REACTOR DESIGN AND COST FOR PRODUCING BIODIESEL FROM CANOLA OIL FOR lOMILLION GALLONS PER YEAR CONCEPTUAL PLANT A Thesis Submitted to the Graduate Faculty of the University of South Alabama in partial fulfillment of the requirements for the degree of Master of Science in Chemical Engineering by Shali Vemparala B.TECH, BRECW (Affiliated with JNT University), 2007 May 2010
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
REACTOR DESIGN AND COST FOR PRODUCING BIODIESEL FROM CANOLA OIL FOR lOMILLION GALLONS PER YEAR CONCEPTUAL PLANT
A Thesis
Submitted to the Graduate Faculty of the University of South Alabama
in partial fulfillment of the requirements for the degree of
Master of Science
in
Chemical Engineering
by Shali Vemparala
B.TECH, BRECW (Affiliated with JNT University), 2007 May 2010
UMI Number: 1484491
All rights reserved
INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
Dissertation Publishing
UMI 1484491 Copyright 2010 by ProQuest LLC.
All rights reserved. This edition of the work is protected against unauthorized copying under Title 17, United States Code.
ProQuest LLC 789 East Eisenhower Parkway
P.O. Box 1346 Ann Arbor, Ml 48106-1346
THE UNIVERSITY OF SOUTH ALABAMA COLLEGE OF ENGINEERING
REACTOR DESIGN AND COST FOR PRODUCING BIODIESEL FROM CANOLA OIL FOR 10 MILLION GALLONS PER YEAR CONCEPTUAL PLANT
BY
Shali Vemparala
A Thesis
Submitted to the Graduate Faculty of the University of South Alabama
in partial fulfillment of the requirements for the degree of
Master of Science
in
Chemical Engineering
May 2010
Date: Q - 2 ) d 4 / 10
oijerhj to
z/y/o
Approved:
Chair of Thesis Com] Jagdhish C. Dhawan
Committee Member: Dr. Srinivas Palanki k l H i ^-.
K^ang-Tii Committee Member: Dr. Kwang-Ting Hsiao CCU ._..
Chair of Department: Dr. Srinivas Balanki
Dean of the Graduate School: Dr. B. Keith Harrison
ACKNOWLEDGEMENTS
This work would not have been possible without the support and encouragement of
Dr. Jagdhish C. Dhawan, under whose supervision I chose this topic and began this
thesis. I appreciate the assistance that he has given me throughout my research at
University of South Alabama. I would like to thank Dr. Srinivas Palanki and Dr. Kuang-
Ting Hsiao for serving on my committee. I would also like to thank the rest of the
academic staff of Department of Chemical and Biomolecular Engineering for their
cooperation. I cannot end up without thanking my family and friends Mohan, Chaitanya,
Shwetha and Vijyanthi on whose constant encouragement and love, I have relied
throughout my time at the university.
