PART B

Actuators

Aim: To study the inherent characteristics and installed characteristics of control valve.

Learning objectives

On successful completion of this experiment, you should be able to:

1. Relate variations of pressure drop across the control valve with the flow
2. Distinguish between inherent characteristics and installed characteristics of a control valve
3. Understand modeling of a control valve

Introduction and applications:

The control action taken by a controller is delivered by final control elements. Control valve is
one of the most commonly used final control element. A valve has three main parts: actuator,
positioner and body. Control valves are automatic and are operated by electrical, hydraulic or
pneumatic actuators. A pneumatic actuator mainly consists of a piston and a cylinder. The piston
is covered by a diaphragm, or seal, which keeps the air inside the cylinder, allowing air pressure
to force the diaphragm, moving the piston, which in turn moves the valve stem, which is linked
to the internal parts of the actuator. The valve under study here is an air-to-open valve which is
obtained by a reverse actuator and direct body (plug above the seat in Figure 1). A reverse acting
actuator will push the actuator stem upwards as air is applied to the bottom of the diaphragm. A
generic schematic diagram of a pneumatic control valve is presented in Figure 1 (Noltingk,
1987).

Figure 1: Schematic diagram for pneumatic diaphragm control valve (Reference: B. E. Noltingk,
Instrumentation Systems, Jones’ Instrument Technology, Vol. 4, 4th Edition, Butterworth, 1987)
Principles and Theory:

Inherent characteristics:
All control valves have an inherent flow characteristic that defines the relationship between
valve opening and flow rate under constant pressure conditions. These characteristic curves are
developed by doing experiments where the pressure drop across the valve is kept constant.

The flow rate through a valve is expressed as follows:

ΔP
Q=Cf x
ρ

where,
Q = Volumetric flow rate
C = Valve coefficient
x = fractional valve opening
ΔP = Pressure drop across the valve
ρ = Density of the fluid

Depending upon the shape of the valve plug we get different valve characteristics (and hence
different types of f(x)). Three main types are Quick opening, Linear and Equal percentage.
For quick opening type of valves, f(x) is generally taken as 𝑥. These valves give a large
increment in flow for relatively smaller valve opening.
The linear flow characteristic has a constant slope, meaning that valves of this type have constant
gain through complete range of valve openings. The function f(x) is equal to x.
In Equal percentage the change in flow is proportional to the flow just before the incremental
valve opening is performed. The function f(x) is given by a(x-1) for these valves.

Installed characteristics:
When the valve is installed in a plant with several other equipments, valve does not exhibit its
inherent characteristics and we get installed characteristics. This is because the elements will
have their own flow vs. pressure characteristics, there will be additional frictional loss in the
system and the effective characteristics of the valve will be different from the ideal system in
which inherent characteristic studies are performed.

Problem statement:

For the given control valve, (i) obtain parameters of the f(x) function characterizing the
relationship between fractional valve opening and the flow rate, (ii) compare the flow rate versus
fractional valve opening behaviour for the inherent and installed characteristics.

Experimental setup:

The valve under study is an air-to-open valve. The pump draws water from the tank and pumps it
through the valve. The flow rate is measured by a rotameter placed after the control valve. There
are two manually operated ball valves. One is placed in bypass line and other one just before the
control valve. The voltage supply to the control valve is provided by a voltage source. Voltage
can be varied between 0 and 5V to get 0% and 100% openings respectively. However, in the
setup we don’t recommend going below 3 V to avoid nonlinearities in the voltage to control
valve opening relationship. The pressure drop across the valve is measured by a mercury
manometer.

1.1 Schematic:

The schematic diagram showing the flow arrangement around the control valve is given below.

Figure 2: Schematic flow diagram for the experiment

1.2 SOP (startup, controls, shutdown)

1. Follow the switch-on sequence: Air supply, centrifugal pump, electrical input.
2. At startup, keep the bypass valve always open so that load on the centrifugal pump is reduced.
3. During the experiments, the bypass valve should be always kept closed.
4. The trade-off between the opening of manipulating valve and pressure drop to be maintained,
should be done in a controlled manner.
5. Follow the switch-off sequence: Electrical input, centrifugal pump and air supply.

Do’s and Don’t:

1. The electric supply to the control valve should be maintained between 3 and 5 V only.
2. Ensure proper electrical connections for the converter (voltage to current converter).
Otherwise, the control valve cannot be operated despite changing the voltage supply.
3. Always note the reading of flow rate and pressure drop on stabilization only.
Measurement, Apparatus and Chemicals:

1. Air-to-open control valve

2. Centrifugal pump
3. Rotameter
4. Ball valves
5. Voltage supply 5V DC
6. Mercury manometer
7. Water

Procedure:

Inherent characteristics:
1. The by-pass valve is kept closed. The manipulated valve is used to keep the pressure drop
constant.
2. Experiment is started with 5 V and 200 mm (Hg manometer). On reducing the opening of the
valve, ΔP increases. The manipulated valve is then slightly closed to bring the ΔP back to 200
mm. The flow rate is noted.
3. The voltage is decreased in steps of 0.2 V and the procedure is repeated till 3 V is reached.
4. Same procedure is followed but with ΔP = 150 mm (Hg Manometer) to obtain another set of flow
rate and fractional valve opening measurements.

