Anda di halaman 1dari 18

CPB30004- Process Dynamics and Control

Experiment 4: Gas Pressure Process Control Page |1

SUMMARY

The objectives of this experiment of Experiment 4 Gas Pressure Process Control were to
identify the important components of the air pressure control system and to mark them in the
P&ID, to study single loop flow control using PID controller, to control the pressure in single
capacity and two capacity processes using PID controller and to tune the PID controller using
Ziegler Nichols closed-loop method. The procedure started with general procedure and the
main procedure have three trial where in first trial, PID controller is set with values of PB1=
70%, T11= 40 sec and TD1= 0 sec and controller is set to mode Auto (A) and then after the
response is stable, it is switched to mode manual (M) to decrease 5-10% of MV and is
switched to Auto (A) mode again. The same procedure is repeated for second and third trial
where in second trial, PB1= 45%, T11= 10 sec and TD1= 0 sec while in third trial, PB1=
20%, T11= 10 sec and TD1= 2 sec. Based on the chart obtained in Figure 4.1, and referring to
the Ziegler-Nichols Closed Loop Method explanation on stability of system, we can say that the
process is stable, as the amplitude of the curves does not varies so much, and the curves also
oscillated at the same frequency for each trial of process. Even though the results obtained
shows that the process was oscillating accordingly and frequently, however, different trials have
different values of proportional band (PB). Thus, the Ziegler-Nichols Closed Loop Method was
proved. The method itself developed their turning rules by stimulating a great number of
different processes and correlating the controller parameters with features of the step response.
Theoretically, the higher the proportional band (PB) values, the lower the control gain (Kc)
obtained. In this process, the results obtained did not proven the theory. Yet, the theory of as
proportional band (PB) values increase, the control gain (KC) values will decrease is not proven.
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control Page |2

1.0 INTRODUCTION

Pressure vessels function is to store and transmit liquid, gas or vapours under controlled
pressure. Pressure vessels played an important role in chemical and mechanical industries as a
storage vessels for gas phase components, reactions vessels to carry out a gas phase reaction,
because it will exert pressure equally in all directions of the walls and ends of pressure vessels.

In pressure vessels, it is necessary for precise control of pressure to maintain constant


concentration of components in vessels, maintain number of moles in vessels and to protect
vessel from being damaged due to a very high pressure.

The two main types of pressure vessels are according to the end construction and according to
the dimensions. In engineering two main shapes of the vessels can be found is cylindrical tank
vessel or spherical tank vessels.

In this process, the plant consists of two pressure vessels which are T91 an T92 with piping for
the flow of gas or vapour, valves to control the flows and fitting. The process plant consists of
pressure control valve (PCV), each vessel has its own vent (V), and bottom drain (D), air
pressure regulator (AR) and pipelines PL).

In single capacity process, only vessel T91 is used while vessel T92 is by-passed. Meanwhile,
for two capacity process, two vessels, T91 and T92 were used. It is important to fully understand
the piping and instrument diagram (P&ID) of the process before conducting the experiment to
ensure the steps is done correctly.

Theoretically, for this process, the Ziegler-Nichols Closed Loop Method is used. In this
method, the ratio of the amplitudes of subsequent peaks in the same direction is approximately
1/4:

A2 1
A1 = 4

And the higher the proportional band (PB) values, the lower the gain control (Kc) values.
Throughout this process, these two theories, were tested.
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control Page |3