ii
TABLE OF CONTENTS
Page
LIST OF TABLES vi
LIST OF FIGURES viii
LIST OF ABBREVIATIONS x
NOMENCLATURE xi
ABSTRACT xiv
CHAPTER 1: INTRODUCTION 1
1.1 Energy Consumption in World 1 1.2 Biodiesel 2 1.3 Scope and Objectives 3 1.4 Significance of the Research 4
CHAPTER 2: LITERATURE SURVEY 5
2.1 History of Biodiesel 5 2.2 Aliphatic Fatty Acid Chains 6 2.3 Transesterification of Oils 8
2.3.1 Base Catalyzed Transesterification Mechanism 9
3.1 Plug Flow Reactor (PFR) Design 20 3.2 Performance Equations for a Plug Flow Reactor (PFR) 22 3.3 MathCAD Solution to PFR Model 25
3.3.1 Effect of Methanol to Canola Oil Feed Ratio on Conversion 34
3.4 Aspen Plus Simulation of PFR 34
3.4.1 Plug Flow Reactor Size for Diameter 0.038m (1.5 inch OD) and 0.051m (2.0 inch OD) Tubes 36
3.4.2 Effect of Methanol to Canola Oil Molar Feed Ratio on Conversion 43 3.4.3 Plug Flow Reactor Length Required for 1.5 inch OD and 2.0 inch OD Tubes43 3.4.4 Pressure Drop Across Plug Flow Reactor 44 3.4.5 Plug Flow Reactor Cost 46
CHAPTER 4: CONTINUOUS STIRRED TANK REACTOR (CSTR) 48
4.1 Performance Equations Model for a Continuous Stirred Tank Reactor (CSTR)... 48 4.2 MathCAD Solution to CSTR Model 50 4.3 Aspen Plus Simulation of CSTR Model 51
4.3.1 Effect of Residence Time on Conversion 55 4.3.2 Effect of Reactor Volume on Conversion 56
4.4 Design Dimensions of the CSTR at Methanol/Oil Molar Feed Ratio of 12 56 4.5 Continuous Stirred Tank Reactor Cost 57 4.6 Continuous Stirred Tank Reactor in Series 57
4.6.1 A System of Three CSTR Biodiesel Reactors in Series 57 4.6.2 Economic Analysis of Three CSTR in Series 62 4.6.3 Design of Methanol Recovery Column 65
4.6.3.1 Column Diameter 65 4.6.3.2 Tray Hydraulics 65 4.6.3.3 Dry Pressure Drop 66 4.6.3.4 Check for Down Comer Residence Time 69 4.6.3.5 Weeping Check 69
CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH 71
5.1 Conclusions 71 5.2 Recommendations for Future Research 74
iv
REFERENCES 75
General References 76
APPENDICES 77
Appendix A: MathCAD Program for Solution of PFR for Diameter 0.038m (1.5 inch OD) and 0.051m (2.0 inch OD) Tubes 77 Appendix B: PFR Sensitivity Results from Aspen Plus Simulation for Diameter 0.038m (1.5 inch OD) and 0.051m (2.0 inch OD) Tubes 82 Appendix C: Shell/ Tube Configuration of a PFR 92 Appendix D: PFR Pressure Drop in 0.038m (1.5 inch OD) and 0.051m (2.0 inch OD) Tubes 93 Appendix E: Plug Flow Reactor Cost for Diameter 0.038m (1.5 inch OD) and 0.051m (2.0 inch OD) Tubes 95 Appendix F: MathCAD Program for a Solution to CSTR 99 Appendix G: CSTR Sensitivity Results from Aspen Plus Simulation 102 Appendix H: Sizing of Single CSTR 110 Appendix I: A Graphical Representation of Purchased Cost of Jacketed and Stirred Reactors 111 Appendix J: AspenPlus Input File for Three CSTR in Series with Pump Around System 112 Appendix K: Design of Methanol Recovery Column Calculations 127
BIOGRAPHICAL SKETCH 139
v
LIST OF TABLES
Page
Table 1: Fatty acid composition (%wt) in different types of oils 6
Table 2: Literature summary of biodiesel production from various types of oils 13
Table 3: k values for soybean oil are reported by Noureddini and Zhu, 1997 16
Table 4: k values for vegetable oil are reported by Sharma, 2008 17
Table 5: k values for palm oil are reported by Leevijit, 2004 17
Table 6: k values for soybean oil are reported by Marchetti, 2007 17
Table 7: k values for vegetable oil are reported by Komers, 2002 (based on particular data regression) 18
Table 8: Reaction kinetics relating rate constants, activation energy and Arrhenius constant (Noureddini et al., 1997) 19
Table 9: MathCAD results for diameter 0.038m (1.5 inch OD) tube 26
Table 10: MathCAD results for diameter 0.051m (2.0 inch OD) tube 27
Table 11: Summary of ASPEN results of PFR tube OD 0.038m (1.5 inch) 36
Table 12: Summary of ASPEN results of PFR tube OD 0.051m (2.0 inch) 36
Table 13: Cost of plug flow reactor for different diameters 47
