1.
Modelling Process Dynamics
Orest Shardt
Part 1. Modelling Process Dynamics
• Classification of variables
• Conservation equations
• Mass balance
• Mole balance
• Energy balance
2
Modelling process dynamics
• Focus on how a process/system responds over time to changes in
conditions
• Depends on fluid mechanics, heat transfer, mass transfer, thermodynamics
• Other modules consider the details of these topics
• We will use appropriately simplified models of these phenomena
• The goal is to write ordinary differential equations that relate process
variables and their evolution over time
3
ISS docking
• “Movement in space is slow and requires patience & precision”
𝐹 𝑡 = 𝑚𝑎
𝑑𝑣
𝑎= = 𝐹(𝑡)/𝑚
𝑑𝑡
𝑑𝑥
𝑣=
𝑑𝑡
• Similar equations for translation in other dimensions and rotation
• Solve for 𝑥(𝑡) and 𝑣(𝑡) for a sequence of thrust applications 𝐹(𝑡)
• Understand dynamics -> design automatic control scheme
𝑑2 𝑥
• (Note: This is an example of a second order process: 𝐹 𝑡 = 𝑚 2)
𝑑𝑡
4
Classification of variables
• Input variable
• any variable that affects the state of a process/system
• Output variable
• any variable that describes the state of a process/system
• output variables change in response to input variables
• Classification may change depending on the context
• output of one process may be an input to the next process
5
Classification of variables
• Controlled variable (CV)
• output variable that is being controlled
• desired value is called the setpoint
• Manipulated variable (MV)
• input variable that is changed to keep a CV at its setpoint
• e.g. flow rate, valve position, motor speed
• Disturbance variable (DV)
• an input variable that affects a CV but is not manipulated
• related to the environment of a process
• e.g. cooling water temperature, ambient temperature
• Variables may be measured, unmeasured, or estimated
6
Chemical process example
• Recall previous example of temperature control
• exothermic reaction
• cooled by water flow through jacket
• CV?
• reactor temperature
• MV?
• cooling water flow rate
• DV?
• cooling water temperature www.pharmaceutical-technology.com
7
Conservation equations
• Mass, mole, energy balance equations but in their transient
forms
• keep 𝑑/𝑑𝑡
• Focus on open flow systems (control volumes) without reactions
8
Mass/mole balance
• Can be written for each component (𝑖) in a system:
𝑑𝑚𝑖
= 𝑚ሶ 𝑖,𝑖𝑛 − 𝑚ሶ 𝑖,𝑜𝑢𝑡
𝑑𝑡
𝑖𝑛𝑙𝑒𝑡𝑠 𝑜𝑢𝑡𝑙𝑒𝑡𝑠
• or for the total mass or number of moles in a system
𝑑𝑚
= 𝑚ሶ 𝑖𝑛 − 𝑚ሶ 𝑜𝑢𝑡
𝑑𝑡
inlets outlets
• Replace 𝑚 (mass) with 𝑛 (number of molecules) for mole
balance equations
9
Energy balance
• Energy balance
𝑑𝐸
= 𝑚ሶ 𝑖𝑛 ℎ𝑖𝑛 − 𝑚ሶ 𝑜𝑢𝑡 ℎ𝑜𝑢𝑡 + 𝑄ሶ 𝑖𝑛 − 𝑄ሶ 𝑜𝑢𝑡 + 𝑊ሶ 𝑖𝑛 − 𝑊ሶ 𝑜𝑢𝑡
𝑑𝑡
inlets outlets
• 𝐸: internal energy (J), 𝐸 = 𝑚𝑢 where 𝑚 is mass (kg) and 𝑢 is specific internal energy
(J/kg)
• ℎ: enthalpy (J/kg)
• heat transfer rate
𝑄ሶ 𝑜𝑢𝑡 = 𝑈𝐴 𝑇 − 𝑇𝑠
• 𝑇: system temperature
• 𝑇𝑠 : surrounding temperature
• 𝑈: heat transfer coefficient
• 𝐴: surface area for heat transfer
• Simplified treatment of enthalpy: ℎ = 𝑐𝑝 𝑇 for constant 𝑐𝑝
• Similarly for internal energy: 𝑢 = 𝑐𝑣 𝑇 for constant 𝑐𝑣
10
Mass balance example: A tank
• Consider a tank acting as a buffer
inlet flow rate: 𝑉ሶ𝑖𝑛
cross-sectional area: 𝐴
liquid level: ℎ
output flow rate: 𝑉ሶ𝑜𝑢𝑡
valve resistance: 𝑅𝑣
11
Classify variables
• Input
• 𝑉ሶ𝑖𝑛 , 𝐴, 𝑅𝑣 inlet flow rate: 𝑉ሶ𝑖𝑛
• Output
• ℎ, 𝑉ሶ𝑜𝑢𝑡
cross-sectional area: 𝐴
liquid level: ℎ
output flow rate: 𝑉ሶ𝑜𝑢𝑡
valve resistance: 𝑅𝑣
12
Conservation equation
• mass conservation
𝑑𝑚 inlet flow rate: 𝑉ሶ𝑖𝑛
= 𝑚ሶ 𝑖𝑛 − 𝑚ሶ 𝑜𝑢𝑡
𝑑𝑡
cross-sectional area: 𝐴
liquid level: ℎ
output flow rate: 𝑉ሶ𝑜𝑢𝑡
valve resistance: 𝑅𝑣
13
Valve model
• Flow rate ∼ Δ𝑃, 𝑉ሶ = Δ𝑃/𝑅𝑣 , where 𝑅𝑣 is the valve resistance
• What is Δ𝑃?