PROCEDURE

START-UP PROCEDURE
1. The first step was to switch "PANEL, SCADA/DDC" selector switch and
switch on the main power supply.
2. The power supply was switched ON. The ACKNOWLEDGE button was pressed
to silence the buzzer.
3. The instrument air supply regulator was set to the pressure indicated at the air
regulators at the pneumatic control valves PCV91 and PCV92.
4. The by-pass valves around the control valves PCV91 and PCV92 were shut
and the pair of manual valves upstream/downstream of these control valves
were opened.
5. The control valve positioner (PP) of PCV91 was ON.
6. The compressed air was available at the pressure inlet air regulator AR91.
7. The vent valve (V) and bottom drain valve (D) of vessel T91 were shut.
8. Valves MV91B and MVIII were shut manually.
9. The process inlet manual valve MV91 was shut fully
10. The panel controller PIC91 was put in manual (M) mode and the MV was
adjusted to - 6.3% so the control valve PCV91 opened.
11. The final discharge manual valve (VP) was opened fully.
12. Pipelines PLI (Single capacity process) opened and PLII at their various
manual valves.
13. The manual by-pass valve B92 around vessel T92 was opened. Manual valves
at the inlet MVI and outlet MV92 of T92 were shut.
14. Valves MV92A and MV92B were both shut.
15. Pressure at AR91 was adjusted. MV91 was fully shut.
16. MV91 was opened and regulated until the flow rate at the variable area flow
meter FI92 reads 35 Nm3/Hr.
17. Check the pressure and flow rate setting:

a. PCV91 from PIC91 was set in manual (M) mode with


MV=106.3%. The pressure at vessel T91 must be around 22 27
psig.
b. PCV91 from PIC91 was set in manual (M) mode with MV=-6.3%.
The pressure at vessel T91 at the PV reading of PIC91 was not
more than 10 psig and the flow rate at FI92 reads 35 Nm3/Hr.
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control Page |4

ON/OFF PRESSURE CONTROL

1. The chart drive of the recorder was started by pressing the `RCD' pushbutton.
2. PCV91 from PIC91 was shut in manual (M) mode with MV=106.3%. The pressure
at PV at PIC91 was between the range of 22-27 psig.
3. PLII was opened through the manual discharge globe valve MV92A at the common
vent of PLII. The gate valve MV92B at PLII was closed.
4. The response of PT91 at the recorder PR91 was noted. PCV92 was opened when
PT91 exceeds 21 psig and closed when PT91 drops below a fixed Dead Band (0.5
psig), i.e., 20.5 psig.
5. The ON/OFF action was repeated by opening and closing of PC 91 by reducing the
MV such that pressure PV drops below 20 psig. Then shut PCV91 fully again with
MV = 106.3%.
6. The record chart drive was stopped and the relevant portions of the chart was
retrieved.

PID CONTROL OF PRESSURE IN SINGLE CAPACITY PROCESS

1. The chart drive of the recorder was started by pressing the `RCD' pushbutton at the
front of the recorder.
2. Pipelines PLI and PLII were opened throughout.
3. Globe valve MV92A at the common vent of PLII was opened manually.
4. Gate valve MV92B at the common outlet of PLII was closed.
5. Vessel T91 was used for single capacity operation.

a. By-pass valve B92 around T92 was opened.


b. The manual valves at the inlet, MVI, and at the outlet, MV92,
of T92 was shut.
c. The air flow path T91-PLI-FI92-PCV91-VP was opened
through B92.

6. The set point of PIC91,SV was set at 15 psig.


7. PID controller with the first (I) trial values: PB1 was set at 70%, T11 = 40 secs, TD1
= 0 sec.
8. The controller was transferred to Auto (A) mode and wait for the response to
become almost stable.
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control Page |5

9. PIC91 was switched to Manual (M) mode and decrease MV by about 5 to 10%.
Then, switch back to AUTO (A).
10. The response was slow so return the set point SV = 15 psig and reduced the TI1 to
20 secs and repeat the step (9) again.
11. The PID controller was set with the second (II) trial values: PB1 = 45%, TI1 = 30
secs, TD1 = 0 sec.
12. PIC91 was switched to Manual (M) mode, the MV was decreased to 5 to 10%,
switched back to Auto (A) mode and the response was observed for a few minutes.
The response was oscillatory.
13. PB1 was increased by 5%. Repeat the step (11).
14. The third (III) trial values for the PID controller: PB 1 = 20%, TI1 = 10 secs, TD1 =
2 sec.
15. The disturbance was introduced in MV and the response was observed.
16. The chart drive was stopped and the relevant portions of the chart paper was
retrieved.

SHUTDOWN PROCEDURE

1. The controller PIC91 was switched to Manual(M) mode, and the MV was set to 100%.
2. Push the recorder RCD button to stop recording.
3. The main power supply was switched off.