Table 14: Summary of MathCAD results for different reactor volume and methanol to oil molar feed ratio: 51
Table 15: Summary of ASPEN results of CSTR volume 7m3 52
Table 16: Material balance data for three CSTR's in series to produce biodiesel from canola oil (Plant capacity: 10 million gallons/year of biodiesel product) 60
vi
Table 17: Below summarizes the results of overall conversion as a function of methanol to canola oil feed ratio 62
Table 18: Total utility cost ($/hr) as a function of methanol to canola oil mole feed ratio 62
Table 19: Cost of methanol recovery distillation column 63
Appendix
Table 20: Aspen stream results for various methanol to canola oil feed ratios as a the function of reactor length for diameter 0.038m (1.5 inch OD) 90
Table 21: Aspen stream results for various methanol to canola oil feed ratios as a the function of reactor length for diameter 0.051m (2.0 inch OD) 91
Table 22: Aspen steam results for various methanol to canola oil molar feed ratio as function reactor volume of CSTR 109
vii
LIST OF FIGURES
Page
Figure 1: World fuel consumption in year 2008 1
Figure 2: Schematic diagram of biodiesel production 3
Figure 3: Schematic diagram of biodiesel production process 11
Figure 4: Canola oil reacts with methanol in PFR to generate biodiesel and glycerol 22
Figure 5: Plug flow reactor length (m) versus conversion of canola oil to biodiesel for diameter 0.38m (1.5 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms) 28
-J Figure 6: Length (m) of plug flow reactor versus component concentrations (kmol/m ) for diameter 0.038m (1.5 inch OD) at 323K (50°C) and 4.053 x 105Pa (4atms) 29
Figure 7: Plug flow reactor length (m) versus conversion of canola oil to biodiesel for diameter 0.38m (1.5 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms) 30
Figure 8: Plug flow reactor length (m) versus conversion of canola oil to biodiesel for diameter 0.051m (2.0 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms) 31
Figure 9: Length of reactor (m) versus component concentrations (kmol/m ) for diameter 0.051m (2.0 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms) 32
Figure 10: Plug flow reactor length (m) versus conversion of canola oil to biodiesel for diameter 0.051m (2.0 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms) 33
Figure 11: ASPEN simulation of PFR diameter-0.038m (1.5 inch) and 0.051m (2.0 inch) OD 35
Figure 12: Plug flow reactor residence time versus conversion of oil to ester for diameter 0.038m (1.5 inch OD) at 323K (50°C) and 4.053 x 105Pa (4atms) 37
Figure 13: Length of plug flow reactor versus component flow rate for diameter 0.038m( 1.5 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms) 38
viii
Figure 14 : Plug flow reactor residence time versus conversion of oil to ester for diameter 0.38m (1.5 inch OD), 323K (50°C) and 4.053 x 105 Pa (4atms) 39
Figure 15: Plug flow reactor residence time versus conversion of canola oil to biodiesel for diameter 0.051m (2.0 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms) 40
Figure 16: Length of plug flow reactor versus component flow rate for diameter 0.038m (2.0 inch OD) at 323K (50°C) and 4.053 x 105Pa (4atms) 41
Figure 17: Plug flow reactor residence time versus conversion of canola oil to biodiesel for diameter 0.051m (2.0 inch OD), at 323K (50°C) and 4.053 x 105 Pa (4atms) 42
Figure 18: 180 degree tube bends of a PFR 44
Figure 19: Canola oil reacts with methanol in a CSTR to produce biodiesel and glycerol 48
Figure 20: ASPEN Simulation of CSTR volume 7m3 52
Figure 21: Continuous stirred tank reactor residence time versus conversion of canola oil to biodiesel for methanol to oil ratio 3, 6, 9, 12 varying along the volume of reactor at 323K (50°C) and 4.053 x 105Pa (4tams) 53
Figure 22: Volume of Continuous stirred tank reactor versus component flow rate for methanol to oil ratio 12 at 323K (50°C) and 4.053 x 105Pa (4tams) 54
