hot water coil. If the control system fails, it would probably be desirable for the control
valve to fail in the open position so as to provide full heat to the space. Normally open is
the typical fail-safe condition selected for most heating valves. However, assume the space
houses a mainframe computer system and any heating required is accomplished with a
reheat coil in a VAV box. Mainframes tend to be somewhat sensitive to high ambient
temperatures. In such a case, we may wish the heating valve to fail in a full closed
position should the control system fail. Although the space may overcool, this is
preferable to overheating in this situation. As one can surmise, the selected fail-safe
condition is dependent upon the application. One should also note the selected fail-safe
condition does not affect the controllability of the system, although it does impact certain
controller settings. The remaining four criteria are, debatably, the most important
parameters to properly determine for good controllability.
Valve Coefficient
Valve coefficient or C
v
is a measure of the volume flow rate of 25
o
C water through a
valve with a 1 psi pressure differential across the valve. It is determined by the following
equation:
v
GPM
C
P
SG
=
A

where:
GPMFluid flowrateexpressed as gallons per minute
PPressure dropacross control valve
SGSpecificgravity of fluid

For example, suppose you know there is a 15 psi pressure drop across a control valve
when the valve is wide open with a flow rate of 150 gpm of water through the valve. The
specific gravity of water is one. The valve coefficient can be calculated as:
1/ 2
150
38.72
15
1
v
gpm gpm
C
psi psi
= =

Once we know the valve coefficient, we can then calculate the pressure drop across the
valve given a flow rate, or a flow rate given a pressure drop. For example, determine the
pressure drop across the above valve if the flow rate increases to 200 gpm.
2
2
200
1 26.68
38.72
v
GPM
P SG
C
| |
| |
A = = =
| |
\
\
psi
In practice, once you know the design flow rate and the desired pressure drop, one can
calculate the required valve C
v
and select a proper valve from manufacturer's literature.
But how does one determine the desired pressure drop? This is accomplished by
designing for a specified valve authority.
Valve Authority
Valve authority is defined as the ratio of the minimum pressure drop across the valve
(that which occurs when the valve is wide open) to the maximum pressure drop across
the valve (that which occurs at the minimum controllable closed position of the valve)
expressed as a percent. In some literature, this is referred to as the Pressure Drop Ratio
(PDR). The maximum pressure drop is generally taken to be the pressure drop of the
system in which the valve controls. As such, we can mathematically define valve authority
as:
valve
pipe coil valve
P
P P P
|
A
=
A + A + A

where:
P
valve
Pressure dropacross control valve
P
pipe
Pressure dropof pipingincludingservice valves , etc.
P
coil
Pressure dropof thecoil


Figure 1 Characteristic curves of some valves
All control valves have an inherent performance characteristic. Figure 1 shows the typical
characteristic curve of an equal percentage, parabolic, linear and quick opening valve.
The equal percentage is typically used to linearize the output of a heat exchanger. The
output of a heat exchanger is also logarithmic and tends to have a characteristic opposite
that of an equal percentage valve. As such, the heat output of a heat exchanger
controlled by an equal percentage valve tends to be nearly linear with valve movement.
Refer to Figure 2.
+ =

A heating coil may have a characteristic similar to that shown in the graph above left. Ideally, one would
control this heat exchanger with an equal percentage valve with a characteristic similar to that shown in
the center. When the characteristics of each piece of equipment interact, they will produce a linearized
output. This is desirable in the area of control in order to ease the process of tuning or calibrating the
control loop
Figure 2 The linearization of a control process through proper selection of equipment

The equal percentage valve is so named because each the flow rate will increase an equal
percentage of the previous flow for every equal step change in valve movement. It is
mathematically defined as:
( ) 1 k x
flow percent e

A =
where:
0 1
0 1
k aconstant uniquetothevalve
x avaluebetween and representing valve position
closed open