4.0 RESULTS & CALCULATIONS

4.1 PIPING & INSTRUMENT DIAGRAM


CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control Page |6

4.2 START-UP PROCEDURE REPORT

In a start-up procedure before starting the Gas Pressure Process Control experiment, the
PANEL, SCADA/DDC selector switch was switched to PANEL, SCADA position. The
main power supply was switched ON. The RCD button was pressed to stop the recorder drive
from running and ACKNOWLEDGE button was pressed to silence the buzzer when
annunciator gets activated.

The instrument air supply regulator was set to pressure indicated at air regulators at pneumatic
control valves PCV91 and PCV92. The purged water in air regulators was purged by opening the
bottom drain valves. The by-pass valves around the control valves PV91 and PCV92 are shut
fully and pair of manual valves upstream/downstream of control valves were opened. The control
valve positioner (PP) of PCV91 is connected, was checked. The compressed air availability at
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control Page |7

pressure inlet air regulator AR91 was ensured. The vent valve (V) and bottom drain valve (D) of
vessel T91 was checked to ensure it is shut.

The manual valves MV91B and MV and process inlet manual valve MV91 were shut fully.
The panel controller PIC91, was placed in manual (M) mode and its MV was adjusted to MV =
-6.3% to open control valve PCV91.the final discharge manual valve (VP) was open fully. The
two pipelines PLI an PLII were traced in the process air supply as following:

Pipeline PLI: single capacity process:

AR91-MV91-T91-PLI-B92-FI92-PCV91-VP

Pipeline PLII:

AR91-MV91-T91-PLII-MVII-PCV92-MV92A-MV92B-V

The pipelines PLI (Single capacity process) and PLII were opened fully at their various manual
valves. The manual by-pass valve B92 around vessel T92 was ensured opened and manual valves
at inlet MVI and outlet MV92 and T92 were ensured shut. Shut both valves MV92A and
MV92B. The pressure at AR91 is approximately at pressure indicated at AR91 was checked and
adjusted, MV91 was fully shut. MV91 was then slowly open and regulated, until flow rate at
variable area flow meter FI92 showed reading of 35Nm3/Hr.

Lastly, the pressure and flow rate setting were checked:

a. PCV91 and PIC91 was shut fully in manual (M) mode with M=106.3%. The
pressure at vessel T91 at PV reading of PIC91 was then read. The pressure
reading was 24.4 to 24.9 psig which is in range (22-27 psig).
b. PCV91 was opened fully from PIC91 in manual (M) mode with MV=6.3%. The
pressure at vessel T91 at PV reading of PIC91 was then read and ensure that the
reading is not more than 10 psig (the reading was at range between 9.0 to 9.4
psig) and the flow rate at FI92 reading was 35 Nm3/Hr which is correctly set.
c. The process air supply pressure and flow rate were correctly set.

4.3 RESPONSE CURVES OF CONTROLLER


CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control Page |8

In a process dynamic & control, it is important to understand the terms in response curve.
understanding the terms in the response curves can help in determine what is a setpoint, and what
is deadband. Setpoint in a process control means desired or target value for an essential variable
of a system, often used to describe a standard configuration or norm for system. Deadband is a
period of dead state of a system (sometimes called neutral zone or dead zone) is an interval of
signal domain or band where no action occurs (system is dead or output is zero).

In Gas Pressure Process Control experiment, the setpoint for PID Control of Pressure in Single
Capacity Process was SV=15 psig. The disturbance was introduced to the process to observed
the change in oscillatory. Disturbance is a variable, an effect of surrounding towards the process.

At first trial, the PID controller was set as below:

PB1=70%, TI1=40 secs, TD1=0 sec

During this trial, after the disturbance was introduce to the process, there are no constant
oscillation and the chart print out a dead band, where the system was in a dead state or neutral
zone. The value of TI1 was then reduced to 20 seconds and the process were repeated to observe
the changes. After the changes of values TI1 reduced to 20 secs, the chart drive print out a
constant oscillation with constant frequency, however the amplitude was not constant.

