PSG COLLEGE OF

TECHNOLOGY

DEPARTMENT OF INSTRUMENTATION AND CONTROL SYSTEMS

ENGINEERING
B.E. INSTRUMENTATION AND CONTROL
ENGINEERING
19U602 - Process
Control
Quadruple Tank
System
(Benchmark System)

Presented by,
Madhumita K (22U226)
Nemi Siva Saravani M J (22U228)
Premalatha P (22U231)
Date : Sivarekha K (22U243)
26/03/2025 1
Quadruple Tank System (QTS)

The Quadruple Tank System (QTS) is a multi-

input, multi-output (MIMO) nonlinear
system used in control engineering for process
control studies. It consists of four
interconnected water tanks arranged in two
levels:

● Tanks 1 & 2 (Lower tanks)

● Tanks 3 & 4 (Upper tanks)

The system's objective is to control the water

levels in Tanks 1 & 2 by adjusting the pump
voltages and while considering the dynamic
interactions caused by water flow between the Figure 1. Quadruple Tank System
tanks.
2
Quadruple Tank System (QTS)

There are three types of flows in the

system:

1. Pump Inflow: Water enters from pumps

into upper tanks (Tanks 3 & 4).
2. Inter-Tank Flow: Water from upper
tanks flows into lower tanks (Tanks 1
& 2).
3. Outflow: Water drains out from each
tank through small holes at the bottom.

Figure 1. Quadruple Tank System

3
Mathematical Notations for the Quadruple Tank
System
Table 1 . Mathematical notions for the
System
Symbol Meaning

h1,h2,h3,h4 Water levels in Tanks 1, 2, 3, and 4 (cm)

A1,A2,A3,A4 Cross-sectional areas of the tanks (cm²)

a1,a2,a3,a4 Outlet hole areas of each tank (cm²)

v1, v2 Pump voltages (input signals)

k1, k2 Pump constants (convert voltage to flow rate)

g Gravity constant (cm/s²)

γ1,​γ2 Valve split factors (decide how much flow goes to

each tank)

4
Derivation of the Quadruple Tank System
Model(CONTD.)
Apply Mass Balance Equation

The fundamental equation to model the water

levels in each tank is:

Since volume = area height, we can write:

Now, let's apply this equation to each

Figure 2. Quadruple Tank System
tank.
5
Derivation of the Quadruple Tank System
Model(CONTD.)
Tank 1:

Tank 2:

Tank 3:

Tank 4:

6
Derivation of the Quadruple Tank System Model
(CONTD.)

We define small perturbations around an equilibrium point as,

where,

● hi* and uj* are the steady state values.

● h̃ i and ũj are small deviations from equilibrium.

By Taylor series expansion (1st order),

7
Derivation of the Quadruple Tank System Model
(CONTD.)

Re-writing mass balance equations in deviation variables,

(Where, hi* = giz + hi*)

8
Derivation of the Quadruple Tank System Model
(CONTD.)

Time constant:

For tank 1,

For tank 2,

9
Derivation of the Quadruple Tank System Model
(CONTD.)

For tank 3,

For tank 4,

State-Space Equations:

10
Derivation of the Quadruple Tank System Model
(CONTD.)

State-Space Equations:

Final State-Space Representation:

11
Derivation of the Quadruple Tank System Model
(CONTD.)

Tank 1 and 2 contribute to the output:

where y1and y2 are the outputs from Tank 1 and Tank 2, respectively, and are proportional to
the respective heights of the tank.

