CHML 404-01 Date: 03/11/2019

Unit Operation: 1 Instructor: Dr. Martinez

Reactors
Victoria Lombardo, Fernando Maxi, Kevin Thomas
Abstract

The objectives of this laboratory are to understand the relations between the conversion

and the residence time by comparing three different reactors. The three reactors in comparison

were CSTR, PFR and PBR. To fully understand the different types of reactors, we also changed

a few dependent variables. In the case of CSTR, we adjusted rotation speed and flow rates and in

the case of PBR, we changed flow rates as well. It is known from prior laboratory studies that the

hydraulic residence time can be different to the experimental mean residence time. This

laboratory report is focused on the conversion of a simple chemical reaction with well

characterized kinetics.

1
Table of Contents

List of Tables
…………………………………………………………………………….………..3

List of Figures…………………………………………………………………………….……….3

Background and Theory…………………………………………………………………....……...4

Results and Discussion…………………………………………………………………….……...9

Statistical Error Analysis, Regression, and Correlation of Data …………………………...……12

Conclusions……………………………………………………………………………..…….….13

Recommendations…………………………………………………………….………......……...14

References ……………………………………………………………………………..………...15

Appendix A: Original Data and Preliminary Data Analysis……………………………………..16

Appendix B: Sample of Numerical Calculations……………………………………………..….18

2
List of Tables

​ Table 1. Maximum Observed Conversion……………………………………………………...10

​Table
2. Conversion Based on Hydraulic Residence
Time……………………………………..10

Table 3. Conversion based on Mean Residence Time and Aspen

Modeling…………………...12

List of Figures

Figure 1. Conductivity profiles across reactors to steady-state operation………………………...9

Figure 2. Observed kinetic constant, 3.35 lmol​-1​s​-1​……………………………………………...11

Figure 3. 1 CSTR residence time

distribution……………………………………………………11

Figure 3.2 PFR/PBR residence time distribution………………………………………………...11

Figure 4. Conductivity test……………………………………………………………………….13

3
Background and Theory

This experiment concerns the observation of the evolution of a chemical reaction in

different types of continuous reactors known as a Continuous Stirred Tank Reactor (CSTR), Plug
Flow Reactor (PFR), and Packed Bed Reactor (PBR). The experimental plan is such that the
volume of each of the reactors is equal. In order to compare, with identical throughput time, the
respective performance is watched. The model reaction used is saponification of ethyl acetate by
sodium hydroxide with the stoichiometry is as follows:

EtOAc + NaOH → EtOH +NaOAc

A+B→R+S

Where;
A is EtOAc (ethyl acetate)
B is NaOH (sodium hydroxide)
R is EtOH (ethanol)
S is NaOAc (sodium acetate)

The kinetics of this reaction is second order and is expressed according to the rate law:

r = kC​A​C​B
Where;
k is the kinetic constant from Arrhenius law

k=k​0​e​-Ea/RT

Where;
k​0​ = 1.344×10​8​ /lmol×min
Ea is the activation energy = 42810 J/mol
R = 8.314 J/molK

The experiment includes the use of a Continuous Stirred-Tank Reactor (CSTR). CSTRs
are open systems, where material is free to enter or exit the system that operate on a steady-state
basis, where the conditions in the reactor don't change with time. Reactants are continuously
introduced into the reactor, while products are continuously removed. CSTRs are very well
mixed, so the contents have relatively uniform properties such as temperature, density, etc.

4
throughout. Also, conditions in the reactor's exit stream are the same as those inside the tank.
Most CSTRs usually are a tank of constant volume with an addition of a stirrer to mix reactants
together. Feed and exit streams are present to enter reactants and remove products. The design
equation for a CSTR is as shown:

V=F​AO​X/-r​A

Where;
F​AO​ = the initial flow rate of A entering the reactor
X = conversion, which is the moles of A reacted up to that point
-r​A​ = the rate of the reaction with respect to A

CSTRs are most commonly used in industrial processing, primarily in homogenous

liquid-phase flow reactions, where constant agitations is required. They may be used by
themselves, in series, or in a battery. Pharmaceutical industry uses them as a loop reactor.
Another application of CSTRs is the use of fermenters. Fermenters are used in many industries
such as brewing, pharmaceuticals, wastewater treatment and hydrogen processing. For a CSTR,
the matter balance is derived and written as follows:

QC​BO​ = QC​B​ + rV
Where;
Q = Q​A​ + Q​B​ and represents the global volume flow through the reactor

C​AO​ = C’​AO​×Q​A​/(Q​A​+Q​B​)
C​BO​ = C’​BO​×Q​A​/(Q​A​+Q​B​)
Where;
C​AO​ and C​BO​ are the concentrations of A and B entering the reactor
C’​AO​ and C’​BO​ are the concentrations of A and B in each of the feed circuits

