University of La Sabana

Facultad de Ingeniería

Process dynamic and control laboratory

Teacher: Nina Marcela Pérez Barrero

Chía-Cundinamarca

____________________________________________________________________________________

EVAPORATOR – Session two

Laura Catalina Marcelo – Juan José Marulanda –Andrés Felipe Mozo – Isabella Martina Ortiz

September 5th, 2019

Theoretical framework

The control of the flow is important for any equipment from the industrial point of view, taking into
account that through it you can maintain an operating regime either in refrigeration systems, separation
or heating operations as in the case of the preheater the evaporator knowing that this will ensure the
conditions desired at the inlet flow of the evaporator.

The evaporator uses a preheater system to regulate the temperature of the inlet current, the heat
exchanger ensure that the fluid enters with the necessary temperature to obtain the desired output
composition. The main dynamic characteristic to consider is the control of heat transmission,
characteristic defined as follows:

𝑄 𝑖𝑛 = 𝑄𝑜𝑢𝑡 + ℎ𝑒𝑎𝑡 𝑙𝑜𝑠𝑠

Equation 1. General heat balance. (Sánchez, 2019)

In control terms the heat loss can be considered as negligible by setting a feedback system, because the
heat amount is provided proportional to the mass, where the general equation is:

𝑄 = 𝑊 ∗ 𝐶𝑝 ∗ (𝑇𝑒 − 𝑇𝑠) = 𝑤 ∗ 𝑐𝑝 ∗ (𝑇𝑠 − 𝑇𝑒)

Equation 2. Heat balance to outlet temperature (Sánchez, 2019)

Where:

Q= Heat released

W,w= Hot and cold product flow (kg/h)

Cp,cp= Specific heat hot and cold product (kcal/kg*°C)

Te,te=Inlet temperature hot and cold product (°C)

Ts,ts=Outlet temperature hot and cold product (°C)

 Figure 1. Exchanger variables. (Levenspiel, 1993)

The conventional control system measures the outlet temperature that is defined as the controlled
variable, the manipulated variable is the fluid flow, reason why a disturbance in the fluid inlet flow could
derive in mismatches at the heating fluid temperature. This disturbance is minimized using an advanced
system like a feedforward configuration like the following:

Figure 2. Feedforward control system heat exchanger. (Levenspiel, 1993)

However, to take into account changes in temperature or fluid in the equipment this configuration may
be insufficient and must be considered other settings:

Figure 3. Feedforward and Feedback control system heat exchanger measuring heating agent. (Levenspiel, 1993)

Figure 4. Feedforward and Feedback control system heat exchanger measuring lower flow. (Raya, 2004)
MATHEMATICAL MODEL

 Theoretical Dynamic Model (Part 1 correction)

Taking into a count that the interest variable corresponds to temperature is necessary to define and
energy balance in the preheater which in this case is a tube and shell exchanger represented in the
following figure as a thermal process:

Figure 1. Variables on a thermal process

Where:

 q: input and output volumetric flow (m3/s)

 Ti= Initial temperature
 T= Output temperature

Where the relationship between the inlet and outlet temperature define the energy balance, assuming
initially volumetric flow, specific heat and density constant. To obtain a better modeling heat loss to the
environment are considered obtaining:
𝑑𝑇(𝑡)
𝑞𝜌𝐶𝑝𝑇𝑖(𝑡) − 𝑈𝐴[𝑇(𝑡) − 𝑇𝑠(𝑡)] − 𝑞𝜌𝐶𝑝𝑇(𝑡) + 𝑄 = 𝑉𝜌𝐶𝑣
𝑑𝑡
Equation 1. Theoretical Dynamic Model.

Where:

 U=Global coefficient of heat transfer, J/m2*K*s

 A= Heat transfer area, m2
 Ts(t)= Ambient temperature, °C
 Q, preheater

 Transference function

Because in this case the fluid is water it can be considered Cv≈Cp. To obtain the transference function is
necessary to consider deviation variables obtained initially by energy balance in stationary state and
substrate in equation 1 in the following way.
 𝑞𝜌𝐶𝑝𝑇𝑖 ′ (𝑡) − 𝑈𝐴[𝑇 ′ (𝑡) − 𝑇𝑠 ′(𝑡) ] − 𝑞𝜌𝐶𝑝𝑇 ′ (𝑡) = 0

𝑑𝑇(𝑡)
− 𝑞𝜌𝐶𝑝𝑇𝑖(𝑡) − 𝑈𝐴[𝑇(𝑡) − 𝑇𝑠(𝑡)] − 𝑞𝜌𝐶𝑝𝑇(𝑡) = 𝑉𝜌𝐶𝑣
𝑑𝑡
________________________________________________________

𝑞𝜌𝐶𝑝(𝑇𝑖(𝑡) − 𝑇𝑖(𝑡)′ ) − 𝑈𝐴[(𝑇(𝑡) − 𝑇 ′ (𝑡)) − (𝑇𝑠(𝑡) − 𝑇𝑠′(𝑡)] − 𝑞𝜌𝐶𝑝(𝑇(𝑡) − 𝑇 ′ (𝑡))

𝑑(𝑇(𝑡) − 𝑇 ′ (𝑡))
= 𝑉𝜌𝐶𝑣
𝑑𝑡
Equation 2. Theoretical Dynamic model in relation with stationary state balance.

Using deviation variable as:

𝑇(𝑟) = 𝑇𝑠(𝑡) − 𝑇′𝑠(𝑡)

Equation 3. Definition of deviation variable.