The experiment was then prepared for the second trials, where the values were set as shown
below:

PB1=45%, TI1=30 secs, TD1=0 sec

In this trial, the response curves printed out at the chart paper shows that the response were
oscillatory. Even after the disturbances was introduced, the oscillatory was not affected, and the
process operated as usual. At this trial, the controller managed to reject the disturbances
introduced to the process and force the process variable back to desired condition of process, the
setpoint.

The experiment was then operated, with the values of the of PID controller being set to:

PB1=20%, TI1=10 Secs, TD1=2 secs


CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control Page |9

The results in the chart paper shows that the response curves were oscillating at constant
oscillations. Even after the disturbances was introduced, the curves show no effect and the
controller managed to operated accordingly by rejecting the disturbances and force back the
process to the setpoint of the process.

4.4 RESPONSE CURVES FOR SINGLE CAPACITY & TWO-CAPACITY

Figure 4.1: Single capacity process results

For single capacity process, only vessel T91 was used and vessel T92 was by-passed. The set
point of PIC1 was set to 15 psig. After setting the PID controller with values of PB1 to 70%, TI1
to 40 seconds and TD1 to 0 second, the controller was set to Auto (A) mode and wait for
response to become stable. By stable, it means that the oscillation of the curves was oscillating
properly. When the oscillation is stable, the process was switched to Manual (M) mode, and
decreasing the MV to range of 5 to 10%. During this process, the response curves produced dead
band, and too slowed.

Due to response of process too slow, by following the procedure, the value of TI1 was reduced
to 20 seconds, and the process were repeated. The response curves produced shows a constant
oscillation, and the process was stable.

The process was then continued for the second trial by setting the parameter at the PID
controller with values of PB1=45%, TI1=30 secs and TD1=0 sec. The results obtained for this
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control P a g e | 10

trial was almost the same, it is oscillating at constant frequency but a slight difference in
amplitude can be seen. When the disturbance were introduced to the process, the process were
not affected and still operate as usual.

Lastly, the third trial for single capacity process were operate, with the parameter values being
change to PB1=20%, TI1=10 secs, TD1=2 secs. During this trial, the response curve oscillated at
the same frequency as the first and second trial, but same amplitude as the second trial. In this
process, the disturbance also did not affect the process and the controller able to operate
correctly.

Even though the results obtained shows that the process were oscillating accordingly and
frequently, however, different trials have different values of proportional band (PB).
Theoretically, the higher the proportional band (PB) values, the lower the control gain (Kc)
obtained. In this process, the results obtained did not proved the theory.

100%
PB
Kc
If the operating point varies and if the process dynamic properties depends on the operating
point, you should consider using adaptive control or gain scheduling, where the PID parameter
are adjusted as functions of the operating point. If the controller parameters shall have fixed
value, they should be tuned in the worst case as stability is regarded. This ensures
proper stability if the operation point varies. The worst operating point is the
operation point where the process gain has its greatest value and/or the time delay
has its greatest value. (Haugen, 2010)

4.5 RESPONSE CHARACTERISTICS

Ziegler and Nichols [1] used the following definition of acceptable stability as a basis for their
controller tuning rules: The ratio of the amplitudes of subsequent peaks in the same direction
(due to a step change of the disturbance or a step change of the setpoint in the control loop) is
approximately 1/4, such as below:
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control P a g e | 11

A2 1
A1 = 4

However, there is no guaranty that the actual amplitude ratio of a given control system becomes
1/4 after tuning with one of the Ziegler and Nichols methods, but it should not be very different
from 1/4. (Haugen, 2010)

Figure 4.2: According to Ziegler-Nichol, if A2/A1 = the process or stability of system is


OK

Based on the results obtained in Figure 4.1, and referring to the Ziegler-Nichols Closed Loop
Method explanation on stability of system, we can say that the process is stable, as the amplitude
of the curves does not varies so much, and the curves also oscillated at the same frequency for
each trial of process. Thus, the Ziegler-Nichols Closed Loop Method was proved. Yet, the theory
of as proportional band (PB) values increase, the control gain (KC) values will decrease is not
proven.
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control P a g e | 12