= =

As an example, suppose k = -3.3648.
Then for x = 0.01 (1%), %flow = 0.03575
For x = 0.02 (2%), %flow = 0.03697
Thus, the change in flow rate for a 1% increase in valve position is:
( ) 0.03697 0.03575
0.034(3.4%)
0.03575

=
The parabolic valve provides flow variation that varies as the square of pressure drop.
From a basic course in fluids, you may remember that the pressure drop across any
hydraulic element varies with the following ratio:
2
1 1
2 2
PD GPM
PD GPM
| |
=
|
\

A valve with parabolic trim will vary the flow in a fashion opposite to this relationship.
This type of valve allows the flow rate of a fluid flow system to vary linearly with a
differential pressure measurement.
The linear valve is rather self-explanatory. The flow through the valve varies linearly with
valve stem position. Such valves are useful for controlling steam flow or providing
temperature blending of two fluid streams.
A look at the characteristic curve of a quick opening valve shows that such a valve make a
poor modulating control valve. However, it is often in on/off control systems.
The inherent characteristic of a control valve is valid only if it is the sole pressure drop in
the system. This can only occur if the valve is the only pressure drop within the system it
is controlling or if it is installed in such a fashion that the pressure at each valve port is
essentially equal. There are few applications when this occurs. Perhaps the most
common is a three-way valve installation at the interface of a primary-secondary system
where the valve is controlling a blending temperature at the secondary. Otherwise, the
control valve will always be some percentage of the total pressure drop in the circuit. As
such, the installed characteristic of the valve can be quite different from its inherent
characteristic. This is what valve authority measures.
For example, consider the system shown in Figure 3. The piping, including all fittings and
service valves, has a calculated pressure drop of 15 ft. w.g. The coil has a pressure drop
of 30 ft w.g. and the control valve has a pressure drop of 20 ft. w.g. Calculate the valve
authority.
20
0.307
15+30+20
= = |

S

Coil
R
Figure 3 Coil and associated subcircuit
The characteristic of an equal percentage
valve with different degrees of authority is
shown below. Note that at low values of
authority, the equal percentage valve
approaches the characteristic of a linear
valve. It can also be shown that a linear
valve approaches the performance characteristic of a quick opening valve when installed
with a low degree of authority. This curve also implies that systems should be designed
with a high valve authority.


Figure 4 An equal percentage valve with varying degrees of authority
The obvious question then becomes, 'What value of authority should I use?'.
Unfortunately, this is a question not easily answered. The selection of an authority value
is a compromise between controllability and the life-cycle cost of the system. A high
value of authority improves controllability, but at the expense of high pump head. This
means we need a larger, more expensive pump which not only drives up first cost, but
also operating cost. As such, a designer must make a judgment call based upon the needs
and desires of the client and the criticality of the process being designed as to whether
one should select a higher or lower value of authority.
For example, suppose you are selecting a valve for a critical industrial process. Further
assume that although the client wishes as low an operating cost as possible, accurate and
precise control is more important. In such a case, valve authority as high as 50% may be
justified. However, if we have a comfort heating process, one may be able to tolerate a
certain level of reduced controllability for purposes of reducing the size and operating
cost of the installed pumping system. In such a case, a value of authority around 30% may
be appropriate.
Depending on who you talk to, 'proper' value of authority may range from a low of 10%
to a high of 50%. However, most people agree valve authority should be above 0.23 (23%)
for acceptable controllability. On the other hand, one well known manufacturer of valves
and pumps recommends an authority of not less than 33%. In the end, the proper
selection of authority is often a judgment call based on one's experience with fluid flow
systems. For purposes of this module, authority should range between 30% and 50%.
Valve Rangeability
Another aspect of controllability relates to valve rangeability. Rangeability is a measure of
the minimum controllable flow the valve is able to handle. Mathematically, it is defined
as:
1/ 2 max
min
Q
R
Q
|
| |
=
|
\

where:
min
max
R valve rangeability
Q minimun controllable flow rate
Q design flow rate
installed valve authority |
=
=
=
=