Thus,

The Output Equation,

12
Transfer function of the Quadruple Tank System
Model
Final Transfer Function of the System:

where,

13
 Transfer function of the Quadruple Tank System
Model (CONTD.)

Steady-State Parameters for Minimum and Non-Minimum Phase Systems

Table 2 Steady-State Parameters

Parameter Minimum Phase Non-Minimum Phase

Tank Levels (h₁, h₂) [cm] (12.4, 12.7) (12.6, 13)

Tank Levels (h₃, h₄) [cm] (1.8, 1.4) (4.8, 4.9)

Input Voltage (u₁, u₂) [V] (3.00, 3.00) (3.15, 3.15)

Pump Gain (k₁, k₂) [cm³/Vs] (3.33, 3.35) (3.14, 3.29)

Flow Ratio (γ₁, γ₂) (0.70, 0.60) (0.43, 0.34)

14
 Transfer function of the Quadruple Tank System
Model (CONTD.)
Transfer function matrix for Minimum Phase System

Transfer function matrix for Non-Minimum Phase System

15
 System Properties
1. Order of the System

The Quadruple Tank System is a fourth-order system because it has four state variables,
corresponding to the water levels in the four tanks.

2. Self-Regulating or Not?

🔹 Depends on the valve configuration:

● Minimum-phase configuration: Self-regulating (system naturally stabilizes)

● Non-minimum-phase configuration: Non-self-regulating (requires active control)

3. Batch or Continuous Process?

🔹 The QTS is a Continuous Process because,

● Water flows continuously through the system.

● The system does not operate in fixed cycles like batch processes.

16
 System Properties
4. Degrees of Freedom (DOF)

🔹 The degrees of freedom (DOF) are determined by the number of independent control inputs and
constraints:

● 4 state variables (tank levels: )

● 2 manipulated inputs (pump flow rates: )
● DOF = 2 (in a standard configuration)

5. Process Variables

Controlled Variables (CV): Water levels in the tanks.

Manipulated Variables (MV): Flow rates from the two pumps.

Load Variables (Disturbances): External disturbances such as fluctuations in pump flow, valve changes,
or leaks.

17
Pole zero Map of the Process Model

Figure 3. Pole Zero plot Minimum Phase (y1-u1) Figure 4. Pole Zero plot Minimum Phase 2 (y2-u2)
18
Pole zero Map of the Process Model

Figure 5. Pole Zero plot Minimum Phase 2 (y1-u2) Figure 6. Pole Zero plot Minimum Phase 2 (y2-u1)

19
Pole zero Map of the Process Model

Figure 7. Pole Zero plot Non Minimum Phase 2 (y1-u1) Figure 8. Pole Zero plot Non Minimum Phase 2 (y2-u2)

20
Pole zero Map of the Process Model

Figure 9. Pole Zero plot Non Minimum Phase 2 (y1-u2) Figure 10. Pole Zero plot Non Minimum Phase 2 (y2-u1)
21
 Suitable Mode of Control

Selection of Standard Control Mode (P, PI, or PID)

● System Characteristics:

○ The process has multi-input, multi-output (MIMO) behavior.

○ Interactions exist between the variables, requiring decoupling or advanced
control.
○ Non-minimum phase behavior may cause initial inverse response,
making standard PID challenging.

● Control Mode Selection:

○ For minimum phase system: PI control

○ For non-minimum phase system: PI control

22
Control Approach for Quadruple Tank System

Without Decoupler (Minimum & Non-Minimum Phase)

● Strong cross-interactions causing oscillations.

● Slower response with significant overshoot.

With Decoupler (Minimum Phase)

● Reduced cross-effects, leading to better setpoint tracking.

● Faster stabilization and improved disturbance rejection.

With Decoupler (Non-Minimum Phase)

● Inverse response is minimized, improving control stability.

● Enhanced performance despite non-minimum phase characteristics

23
Without Decoupler (Minimum Phase)- SIMULINK MODEL

Figure. 11.Without Decoupler (Minimum Phase) 24

Sub system (Minimum Phase)- SIMULINK MODEL

Figure. 12.Sub System (Minimum Phase)

25
Without Decoupler (Minimum Phase)- SIMULINK MODEL
response

Figure. 13.Without Decoupler (Minimum Phase) response

26
Without Decoupler (Non-Minimum Phase)- SIMULINK MODEL

Figure. 14.Without Decoupler (Non-Minimum Phase)

27
Without Decoupler Sub system (Non-Minimum Phase)- SIMULINK MODEL

Figure. 15. .Without Decoupler (Non Minimum Phase)

28
Without Decoupler (Non-Minimum Phase)- SIMULINK MODEL response

Figure. 16.Without Decoupler (Non Minimum Phase) response

29
Decoupling Control Approach
Why decoupling?

● In multivariable systems, inputs can interfere with multiple outputs.

