CL3601 – Process Control

Lecture 2 – Process Dynamics I

Lecture outline:
• Introduction to first order process and their dynamics
• Explanation of process gain, lag, deadtime, time constant
• Explanation of different input functions(step, ramp, sinusoidal )
• Explanation of transfer function and their use in modelling system.
• Examination of how first order process respond to different input functions.

LIQUID TANK EXAMPLE

In previous lecture the outlet and inlet flow were controlled by pumps, In integrating
processes like this, the system doesn't naturally stabilize or return to a baseline condition
after a disturbance. Instead, it keeps "integrating" the effects of changes, leading to
potentially unbounded behaviour unless managed.

What will happen if inlet and outlet flow is controlled by valves?

The introduction of valve make changes in the system. Using pumps gives proper and
desired steady flow but in case of valves its more related to pressure difference. The
dynamics is now more closely related to the pressure difference across the valves and the
height of the liquid in it.

1. Objective: The primary goal here is to calculate the outlet flow rate (Fout) when a
valve is present.
2. Valve Resistance: The valve inherently provides a resistance to the flow. This
resistance will influence the flow rate out of the tank, especially in comparison to a
pump which might provide a more consistent flow rate.
3. Flow Proportional to Pressure Drop: A critical point mentioned is that the flow
through the valve is proportional to the pressure drop (ΔP). This relationship implies
that as the pressure difference across the valve increases, the flow rate will also
increase, and vice versa.
The pressure drop across the valve is influenced by the liquid's height in the tank, calculated
using the formula ρgh, where ρ is liquid density, g is gravity, and h is liquid height. The
valve's resistance affects the flow rate, and its characteristic curve or flow coefficient is
essential to determine the outlet flow, Fout. Knowledge of the valve's specifics and other
system pressures, like Pin, will provide an accurate outlet flow rate.

1. The valve inlet pressure is influenced by the liquid's height and is represented by:

Pin=ρgh+Pamb
Where:
o Pin is the valve inlet pressure.
o ρ is the liquid's density.
o g is gravitational acceleration.
o h is the liquid's height.
o Pamb is the ambient pressure.

2. The pressure drop, considering the outlet at ambient pressure, is:

ΔP = Pin - Pamb =ρgh

This model describes how the height of liquid impacts the valve inlet pressure and the
resulting pressure drop.

Key points:
1. Flow Condition: If the flow through the valve is turbulent and there's no phase
change, the outlet flow is influenced by the pressure drop across the valve.
2. Equation: The outlet flow, Fout, is described by the equation:
Where:
∆𝑷
𝑭𝒐𝒖𝒕 = 𝒌𝒗𝒂𝒍𝒗𝒆 𝐱 &
𝝆𝒍
• ΔP is the pressure drop.
• 𝜌( is the liquid density.
• 𝑘)*()+ is a constant that's related to the design of the valve and also depends on how
open the valve is.

Simplification: The equation for outlet flow, Fout, can be simplified further by substituting
some terms. The given equation is:

Constant Explanation: The constant kg is derived from ,𝑔. It essentially relates the flow
out of the tank to the gravitational acceleration.
 Mass Balance: From a previous lecture on mass balance, we have the following equation
that relates the change in liquid height, h, over time, t, with respect to the inlet flow, Fin, and
outlet flow Fout. Substituting Fout derived above into the equation relating liquid height and
over time we get:

A critical takeaway from this equation is that the rate of change of the liquid level in the tank
(with respect to time) is directly influenced by the current liquid height. More specifically, the
outflow from the tank (which depends on √ℎ) affects how quickly or slowly the level in the
tank changes. This means, for instance, when the liquid level is high, the outflow is larger,
leading to a faster decrease in the level. Conversely, when the liquid level is low, the outflow
reduces, resulting in a slower decrease in the level.

What happens if the system is disturbed?

Disturbances can be changes in input flow rates, sudden removal or addition of liquid, or
any other unexpected event affecting the tank. When a system is disturbed:
• The liquid level might deviate from its steady state.
• Depending on the type and magnitude of the disturbance, the system might take
some time to return to a new steady state or might oscillate around the steady state.
• The response of the system to such disturbances is vital for control engineering, as it
helps in designing controllers that can bring the system back to the desired state.