ANALYSIS

Pressure vessels are use as storage vessels for gas phase components. Precise control
of pressure in the pressure vessels is necessary in order to maintain a constant concentration
of components in the vessel and maintain the number of moles in the vessel.
The flow of air was observed by manipulating the parameters. Considering, that this is
a single capacity process and certain amount of pressure was applied, we are required to use
a proper control to the valve manually. For this experiment we are using pneumatic control
valve. There are two types of actuators in pneumatic which are air-to-close and air-to-open
control valve. The valve contains upper and lower chamber. The spring was located upwards or
downwards of diaphragm to push or pull control valve (open or close), the stem and control
valve.
Based on the results obtained we can say that the process is stable. The amplitude showed
that the curves oscillated at the same frequency for each trial of process. Hence, the Ziegler-
Nichols Closed Loop Method was proved. Unfortunately, the value of (PB) and (KC) does not
proved the theory. Theoretically, if the value of proportional band (PB) is increasing then, the
value of control gain (KC) must be decreasing.
Basically, pressure vessels are meant to operate above than 15 psig. If the vessels
cracked or damaged can result in gas leakage or rupture failure. The employees are at risk of
being expose to poisonous gas, suffocations and explosions. Protect the vessel from being
damaged due to very high pressures
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control P a g e | 13

DISCUSSION

In this process, we are about to make an observation about flow of air when certain
parameter had been manipulated. Parameter that had been changed during this process is value
of MV to get value of PV below than 20 psig. Other parameters that need to change are values of
PB1, T11, TD1 and also change the process from manual to auto. In this experiment, we only
conduct control of pressure in single capacity process.

Throughout this experiment, we can conclude that this experiment had been conducted
successfully. As an observation, this experiment required final control element which is control
valve to operate. Since this process involve pressure applied, hence the requirement to use a
proper control valve must be done to avoid any accident occurs in the future. There are three
types of control valve that commonly used in industry which are pneumatic control valve,
hydraulic control valve and electrical control valve. Basically, pneumatic control valve requires
air pressure to activate it, hydraulic control valve requires water to be activated and electrical is
using electric to operate. (TotalValveSolutions, 2014)

In this experiment process, we are using pneumatic control valve since the air pressure is
using to activate it. There are two types of actuators in pneumatic which are air-to-close and air-
to-open control valve.
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control P a g e | 14

Figure 1: Air-to-open and Air-to-close control valve

Basically, these type of control valve using a same mechanical to operate. The important
mechanism is the diaphragm; which contain upper and lower chamber, the spring; located
upwards or downwards of diaphragm to push or pull control valve (open or close), the stem and
control valve.

To support information stated from TotalValveSolutions, according to (Bartelt, 2015), the


only different between the two actuators is the position of air pressure inlet into the control
valves whether it is located on the upper chamber of diaphragm or lower chamber of diaphragm.
This is due to safety precaution when control valve is operated.

In this experiment, it is safer to apply air-to-close control valve in the process. When
pressure is lost during the process, the control valve will open to avoid process overheat. By
referring to Figure 1, since the air is lost, the diaphragm will moves upwards and the spring will
move upwards to open the valve. This is called fail-open. When a failure occurs in a process,
control valve will open to avoid overheating in a process flow.

In this experiment, we also get to know about on / off control. This control basically
assigns other control element. They can be one or two position control. They control final
element which is either they will fully close or open (Control Station). But in industry,
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control P a g e | 15

sometimes we need to change the set point immediately due to current situations but on / off only
works to open or close final control element. Thus, a set of boundaries need to be introduce to
protect the final control element from become wear. This set of boundaries is calling a dead
band. It is a range operation around set point and have lower and upper boundary. Within this
limitation, dead band will reduce the wear of final control element and the changes can be easily
makes without damage the instrument.

Figure 2: Range of dead band.

To support information from Control Station, in similarity of dead band according to


(Technology Transfer, 2014), dead band can be define asthe time difference between when a
change occurs in a process and when the change is detected. Technology Transfer has same
information regarding the dead band which is support the previous information.