Installed characteristics:
1. The by-pass valve is completely closed and manipulating valve is kept completely open to get the
maximum range of flow rate.
2. Opening of the control valve is changed by manipulating the voltage supply. However, the
pressure drop across the valve is now allowed to vary.
3. Flow rates at different voltages are noted down. An interval of 0.2 V is taken between two
successive measurements. The voltage is varied from 5 V to the minimum (3 V) and again to 5 V.
Hence two sets of data are obtained for installed characteristics.

Sample Observation Table format:

Table 1: Inherent characteristics. ΔP = 200 mm Hg. Trial 1

Voltage Opening Volumetric Volumetric Flow ΔP (N/m2) y=Q* Error%
Predicted
(V) fraction flow rate flow rate fraction mm N/m2 sqrt(ρ/ΔP) ln y
Q_frac=b(x-1)
(LPH) (m3/s) Q_frac Hg

Table 2: Installed characteristics. Varying ΔP. Trial 1

Voltage Opening Volumetric Volumetric Flow ΔP (N/m2) y=Q* Error%
Predicted
(V) fraction flow rate flow rate fraction mm N/m2 sqrt(ρ/ΔP) ln y
Q_frac=b(x-1)
(LPH) (m3/s) Q_frac Hg
Sample Calculation Tables:

Table 1: Inherent characteristics. ΔP = 200 mm Hg. Trial 1

Voltage Opening Volumetric Volumetric Flow ΔP (N/m2) y=Q* Predicted Error%
(V) fraction flow rate flow rate fraction mm N/m2 sqrt(ρ/ΔP) ln y Q_frac=
(LPH) (m3/s) Q_frac Hg b(x-1)
5 1 440 0.000122 1 200 26658 0.0000236 -10.650 1 0
4.8 0.96 410 0.000114 0.93 200 26658 0.0000220 -10.721 0.823 11.478
4.6 0.92 310 0.0000862 0.7 200 26658 0.0000166 -11.000 0.677 3.180
4.4 0.88 245 0.0000681 0.55 200 26658 0.0000131 -11.235 0.557 1.444
4.2 0.84 210 0.0000584 0.47 200 26658 0.0000113 -11.390 0.459 2.271
4 0.8 165 0.0000459 0.37 200 26658 0.0000888 -11.631 0.378 2.199
3.8 0.76 130 0.0000361 0.29 200 26658 0.0000069 -11.869 0.311 7.344
3.6 0.72 120 0.0000334 0.27 200 26658 0.0000064 -11.949 0.256 5.082
3.4 0.68 100 0.0000278 0.22 200 26658 0.0000053 -12.132 0.210 4.100
3.2 0.64 80 0.0000222 0.18 200 26658 0.0000043 -12.355 0.173 3.506
3 0.6 65 0.0000181 0.14 200 26658 0.0000034 -12.562 0.142 2.134
0 0 0 0 0 200 26658 0

Table 2: Installed characteristics. Varying ΔP. Trial 1

Voltage Opening Volumetric Volumetric Flow ΔP (N/m2) y=Q* Predicted Error%
(V) fraction flow rate flow rate fraction mm N/m2 sqrt(ρ/ΔP) ln y Q_frac=
(LPH) (m3/s) Q_frac Hg b(x-1)
5 1 450 0.000125 1 200 26658 0.0000242 -10.627 1 0
4.8 0.96 410 0.000114 0.91 228 30390.12 0.0000206 -10.786 0.823 9.533
4.6 0.92 340 0.0000945 0.75 270 35988.3 0.0000157 -11.058 0.677 9.635
4.4 0.88 300 0.0000834 0.66 305 40653.45 0.0000130 -11.244 0.557 15.463
4.2 0.84 265 0.0000737 0.58 325 43319.25 0.0000111 -11.400 0.459 20.806
4 0.8 225 0.0000626 0.5 345 45985.05 0.00000922 -11.593 0.378 24.372
3.8 0.76 195 0.0000542 0.43 348 46384.92 0.00000795 -11.741 0.311 27.604
3.6 0.72 165 0.0000459 0.36 348 46384.92 0.00000673 -11.908 0.256 28.812
3.4 0.68 140 0.0000389 0.31 370 49317.3 0.0000554 -12.103 0.210 31.942
3.2 0.64 120 0.0000334 0.26 470 62646.3 0.00000421 -12.376 0.173 33.197
3 0.6 110 0.0000306 0.24 476 63446.04 0.00000383 -12.470 0.142 40.421
0 0 0 0 0 0 0 0
Plots expected:

Fractional flow rate, Qfrac 1.2

0.8

0.6 Installed characteristics

Inherent characteristics
0.4

0.2

0
0 0.2 0.4 0.6 0.8 1
Fraction of opening, x

Figure 3: Plot of installed characteristics versus inherent characteristics

Fraction of opening, x
0 0.2 0.4 0.6 0.8 1
-6
-7
-8
-9
y = 4.863x - 15.48
-10
R² = 0.994
ln y

-11
-12
-13
-14
-15
-16

Figure 4: Linear plot for obtaining parameters of control valve characteristics

Expectations of a good report:

1. Report the details of the computations involved in obtaining the parameters of the inherent
valve characteristics. For example, some details are presented below:

Using Table 1, we may have 4 sets of data at a particular ΔP for inherent characteristics.