QC​BO​ - kC​A​C​B​ = QC​B

C​AO -​ C​A​ = C​BO​ - C​B
C​B​ = C​BO​ – kC​B​(C​AO​ – C​BO​ + C​B​)τ

Where;
τ is the transition time V/Q through the reactor

5
 Where;
X​B​ is the conversion rate of sodium hydroxide

Another reactor this experiment includes is the use of a Plug Flow Reactor (PFR). PFRs
are wildly use in chemical industries. The advantages of PFR reactor is that it has high
volumetric unit conversion and run for long periods of time without maintenance. The control of
a Plug Flow Reactor process is a problem frequently encountered in the chemical industries. It is
challenging because of the time varying and nonlinear characteristics of the Plug Flow Reactor
process. Nonlinear model of a PFR process model is used to estimate the key unit operation
variables when using a continuous tubular reactor to reach a specified output. The steps in
developing nonlinear model for PFR are developing of mathematical model in first principal
including simulate under steady state and unsteady state condition, validation the mathematical
model through an experiment. The design equation for a Plug Flow Reactor is as shown:

The last reactor that was used in this experiment was the Packed Bed Reactor (PBR). A
PBR is the most commonly used reactor in Industry because of its efficient design. PBR’s are
tubular reactors filled with packing that is most likely coated with catalysts. These catalysts help
aid in conversion, therefore denoting these reactors as catalytic reactors. PBR’s are known for
their high conversion rates amongst all catalytic reactors. In theory, a PBR performs just like a
PFR in the sense that the material flows the reactor as a plug. As the materials flows the reactor,
all steady state conditions apply and at any given cross-section, the material will have the same
residence time. The governing equation for a PBR is in terms of Catalyst Weight because of the
direct relationship between catalyst weight and conversion.

6
 Where;

F​AO​ = the initial flow rate of A entering the reactor

X = the flow rate of A exiting the reactor
-r​A​= the rate of the reaction with respect to A and g catalyst

For a PFR and PBR, the matter balance is derived and written as follows:

All three reactors will lead to different residence times no matter the kinetics of the
reactant. Residence time is a measure of how much time a material spends in a particular
location. Therefore, the longer a molecule spends in an axial location, the longer its residence
time. The governing equation for the mean residence time is:

Where;
E(t) = the residence-time distribution function

7
 E(t)dt = fraction of fluid exiting the reactor that has spent time, t, and t+dt
t = time

*The mean residence time holds for cases when there is no changes in density and no
backmixing
The reaction rate of a flow reaction can be controlled by controlling the residence time in the
reactor. This is why finding residence times for different reactor types is very important.
There are two different models to estimate the conversion as a function of the
experimental residence time and the residence time distribution. These models are known as the
segregation model or the maximum mixedness model. For simplicity, the segregation model is
used. This model implies that the theoretical conversion, X(t), can be modeled as a lump of batch
reactors with each having one reaction time. The average conversion for any given system in the
segregation model is shown as:

Where;
X(t) is conversion as function of time for an ideal reactor
E(t) is the residence time distribution

For the case of a CSTR, the mean conversion based on the segregation model is shown as:

For the case of a PFR or PBR, the mean conversion based on the segregation model is shown as:

8
Results and Discussion

Conversion was calculated using four methods; tracking concentration of excess

hydroxide with conductivity probe, hydraulic time residence, mean residence time, and aspen
modeling. Concentration profiles were recorded using conductivity probes to measure the
concentration of hydroxide ions, and therefore error depends on the reliability of the scale.
Profiles were collected for different reactors and are presented in figure 1. Runs were carried on
till steady-state conditions, indicated by flattening of slopes.

Figure 1. Conductivity profiles across reactors to steady-state operation

As seen above, all reactors with high mixing reached maximum conversion relatively fast
compared to the single CSTR at 50 rpm. PBRs and PFRs operate under turbulent flow thereby
eliminating the need to external shaft work for mixing. While these patterns are expected, the
CSTR running at 1000 rpm should reach maximum conversion faster, it is reported 5 times faster
than CSTR at 50 rpm, the conversion should be similar and it is not. The time may be acceptable

9
however the conversion is much lower for the faster CSTR even though flow and feedstocks are
idententical. Those values are reported in table 1 for all flow reactors. It was later determined the
pumps were not calibrated properly thereby delivering disportionate flow rates of the feedstocks,
therefore in some cases, there was not enough acetate to react with hydroxide. This is the best
explanation of why the flow reactors were able to obtain much lower conductivity numbers than
the batch sample.
Table 1. Maximum Observed Conversion