So, these expressions can be replaced in equation 1 to finally obtain:

𝑑𝑇(𝑡)
𝑞𝜌𝐶𝑝𝑇𝑖(𝑡) − 𝑈𝐴[𝑇(𝑡) − 𝑇𝑠(𝑡)] − 𝑞𝜌𝐶𝑝𝑇(𝑡) = 𝑉𝜌𝐶𝑝
𝑑𝑡
Equation 4. Theoretical Dynamic Model expressed in deviation variables terms.

Reorganizing the equation and dividing simulant terms the expression is converted into:
𝑑𝑇(𝑡)
𝜏 + 𝑇(𝑡) = 𝐾1𝑇𝑖(𝑡) + 𝐾2𝑇𝑠(𝑡)
𝑑𝑡
Equation 5. Theoretical Dynamic Model reorganized with tau.

Where:
𝑉𝜌𝐶𝑝
𝜏= , 𝑡ℎ𝑖𝑠 𝑣𝑎𝑙𝑢𝑒 𝑖𝑠 𝑘𝑛𝑜𝑤𝑛 𝑎𝑠 𝑡ℎ𝑒 𝑡𝑖𝑚𝑒 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 𝑎𝑛𝑑 𝑠ℎ𝑜𝑢𝑙𝑑 𝑏𝑒 𝑜𝑛 𝑡𝑖𝑚𝑒 𝑢𝑛𝑖𝑡𝑠
𝑞𝜌𝐶𝑝 + 𝑈𝐴
𝑞𝜌𝐶𝑝
𝐾1 =
𝑞𝜌𝐶𝑝 + 𝑈𝐴
𝑈𝐴
𝐾2 =
𝑞𝜌𝐶𝑝 + 𝑈𝐴
So, using Laplace transformed and reorganizing, if ambiental temperature remains constant the
transference function is the next:
𝑇(𝑠) 𝐾
=
𝑇𝑖(𝑠) 𝜏𝑠 + 1
Equation 5. Transference function with constant environmental temperature.

Where K is defined as the gain, so to analyze how it behaves in relation to the increasement in
temperature (A) it can be expressed as:
 𝑲𝑨
𝑻(𝒔) =
𝒔(𝝉𝒔 + 𝟏)

𝑻(𝒕) = 𝑻′ + 𝑲𝑨(𝟏 − 𝒆−𝒕/𝝉 )

Equation 6. Transference function to sensibility analysis.

 Step function

Taking into a count the results obtained in session one it is possible to see how the temperature has a
step function represented by an increasement in A number of grades, making that the function can be
mathematically represented as:

𝑇𝑖 (𝑡) = 𝑇𝑖 ′ 𝑡<0
𝑇𝑖 (𝑡) = 𝑇𝑖 ′ + 𝐴 𝑡>0
Equation 7. Transference function to sensibility analysis.

That also can be represented using Laplace transformed

𝐴
𝑇(𝑠) =
𝑠(𝜏𝑠 + 1)
And with inverse Laplaced transformed:
𝝉
𝑻(𝒕) = 𝑻′ + 𝑨(𝟏 − 𝒆−𝒕 )
Equation 8. Transformed expression to step function.

EXPECTED RESULTS

 Theoretical Dynamic Model (Part one)

Using python to run the dynamic model showed in equation 1 is possible to model how the
temperature will be obtaining the next behavior. Using the code that is attached in annexes.

Figure 2. Dynamic model in Python.

  Transference function

Equation 6 can be represented graphically the behavior is showed in figure 3.

Figure 3. Response of a first order process to the step change of the function of force-gain process.

How it can be observed in the equations and in the figure 2 the gain is a factor that is dependent from
different factors, so a change in one of them will have a effect over it. And as the same way as the
step function it is expected that the function has an increasement, in this case exponential and have
a time to stabilize, that is defined by the time and the time constant.

El c

Operational conditions

Mode A: Single effect

Solution feed tank 20 liters
Solution concentration 0,05%
Water feed tank 10 liters
Initial cooling water feed 5 liter/min
Feed pump 30%
Preheater temperature 70°C
Table1: Operational conditions

Referencias
Levenspiel, O. (1993). Flujo de fluidos e itercambiador de calor. Reverté.

Raya, A. (2004). Tecnología de sistemas de control. UPC.

Roger A. Ordoñez, C. A. (2012). Scielo. Obtenido de Modelado de un sistema de evaporación de multiple

efecto para la producción de panela:
https://scielo.conicyt.cl/scielo.php?script=sci_arttext&pid=S0718-07642012000600012,

Sánchez, J. A. (23 de 08 de 2019). Control de un intercambiador de calor. Obtenido de

https://es.slideshare.net/joseptchurquicondori/control-en-un-intercambiador-de-calor
ANNEXES

 Code used to dynamic model

import numpy as np

from scipy.integrate import odeint

import matplotlib.pyplot as plt

# function that returns dy/dt

def model(y,t):

Q=10

ρ=1000 #g/l

F=53 #l/min

Cp=4.186 #julios/gramo °C

Ts=23 #°C

Ti = 23

V=10 #l

u= 8.56 #julios/h·l°C

dydt = (((F* Ti)-(F* y))/V)- ((u*(y-Ts))/(ρ*V*Cp))+(Q/(ρ*V*Cp))

return dydt

# initial condition

y0 = 23

# time points

t = np.linspace(0,10)

# solve ODE

y = odeint(model,y0,t)
# plot results

plt.plot(t,y)

plt.xlabel('time')

plt.ylabel('y(t)')

plt.show()