• Assume outlet at 𝑃𝑎𝑡𝑚
• Pressure at liquid level in tank also 𝑃𝑎𝑡𝑚
• Pressure at base of tank is 𝑃𝑎𝑡𝑚 + 𝜌𝑔ℎ
• Neglect losses in piping
• ∴ Δ𝑃 = 𝜌𝑔ℎ ⇒ 𝑉ሶ𝑜𝑢𝑡 = 𝜌𝑔ℎ/𝑅𝑣
• The resulting ODE is
𝑑ℎ 𝜌𝑔
𝐴 ሶ
= 𝑉𝑖𝑛 − ℎ
𝑑𝑡 𝑅𝑣
14
Question
• For a steady state with 𝑑ℎ 𝜌𝑔
0=𝐴 = 𝑉ሶ𝑖𝑛 − ℎ
• 𝐴 = 5 m2 𝑑𝑡 𝑅𝑣
L
ሶ
• 𝑉𝑖𝑛 = 400 𝜌𝑔ℎ
s • 𝑅𝑣 = =
kg 𝑉ሶ 𝑖𝑛
•𝜌= 1000 3
m
m
•𝑔= 9.8 2
s
• ℎ =2m
what is 𝑅𝑣 ?
15
Question
• If 𝑉ሶ𝑖𝑛 is suddenly changed to 450 L/s from 400 L/s, what is ℎ 𝑡 ?
• First determine the new steady state height ℎ∗ :
𝜌𝑔ℎ∗
0 = 𝑉ሶ𝑖𝑛 −
3
𝑅𝑣
L 1 𝑚 𝑘𝑔
𝑉ሶ𝑖𝑛 𝑅𝑣 450 49000 4
∗ s 1000 𝐿 𝑚 ⋅𝑠
ℎ = = = 2.25 m
𝜌𝑔 kg m
1000 3 9.8 2
m s
16
• Define ℎ′ = ℎ − ℎ∗
• Then
𝑑ℎ 𝑑 ′ 𝑑ℎ ′
= ℎ + ℎ∗ =
𝑑𝑡 𝑑𝑡 𝑑𝑡
𝑉ሶ𝑖𝑛 𝜌𝑔 ′
= − ℎ + ℎ∗
𝐴 𝐴𝑅𝑣
𝑉ሶ𝑖𝑛 𝜌𝑔ℎ∗ 𝜌𝑔ℎ′
= − −
𝐴 𝐴𝑅𝑣 𝐴𝑅𝑣
𝜌𝑔ℎ′
=−
𝐴𝑅𝑣
17
• We have a linear, first order ODE, and its solution is
′ ′
𝜌𝑔𝑡 ′
𝑡
ℎ 𝑡 = ℎ 0 exp − = ℎ 0 exp −
𝐴𝑅𝑣 𝜏
𝐴𝑅𝑣
• 𝜏= = 25 s is called the time constant.
𝜌𝑔
• Per unit change in the flow rate (input), how much does the liquid level
(output) change (at steady state)?
𝑉ሶ𝑖𝑛,2 𝑅𝑣 𝑉𝑖𝑛,1
ሶ 𝑅𝑣
Δℎ ℎ2∗ − ℎ1∗ −
𝜌𝑔 𝜌𝑔
= =
ሶ
Δ𝑉𝑖𝑛 𝑉ሶ𝑖𝑛,2 − 𝑉ሶ𝑖𝑛,1 ሶ
𝑉𝑖𝑛,2 ሶ
− 𝑉𝑖𝑛,1
𝑅𝑣 ሶ ሶ
𝑉 − 𝑉𝑖𝑛,1 𝑅𝑣
𝜌𝑔 𝑖𝑛,2
= =
𝑉ሶ𝑖𝑛,2 − 𝑉ሶ𝑖𝑛,1 𝜌𝑔
18
• Finally, substitute the definition of ℎ′:
ℎ 𝑡 = ℎ∗ + ℎ 0 − ℎ∗ exp −𝑡/𝜏
2.25 m
ℎ 𝑡
2m
Δ𝑡 = 𝜏 = 25 s
• when 𝑡 = 𝜏, 63% of the change in ℎ has occurred
• when 𝑡 = 2𝜏, 86% of the change in ℎ has occurred
• when 𝑡 = 3𝜏, 95% of the change in ℎ has occurred
• when 𝑡 = 5𝜏, 99% of the change in ℎ has occurred
19
Process gain and time constant
• Two important characteristics of a process:
• Gain
• Magnitude
• How much does the output change in response to a change in the input?
• Time constant
• Rate
• How quickly does the change happen?
20