● Decoupling control minimizes cross-interactions by compensating for
disturbances.
● Ensures each input primarily affects only its intended output.

Condition Number (CN) for Quadruple Tank System (QTS)

● Minimum Phase: 3.3311 (Easier to control)
● Non-Minimum Phase: 4.2614 (More challenging, but manageable)

30
RGA (Relative Gain Array)
● Determines the best input-output pairings.
● Helps in decoupling control loops.
● Reduces loop interactions in multivariable systems.

For Minimum Phase and Non minimum Phase systems

🔹 Best Pairing: 🔹 Best Pairing:

Input U1 → Output Y1 Input U1 → Output Y2
Input U2 → Output Y2 Input U2 → Output Y1

31
Decoupler Design
Where the Decoupler Matrix,

Where, For minimum Phase and Non-

Minimum Phase,

32
With Decoupler (Minimum Phase)- SIMULINK MODEL

Figure. 17.With Decoupler (Minimum Phase)

33
Decoupler (Minimum Phase)- SIMULINK MODEL

Figure. 18. Decoupler (Minimum Phase)

34
With Decoupler (Minimum Phase)- SIMULINK MODEL response

Figure. 19.With Decoupler (Minimum Phase) response

35
With Decoupler (Non-Minimum Phase)- SIMULINK MODEL

Figure. 20.With Decoupler (Non Minimum Phase)

36
Decoupler (Non-Minimum Phase)- SIMULINK MODEL

Figure. 21. Decoupler (Non Minimum Phase)

37
With Decoupler (Non-Minimum Phase)- SIMULINK MODEL
response

Figure. 22. .With Decoupler (Non Minimum Phase) response

38
FeedBack Control Structure for QTS

39
FeedBack Control Structure for QTS

Components Selected - for Feedback Control Structure

Table 3. Components List for FBCS

Component Suggested Choice Reason

Sensor Ultrasonic Level Sensor Non-contact, high accuracy,

low maintenance.

Final Control Element Variable Speed Pumps Provides smooth, precise

(FCE) control of water inflow.

Controller PI Controller (Proportional- Effective for level control

Integral) without excessive sensitivity
to noise.

40
Control Structure Response - MInimum Phase (Servo control
Loop 1)

Figure. 23 Servo Control loop 1 (Minimum Phase)

41
Control Structure Response - MInimum Phase (Servo control
Loop 2)

Figure. 24 Servo Control loop 2 (Minimum Phase)

42
Control Structure Response - MInimum Phase (Regulatory
control Loop 1)

Figure. 25 Regulatory Control loop 1 (Minimum Phase)

43
Control Structure Response - Minimum Phase (Regulatory
control Loop 2)

Figure. 26 Regulatory Control loop 2 (Minimum Phase)

44
Control Structure Response - Non MInimum Phase (Servo
control Loop 1)

Figure. 27 Servo Control loop 1 (Non Minimum Phase)

45
Control Structure Response - Non MInimum Phase (Servo
control Loop 2)

Figure. 28 Servo Control loop 2 (Non Minimum Phase)

46
Control Structure Response - Non MInimum Phase (Regulatory control Loop 1)

Figure. 29 Regulatory Control loop 1 (Non Minimum Phase)

47
Control Structure Response - Non Minimum Phase (Regulatory control Loop 2)

Figure. 30 Regulatory Control loop 2 (Non Minimum Phase)

48
IMC Tuning Method

What is Internal Model Control (IMC) Tuning?