Figure 23: Continuous stirred tank reactor residence time versus conversion of canola oil to biodiesel for methanol to oil ratio 12 varying along the volume of reactor at 323K (50°C) and 4.053 x 105Pa (4tams) 55
Figure 24: ASPEN simulation of three CSTR's in series 59
Figure 25: Methanol to oil mole ratio versus total cost 64
Figure 26: Schematic diagram of distillation column 65
Appendix
Figure 27: Conversion of canola oil to biodiesel versus length of reactor for diameter 0.038m (1.5 inch OD) tube 79
Figure 28: Conversion of canola oil to biodiesel versus length of reactor for diameter 0.051m (2.0 inch OD) tube 81
Figure 29: Purchased cost of jacketed and stirred reactors (Peters et al., 2003) 111
ix
LIST OF ABBREVIATIONS
FAME Fatty Acid Methyl Esters
PFR Plug Flow Reactor
CSTR Constant Stirred Tank Reactor
LHSV Liquid Hourly Space Velocity
x
NOMENCLATURE
English Letters
Symbol Description Units
A Arrhenius constant L/mol*min
Ar Area of the reactor m
Area Area of plug flow reactor m2
BD Biodiesel unitless
C concentration mol/m
CONV Conversion of canola oil to biodiesel unitless
Cost Cost of Plug Flow Reactor (2008) $
DG Diglyceride unitless
DSheii Shell inside diameter m
E Activation Energy J/mol
F Friction factor unitless
F flow rate mol/s
Fm Material Factor for a 316 stainless steel pipe unitless
GY Glycerol unitless
ID Tube inside diameter m
L Assumed length of a shell m
xi
Lreactor Total length of a PFR m
J TG, DG, MG, A, BD, GY. Unitless
k Rate constant m3/mol.min
Kf Frictional loss in a 180 degree bend unitless
mins minutes minutes
MG Monoglyceride unitless
ME Methanol unitless
NtUbes Total number of tubes in a PFR unitless
OD Tube out side diametr m
APrcactor Pressure drop in PFR Pa
APtotal Pressure drop in all the tubes of a PFR Pa
r Triglyceride of canola oil unitless
R Rate of reaction min"1
R' Alcohol group unitless
TG Triglyceride unitless
V Volume of reactor m3
Wj Conversion of canola oil to biodiesel for a methanol unitless to oil feed ratio of 3
Xj Conversion of canola oil to biodiesel for a methanol unitless to oil feed ratio of 6
Yj Conversion of canola oil to biodiesel for a methanol unitless to oil feed ratio of 9
Zj Conversion of canola oil to biodiesel for a methanol unitless to oil feed ratio of 12
xii
Greek Letters
Symbol Description Units
P Density of liquid in side a tube kg/ m3
v volumetric flow rate m /s
v velocity in a tube m/s
0 residence time s
r residence time s
xiii
ABSTRACT
Vemparala, Shali, M.S., University of South Alabama, May 2010, Reactor Design and Cost for Producing Biodiesel from Canola Oil for 10 Million Gallons Per Year Conceptual Plant. Chair of Committee: Dr. Jagdhish C. Dhawan.
Biodiesel can be produced from many natural renewable sources (vegetable oils,
animal fats, algae etc). The present study concentrates on production of biodiesel using
canola oil and methanol as reactants at 298K (25°C) and 4.053 x 105 Pa (4atms) in the
presence of sodium methylate acting as catalyst. A plug flow (PFR) and continuous
stirred tank (CSTR) reactors are designed using the rate expressions available from the
literature. Reactor performance was evaluated with respect to conversion versus reactor
volume and the effect of methanol to oil molar feed ratio on conversion at 323K (50°C)
and 4.053 x 105 Pa (4atms) was also evaluated. A plug flow reactor of 1.50 inch OD
requires a total length of 2580 meters. The reactor length can be decreased to 1140 meters
when the diameter is 2.0 inch OD. The reactor pressure drop is significantly high and
ranges from 1.317 x 106 Pa (13atms) to 3.445 x 106 Pa (34atms) depending upon the
reactor length. A single CSTR of 7m volume provides 90.3% conversion. However, if
three 5m3 volume CSTR reactors are used in series, an overall conversion of 99.9% can
be achieved. Three CSTR reactors in series with a pump around system for thorough
mixing are recommended. A conversion of 99.9% eliminates the product purification
step to recover the un-reacted material from the product stream.