A slightly different approach was used to estimate conversion from raw data and is
explained in great detail in Appendix B. It was expected for the PBRs and PFRs to have
significantly greater conversion values than CSTRs yet table 1 does not reflect this trend. This is
also clear indication that the actual flow to each reactor is different hence different outcomes.
Because inlet conditions are not identical, conversion values are not comparable. CSTR at 50
may have been richer in acetate than any of the reactors hence its high hydroxide usage. Higher
flows in the PFRs and PBRs indeed result in better conversion because of higher flows indicate
higher fluid velocity and resulting in higher turbulence for mixing.
The second approach to determine conversion using hydraulic residence time. This was
carried on CSTR and PFR given know relationships explained earlier in theory and results are
displayed in table 2. Notice how PBR is identical to say equations are applicable. Use of
hydraulic residence time indicates conversion values should be lower than those measured
however these calculated assuming pumps were delivering 0.1M of acetate and 0.15 of NaOH.

Table 2. Conversion Based on Hydraulic Residence Time

10
 Slight differences in CSTR conversion are result of temperature change. The kinetic
constant is very sensible to to temperature changes. The kinetic constant was also calculated for
PBR reactor and using linearization technique as shown in figure 2.

Figure 2. Observed kinetic constant, 3.35 lmol​-1​s​-1

Lastly, the segregation model was used to determine the conversion based on mean
residence time. First step is to determine the residence time distribution function for CSTR and
PFR. A CSTR is modeled as perfectly stir reactor, while a NaCl tracer was previously injected to
determine its function. Both are shown in figures 3.1 and 3.2 respectively.

11
Figure 3. 1 CSTR residence time distribution Figure 3.2 PFR/PBR residence time distribution

Upon multiplication with the conversion function of an ideal reactor, X(t) and integration,
the following conversion values were calculated. Similarly, aspen modeling was used to compare
and contrast values and are shown in table 3. The CSTR indeed has lower conversion than PFR
and PBR yet the values differs from the aspen simulation significantly. Nonetheless, the trend is
there.

Table 3. Conversion based on Mean Residence Time and Aspen Modeling

Reactor Mean Residence Time Aspen Modeling

CSTR 0.586 0.388

PFR 5 l/hr 0.784 0.458

PFR 10 l/hr 0.784 0.572

Statistical Error Analysis, Regression and Correlation of Data

To find the error for the conversion for the CSTR runs, the equations below can be used.

12
 The volume of the reactor (0.8 L), the error in the volume (𝞭v = 0.008 L), and the error
for the flow rates (𝞭​Q​= 1 L/h) are given to us, the 𝞭​Cb can be determined for each CSTR flow
rate. However, because conditions are not identical comparison has no tangible meaning and
hence no further calculation is carried on these end. The observed kinetic value however can be
determined as shown in figure 2. Two runs were average and the value was rounded to 3.35.
Given the Arrhenius equation and a temperature of 298 K, the calculate value was 3.75 There is
a percent difference of 10% given but overall close enough.
Similarly, the conductivity probe was contrasted against expected values as shown in
figure 4. These were calculated using the specific electrical conductivity for Na and OH ions.
These shows the probe needs to be replace as there is small concentrations of NaOH mixtures.

13
 Figure 4. Conductivity test

Conclusions

The experiment was successful in demonstrating the effects of flow rate, and rotameter

speed on reactor residence time. All reactors present a decline in concentration over time, as the

initial load of the chemical flushes out of the system. With increased flow rate, and rotameter

speed smaller residence times are observed. Residence time distributions of reactors of the same

type all presented very similar behaviors across the runs regardless of flow or mixing speed.

Results follow general trends however no true comparison can be done given conditions are not

identical for every inlets because of inaccurate flow rates readings.

Recommendations

Overall, the experiment was very effective to understanding the principle behind

residence time distribution for different reactor types. However, certain aspects of the procedure

could be improved to allow for more accurate and easier data collection. Most importantly, some

form of electronic data collection system would help ensure that enough data is collected quickly

enough to be able to extract reliable results. Additionally, a direction connection to a water

source would be helpful so that we never run into issues with running out of a water supply at

any point during the experiment. Change of probe or recalibration and add calibration of pumps’

rotameters prior to lab procedure.

14
References

1. Marina. ​Hyper-TVT: on Line Thermische VerfahrensTechnik​,

www.hyper-tvt.ethz.ch/distillation-multicomponent-gilliland.php

2. McCabe, W. L., Smith, J. C., & Harriott, P. (1993). ​Unit operations of chemical
engineering​ (5th ed.). New York ; London: McGraw-Hill.

15
Appendix A: Original Data and Preliminary Data Analysis

CSTR at 50 rpm

16
CSTR at 1000 rpm

17
PFR and PBR

18
19
Appendix B: Sample of Numerical Calculations

20
Appendix C: Miscellaneous Figures and Tables

ASPEN

Type of Reactor Conversion from Aspen

CSTR .65

PFR .66

PBR .68

21
22