The minimum controllable flow rate through a valve is a function of the valve design. It is
directly affected by all sources of friction within the valve assembly. In an ideal valve, any
change in signal applied to the actuator, even an infinitely small change, will force the
valve stem to move, even if that movement is infinitely small. However, friction
represents a force that must be overcome by the actuator. When the actuator exerts
sufficient force to overcome friction, the valve stem will move some finite amount. When
this occurs when the valve is full closed, this finite movement results in a certain minimum
flow rate.
For example, suppose a valve has an installed authority of 35%. The design flow rate is
350 gpm. For good low-flow control, we wish a minimum flow rate of not more than
2 gpm. We can calculate the required rangeability as:
1/ 2
350
0.35 123
2
R
| |
= =
|
\

This means the valve must have a manufacturer's rangeability rating of 123:1 or greater. A
typical commercial valve generally has a rangeability of about 50:1. Industrial valves can
have a rangeability as high as 200:1. This is part of the reason why industrial valves are so
much more costly than a commercial valve.
Valve Cavitation
A final criteria is valve cavitation. As the fluid passes through the restricted opening of
the valve, the local static pressure drops. If it drops below the vapor pressure of the fluid,
the fluid flashes to vapor; in other words, it boils thus forming gas bubbles. As the fluid
continues through the valve, the static pressure begins to rise. As this pressure rises
above the vapor pressure of the fluid, the bubbles begin to collapse. This collapse
releases a significant amount of energy in the form of a shock wave moving at mach one.
The energy released is sufficient to chip away at the material making up the valve plug
and valve seat thus seriously damaging the valve.
Valve cavitation often sounds like gravel flowing through the valve. Valve cavitation tends
to occur most frequently when flowing hot fluids. However, it is possible for cavitation to
occur in cold fluids as well. Cold water cavitation is far worse than hot water cavitation.
This is due to the fact that the ratio of the specific volumes of the fluid in liquid form to
the liquid in a gaseous state is significantly higher when the fluid is cold than when the
fluid is hot. For example, consider the value of specific volume of water at 60
o
F vs. water
at 240
o
F.
Temp Specific Volume (Fluid) Specific Volume (Gas)
60
o
F 0.01604 1206.32
240
o
F 0.01692 16.314
Table 1 Specific Volume of Water vs. Steam
Note the ratio of specific volume at 60
o
F is 1206.32 / 0.01604 = 75,200. This same ratio
for 240
o
F water is 16.314 /0.01692 = 964. This means that as a bubble in cold water
collapses, it releases 78 times more energy than when a bubble collapses in hot water.
Fortunately, cold water cavitation occurs far less frequently than hot water cavitation.
However, when it does, it will be significantly more destructive.
Engineers responsible for the design of a valve take great care to reduce the cavitation
potential as much as possible. The cavitation potential of a valve is specified by a value
known as the recovery coefficient. Mathematically, it is calculated as:
u d
c
u v
P P
K
P P

p

where:
( )
c
u
d
vp
K =Valverecoverycoefficient unitless
P=Pressureatapoint3pipediametersupstreamofvalve
P =Pressureatapoint12pipediametersdownstreamofvalve
P =Vaporpressureoffluidatoperatingtemperature

All pressures must be expressed as absolute pressure. Also note the numerator
represents the pressure drop across the valve.
As an example, consider a valve used to modulate the flow of 280
o
F process water. The
inlet pressure to the valve is 50 psig. What is the minimum downstream pressure allowed
for incipient cavitation if the valve has a recovery coefficient of 0.45?
From a steam table, we can determine the vapor pressure of water to at 280
o
F to be
49.226 psia. Let's assume an ambient barometric pressure of 14.7 psia. Then:
64.7
0.45
64.7 49.2
d
P
=

Solving for P
d
, we fine the minimum allowable downstream pressure is 57.7 psia (43 psig).
If we know the installed valve C
v
, we can predict the potential for valve cavitation.