Model equation to which the data has been fit: y = a * b(x-1) (A)

For each set of data,

Plot fractional opening of control valve (x-axis) versus fractional flow rate (y-axis).

To find values of a and b, transform Equation (A) into a linear form as follows:
𝜌
If y = Q* ΔP
Then, ln (y) = x * ln (b) + (ln a – ln b) (B)
where x is the opening fraction of control valve

From the slope and intercept values from ln (y) versus x, we get,
a=? b=?

Sum of squared errors, SSE = sum of the squared differences between each observation and the
predicted observation = 𝑛𝑖=1 𝑦𝑖 − 𝑦𝑖 2
Sum of squared regression, SSR = 𝑛𝑖=1 (𝑦𝑖 − 𝑦)2
The total variation in the given data, SST = SSE+SSR
SSR
Coefficient of determination R2 = SST
(Note: R2 value may be calculated using graph plotting software tools like MATLAB or MS
Excel)

Thus, valve coefficient, a =

If two or more trials are recorded, average values of ‘a’ and ‘b’ may be reported.

2. Report the details of the computations involved in analyzing the data obtained for the installed
characteristics. For example, the following can be reported:

Using Table 2, we have data with varying ΔP as installed characteristics.

Average values of a and b obtained from the inherent characteristics are substituted in the model
equation as in Eqn. (A) to get predicted values (Pred) and actual values obtained from the
𝜌
experiment (Actual = Q* )
ΔP
Pred – Actual
Then, error % = x 100
Actual
ΔP
If the model Q = a * b(x-1) * results in large error %,
𝜌
ΔP
we need to consider other models say, Q=c * xd * (B)
𝜌
The values of c and d in the model in equation (B) can be obtained by appropriately transforming
the data to log scale. Report these estimated values along with the R2 value.

3. Compare the fractional flow rates as a function of the fractional valve opening for both
installed and inherent characteristics on the same plot (Figure 3). Comment on reasons for the
differences between these two curves.

Conclusion:

The model for equal percentage valve was used to fit the data obtained while studying the
inherent characteristics. It was observed that the model failed/was able to predict the installed
ΔP
characteristics of the control valve. Hence another model of the Q = a * xb * 𝜌
was tried/not
tried. This model was found to give better results, which is evident by the % error values.

It is concluded that model B is accurate/not accurate and has to be used/not used to characterize
the control valve studied in the experiment.

Find out (FAQ on experiment setup and procedure):

Q.1.What is the range of voltage applied for obtaining above 50% valve opening?
The required range of voltage is 3 to 5 V DC supply.

Q.2. Which valve in the setup is to be completely closed throughout the entire experiment?
Bypass valve has to be completely closed during the entire experiment.

Q.3. Are there any restrictions on the opening of manipulating valve during the experiment?
The opening of manipulating valve is to be adjusted so that constant pressure drop can be
maintained across the control valve for noting inherent characteristics. For procedure related
to installed characteristics, there is no need for change in the position of the manipulating
valve during the experiment.

Open ended questions:

Q.1. Find out information on various industrial examples for linear and non-linear control valves.

Q.2. Suggest different non-linear model functions relating flow and control valve opening.

Q.3. List down the advantages and limitations of electric (motorized) and pneumatic control
valves.

Q.4. Read further on valve positioner and its usage in control valve characteristics.
Safety:

2.1 Electrical safety

1. Ensure the electric wire does not come in contact with liquid water and that proper insulation
is provided.

2. The valve opening occurs with current to pressure (I2P) converter driving the valve stem
position. It is required that frequent check of the converter performance is done. Otherwise,
the control valve is reduced to a fail-open or fail-close condition.

2.3 Mechanical safety

1. The liquid should not enter into the manometer beyond a certain level. If it comes in contact
with the mercury column, the manometer will not function properly.

2. Regular maintenance of connectors, mechanical seals and other fittings is necessary to avoid
any leakage of liquid from the system.

3. Temperature-dependent condensate formation inside the control valve is possible. It can arise
in cases where inert gas like nitrogen is supplied for the movement of diaphragm.

4. Local corrections on amount of air supply for valve opening is taken care of by valve
positioner. The working condition of the positioner is critical as an additional controller in
smooth functioning of the valve stem.

2.4 Troubleshooting

1. Open the bypass valve with immediate effect on notice of any malfunctioning of the
centrifugal pump. It shall ensure reduced working load on the pump.

2. Operate the manipulating valve in a gradual manner so that any sudden fluctuation in pressure
drop that may affect the valve system is averted.

3. Always ensure that sufficient pressure (generally greater than 1 bar) is available on the
instrument air supply side during the experiment.