This experiment also as expose to us about single and two capacity process. For gas
pressure process control experiment, we just conduct the single capacity process. There are
different between these two processes. The different that can be discussed are overshooting
occurred and the disturbance. It is clearly show that the overshooting at single capacity is much
lower and became increasing from the first trial until third trial by theoretical. While at the two
capacity, the overshooting is higher and became decreasing from first trial until third trial by
theoretical.
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control P a g e | 16

CONCLUSION

To conclude, in this experiment, P&ID controller can be tuned by using Ziegler Nichols
Close-Loop Method. The method developed their turning rules by stimulating a great number of
different processes and correlating the controller parameters with features of the step response.
Basically, close-loop more accurate than open loop in terms of less sensitive to noise and
disturbance and steady state error and transient response can be controlled easily in close-loop
system by simple adjustment of amplification in loop. The experiment that was conducted is only
single capacity process based on the result on the chart where overshooting at single capacity
process was lower than and became increasing starting from first trial until third trial compared
to two capacity process. The set boundaries or dead band is a range operation around set point
and have lower and upper boundary where it reduced the wear of final control element and
changes without damage the instrument. Based on the chart that was obtained, it can be conclude
that the process is stable and successful. The amplitude showed that the curves oscillated at the
same frequency for each trial of process. Hence, the Ziegler-Nichols Closed Loop Method was
proved. Based on theory, if the value of proportional band (PB) increases then, the value of
control gain (KC) will decreases.

SUGGESTIONS AND RECOMMENDATIONS

As the recommendation, in order to obtain more accurate and desirable result and data of
this experiment in the future, it is necessary to make sure manual valves is in correct whether
open or shut and the manual valve and pressure valve reading is in range so that there will no
error occurred during the experiment. Next, for every trial, make sure the PIC91 is set up to Auto
(A) mode first before record the chart and waits for a few minutes for response to stable. Then,
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control P a g e | 17

make sure the result or chart is recorded subsequently the data is set up in monitor so that the
output response in the chart will stable. The number of trial, P&ID values and the set point must
be marked on the chart paper properly so that we will not confused whenever the values are
changed. Lastly, the output response of the chart must be observed properly whether the response
is oscillatory or not where if the response is oscillatory, the PB1 must be increased by 5% to
obtain stable response and low value of control gain values (KC) by increasing the proportional
band (PB) values.

ADDITIONAL QUESTIONS

1. What are the significance of this experiment?

The Pressure Process Control course offers students hands-on training in the measurement and
control of pressure processes. Students raise and lower air pressure to set the levels using manual
control and auto-tuning methods. Students learn how the E/P converter with the control valve
affects the pressure in the air tank. They create variable demand and supply disturbances by
manipulating the panels valves.

2. Why should we change the parameters?

Pressure vessels play an important role in chemical industries as storage vessels for gas phase
components, reaction vessels to carry out a gas phase reaction. Thus, by changing the
parameters we can put the machine to a test where the machine can conduct a precise control
of pressure in the pressure vessels is necessary in order to maintain a constant concentration
of components in the vessel, maintain the number of moles in the vessel, and protect the
vessel from being damaged due to very high pressures.

3. What are the objectives of the experiment?

To identify the important components of the air pressure control system and to mark them in
the P&I Diagram To study single loop flow control using PID controller, to control the pressure
CPB30004- Process Dynamics and Control
Experiment 4: Gas Pressure Process Control P a g e | 18

in single capacity and two capacity processes using PID controller and to tune the PID
controller using Ziegler Nichols closed-loop method.

References

1. TotalValveSolutions (2014, July 14). Introduction to Valve and Actuators. Retrieved from
https://www.youtube.com/watch?v=tiyOHzgSa-s

2. Technology Transfer. (2014, February 25). Retrieved October 24, 2016, from Process
Control Fundementals: https://www.techtransfer.com/blog/process-control-fundamentals/

3. Bartelt, T. (2015, October 15). Air-to-open / Air-to-close. Retrieved October 24, 2016,
from Wisc Online: https://www.wisc-online.com/learn/career-clusters/stem/ele2007/air-
to-openair-to-close-valves

4. Control Station. (n.d.). Retrieved October 24, 2016, from What is on-off control?:
http://controlstation.com/what-is-on-off-control/

5. http://techteach.no/publications/articles/zn_closed_loop_method/zn_closed_loop_method
.pdf