● IMC is a model-based tuning method that improves stability, disturbance rejection, and
robustness.
● It designs the controller based on the inverse of the process model and a low-pass filter
to ensure stability.
● Provides a single tuning parameter (λ - filter factor) to control response speed.

Why Choose IMC Tuning?

● Robustness: IMC provides better stability and disturbance rejection.

● Handles Time Delays: Works well with systems that have delays.
● Easier Tuning: Uses a single λ parameter instead of complex PID adjustments.
● Improves Multivariable Control: Essential for Quadruple Tank System to handle
interactions.
● Minimizes Overshoot: Unlike standard PI/PID controllers, IMC reduces aggressive
transients.
49
IMC Tuning for Quadruple Tank System

Minimum Phase System

● Loop 1: [ Kc = 1.03 ; Ki = 0.0167]

● Loop 2 :[Kc=0.7568 ; Ki = 0.018 ]

Non-Minimum Phase System

● Loop 1 : [ Kc= 0.292 ; Ki = 0.00138 ]

● Loop 2 : [Kc = 0.313 ; Ki = 0.001307]

IMC tuning provides lower settling time, zero overshoot, and improved disturbance
rejection.
Ensures optimal performance in both minimum and non-minimum phase systems.

50
 IMC Tuning Results for Quadruple Tank System

Minimum Phase Response Non-Minimum Phase Response

Figure. 31 IMC (Minimum Phase) response Figure. 32 IMC (Non Minimum Phase) response
51
Industrial Applications for the process

Quadruple Tank System in Water Treatment Plants

● Water treatment plants require precise liquid level control in multiple interconnected tanks.
● Proper control ensures effective sedimentation, filtration, and chemical dosing.
● The Quadruple Tank System (QTS) is an ideal model for designing advanced control strategies.

Application of QTS in a Water Treatment Plant:

✔️Sedimentation: Removes large particles.

✔️Coagulation & Flocculation: Chemical dosing to clump small particles.
✔️Filtration: Passes water through layers to remove impurities.
✔️Disinfection: Uses chlorine or UV treatment for sterilization.

Why Control is Important?

● Prevents overflow and dry tanks, maintains efficiency of chemical dosing & adapts to
variable inflow rates.

52
Industrial Applications for the process

Control Implementation Using QTS

PI Controller with Auto-Tuning:

● Dynamically adjusts valve openings and pump speeds based on real-time water level
measurements.

Control Process Flow:

1️. Reference Setpoint: Desired water levels set based on treatment requirements.
2️. Error Calculation: Difference between measured levels and setpoints.
3. PI Controller: Adjusts pump flow rates and valve openings.
4️. Final Control Element: Pumps increase/decrease flow rates based on control output. Valves
regulate inter-tank water flow.
5️. Feedback from Sensors: Water level sensors continuously monitor real-time conditions.
Ensures response to sudden disturbances (e.g., heavy rainfall).
The system self-adjusts to maintain optimal operation.

53
REFERENCES

1. K. H. Johansson, “The Quadruple-Tank Process: A Multivariable Laboratory Process with

an Adjustable Zero,” ResearchGate, Accessed: Mar. 2025. [Online]. Available:
https://www.researchgate.net/profile/Karl-Johansson/publication/260586337_The_Quadr
uple-Tank_Process_A_Multivariable_Laboratory_Process_with_an_Adjustable_Zero/links/0
2e7e51aa0dba5fbdb000000/The-Quadruple-Tank-Process-A-Multivariable-Laboratory-Pr
ocess-with-an-Adjustable-Zero.pdf

2. Dr.Angeline Vijula D,“Quadruple Tank System Control,” Shodhganga, Accessed: Mar.

2025. [Online]. Available: https://shodhganga.inflibnet.ac.in/handle/10603/120333
3. A. Kumar, V. Panicker, and R. Lal, “A Quadruple-Tank Process Control Experiment,”
ResearchGate, 2015. [Online]. Available:
https://www.researchgate.net/publication/289265416_A_Quadruple-tank_process_contr
ol_experiment

54