INPUT FUNCTIONS:

1. Step Input:
o Represents a sudden and immediate change in the input to a process,
analogous to suddenly opening a valve.
o The graph shows a rapid rise from one value to another, stabilizing at the new
value. This is typical for systems subjected to a step change, where the
system transitions from one steady state to another due to a sudden
alteration in input.

2. Ramp Input:
o Represents a slow and steady change in the input.
 o The graph depicts a linear increase, illustrating a scenario where there's a
consistent change over time, such as slowly increasing the flow rate of a
liquid into a tank.

3. Sinusoidal Input:
o Represents cyclical and periodic fluctuations in the input.
o Such inputs can correspond to natural variations or disturbances that a
system might experience, like the daily temperature fluctuations.
o The graph displays an oscillatory behaviour, characteristic of systems
subjected to sinusoidal inputs.

When designing control systems or analysing process dynamics, it's vital to understand how
a system reacts to different types of inputs. These input functions allow engineers to predict
system behaviour, design appropriate controllers, and ensure that the system operates
safely and efficiently under various conditions.

RESPONSE TO CHANGE:

Step Change in Input:

• Fin = 2 m³/s for t<1000

• Fin = 2.5 m³/s for t>1000

Process Behaviour:
The system's response to the input change is graphically depicted. Initially, the tank level
rises with a 2 m³/s input. Post 1000 seconds, the increased flow rate accelerates this rise,
evident from the curve's steeper slope.

Reaching a New Steady State:

Eventually, the system attains stability, indicated by the curve plateauing. At this steady
state, the tank's liquid level is constant due to balanced inflow and outflow rates.

Physical Interpretation:
The system's behaviour is intrinsically linked to its physics. An increase in liquid height
boosts the flow rate through outlets like valves, owing to the gravity-induced pressure
increase from the heightened liquid column.

PROCESS GAIN:
Process gain measures the system's response to changes in input. It's calculated as the
ratio of the change in output to the change in input when the system reaches a steady state.

For this specific example:

• The gain illustrates the relation between the change in the liquid's height and the
variation in flow rate.
• Using the given values:

ΔOutput (change in height) = 15.9 - 10.2 = 5.7 m

ΔInput (change in flow) = 2.5 - 2 = 0.5 m³/s

Hence, the Gain = ΔOutput/ΔInput = 5.7/0.5 = 11.4 s/m

Thus, the process gain is 11.4 s/m.

LAG:

Lag represents the time taken for a system to achieve a new steady state following a
change. Its duration depends on the process time constant. The lag amount is associated
with the system's ability to store mass and/or energy, often referred to as the 'capacitance'
of the system.

TRANFER FUNCTION:

A transfer function, often denoted as G(s), defines the relationship between the input and
output of a system in the frequency domain. It can be represented as:
Laplace Transform: The Laplace transform is a powerful mathematical tool used primarily
for solving linear differential equations and analysing linear time-invariant (LTI) systems. It
transforms a function from the time domain, usually denoted by t, into the frequency domain,
usually denoted by the complex variable s.

Why use Laplace Transforms?

• It transforms differential equations into algebraic equations, which are generally

easier to solve.
• It provides a unified framework for analysing and designing systems in the frequency
domain.
• It's a standard tool in control theory and systems analysis.

Connection to Transfer Functions: Transfer functions are a crucial concept in control

theory and systems engineering. They describe the relationship between the input and
output of a system in the frequency domain.

If you have a linear differential equation that describes a system, taking the Laplace
transform of the equation will result in an algebraic equation in s. From this, the ratio of the
Laplace transform of the output to the Laplace transform of the input gives the system's
transfer function.

• Laplace transforms are a way to solve linear ordinary differential equations.

• in this course we will stick to linear equations.

Simple Linear Model: The slide is introducing a simplified linear model to describe a
process. The equation given is:
This equation is a basic first-order differential equation representing the change in height
(dh/dt) of the fluid in a tank as a function of the inflow rate and the outflow rate which
depends on the height and valve resistance.

Deviation Variables: The concept of deviation variables is introduced in the slide.

Instead of dealing with the absolute values of variables (like h and Fin), we look at their
deviation from a steady state or a reference value:

• h′ is the deviation in height from the steady-state value, defined as h′=h−hˉ, where hˉ is
the steady-state height.
• Similarly, Fin′ is the deviation in the inflow rate from its steady state, defined as:
Fin′=Fin−Finˉ.

Using deviation variables can simplify the analysis, especially when dealing with systems
around their operating or steady-state points.

Given the differential equation:

𝒅𝒉′ 𝒉′
𝑨 = 𝑭! 𝒊𝒏 −
𝒅𝒕 𝑹

!",
Laplace Transform of Derivatives: The Laplace transform of the first derivative of
!#
is
sH′(s) − h′(0), where:

• s is the Laplace variable.

• H′(s) is the Laplace transform of h′.
• h′(0) is the initial value of h′.

If we assume zero initial conditions (which is common for deviation variables), h′(0)=0.
Hence, the transformed derivative becomes simply sH′(s).

2. Laplace Transform of Other Terms:

• The Laplace transform of h′ is H′(s).

• The Laplace transform of Fin′ is Fin′(s).

Transforming the Entire Equation: Using the above results, our differential equation in the
Laplace domain becomes:
𝑯′(𝒔)
𝑨𝒔𝑯! (𝒔) = 𝑭! 𝒊𝒏 (𝒔) −
𝑹

Rearranging: Grouping like terms and isolating the transfer function (ratio of output to input
in the Laplace domain), we get:
𝟏
𝑯! (𝒔) /𝑨𝒔 + 2 = 𝑭! 𝒊𝒏 (𝒔)
𝑹

This can be further rearranged to get the transfer function

𝑯′(𝒔) 𝑹
!
=
𝑭 𝒊𝒏 (𝒔) 𝑹𝑨𝒔 + 𝟏
 𝑴
For a step change, 𝑭! 𝒊𝒏 (𝒔) = . Substitution and rearranging:
𝒔

𝑹 𝑴
𝑯′(𝒔) = / 2
𝑹𝑨𝒔 + 𝟏 𝒔

Inverse Laplace Transform:

&𝒕
𝒉! = 𝑹𝑴(𝟏 − 𝒆𝑹𝑨 )

Combining Steady-State and Deviation:

&𝒕
𝒉 = 𝒉& + 𝒉! = 𝒉& + 𝑹𝑴(𝟏 − 𝒆𝑹𝑨 )

1. Response of the Process:

&𝒕
𝒉 = 𝒉& + 𝑹𝑴 /𝟏 − 𝒆𝑹𝑨 2
This equation, which is similar to the one from the previous slide, represents the total
height 𝒉 as a combination of the initial steady-state height 𝒉& and the transient
response due to a change in the input.
For convenience, they introduce a time constant τ where: τ = RA The equation can
then be simplified to:
&𝒕
𝒉 = 𝒉& + 𝑹𝑴 /𝟏 − 𝒆 𝝉 2

2. Steady-State Gain and Resistance: The slide mentions that the steady-state gain in
this context is equivalent to the resistance, R, of the process. This implies that the gain
represents how much the output (height, in this case) changes for a given change in
the input. The final value that the process reaches (or the difference in height from the
initial state) is determined only by this gain.

3. Interpretation of R: R is described as a measure of resistance to flow. When R

increases, the flow out of the valve decreases. This means that for a given change in
the inlet flow (input), the height of the fluid in the tank (output) will undergo a more
significant change. In simpler terms, a higher resistance will lead to a more
pronounced response in height for a given change in inlet flow

Behaviour of First-Order Processes: For first-order processes, the time taken to reach or
approach a steady state is mainly governed by the time constant ττ. Some key points related
to this are:

• When t=τ: The process output achieves 63% of the total change from its initial value
to its final steady-state value. This is often used as a measure of the system's speed
of response.
• When =3τ: The process output is 95% of its final value.
• When t=5τ: The process output is 99% of its final value.
Lag Time:

• The time it takes for a process to reach steady state is referred to as the lag time.
This is typically equivalent to five times the time constant (t=5τ).
• This lag arises from the inertia inherent in the system. Inertia, in this context, is the
system's inherent resistance to change due to its ability to store mass or energy.

Transfer Function for First-Order Process:

• The generic transfer function for a first-order process is given by:

RAMP INPUT:

For a ramp of slope m:

• Y(s) is the Laplace transform of the output response of the system.

• K is the steady-state gain of the system.
• τ is the time constant of the system.
• s is the Laplace variable.
Time Domain Response:

Output of system as time function

Graphical Representation:

The graph on the right illustrates the response of the first-order system to the ramp
input. The x-axis represents time, and the y-axis represents the output level. The curve
shows how the system's output changes with time when subjected to a ramp input.

SINUSOIDAL INPUT :