xiv
CHAPTER 1: INTRODUCTION
1.1 Energy Consumption in World
Energy is an integrated part of human life. Energy can be obtained from two different
sources, renewable and non renewable. The primary source of energy has been from non
renewable sources which are fossil fuels. These fuels include coal, oil and natural gas.
Fuel consumption in the world is shown in the Figure 1 (BP Statistical Review, 2008).
Hydro ^^.Nuclem
6.50%
Figurel: World fuel consumption in year 2008.
The fossil fuel reserves in the world are so unevenly distributed that many countries
have to depend on other countries for their requirements to be fulfilled. The recovery and
1
processing of fossil fuels is known to damage the environment we live in. When fossil
fuels undergo combustion acids like carbonic, sulfuric and nitric are released, which are
the main cause of acid rains. Small amounts of radioactive materials like uranium and
thorium are also present in fossil fuels; hence these together harm the environment when
released into the atmosphere (Gabbard, 1993). All these factors necessitate continued
search and sustainable development of renewable energy sources such as biofuels that are
environmentally friendly.
1.2 Biodiesel
Energy derived from biological sources is Bio-energy. Bio-energy sources are
Table 10: MathCAD results for diameter 0.051m (2.0 inch OD) tube.
Length of reactor MeOH/Oil- 3 MeOH/Oil- 6 MeOH/Oil- 9 MeOH/Oil-12
Conversion
(m) (%)
200 20.80 36.77 49.29 59.26
400 34.65 55.70 69.78 79.57
600 43.59 65.80 79.54 88.24
800 49.34 71.70 85.01 92.65
1000 53.06 75.50 88.44 95.12
1140 54.89 77.45 90.14 96.21
1200 55.52 78.17 90.75 96.57
27
Figure 5: Plug flow reactor length (m) versus conversion of canola oil to biodiesel for diameter 0.38m (1.5 inch OD) at 323K (50°C) and 4.053 x 10s Pa (4atms).
28
12
10
TGj 8
DQ
MGj 6
MEj
BDj 4
GY;
2
0
Figure 6: Length (m) of plug flow reactor versus component concentrations (kmol/m3) for diameter 0.038m (1.5 inch OD) at 323K (50°C) and 4.053 x 105Pa
(4atms).
29
Li
Figure 7: Plug flow reactor length (m) versus conversion of canola oil to biodiesel for diameter 0.38m (1.5 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms).
30
Figure 8: Plug flow reactor length (m) versus conversion of canola oil to biodiesel for diameter 0.051m (2.0 inch OD) at 323K (50°C) and 4.053 x 10s Pa (4atms).
31
Figure 9: Length of reactor (m) versus component concentrations (kmol/m ) for diameter 0.051m (2.0 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms).
32
Figure 10: Plug flow reactor length (m) versus conversion of canola oil to biodiesel for diameter 0.051m (2.0 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms).
33
3.3.1 Effect of Methanol to Canola Oil Feed Ratio on Conversion
The reversible transesterification reaction requires high concentration of methanol for
the production of Biodiesel. Desired methanol to oil ratio is 3 but at this concentration
separation is not possible, hence methanol to oil ratio is increased to 6, 9 and 12 but at
higher concentrations separation of methanol from biodiesel is tedious. For the practical
purposes methanol concentration is maintained at optimum levels i.e 12. In the reactor,
the concentration of compounds changes with the length of the reactor as the
transesterification reaction progresses. Reactants (TG, ME, DG and MG) concentrations
decrease along the length of reactor. Products (BD and GY) concentration increase along
the length of reactor. Figures 6 and 9 are the concentration profile for methanol/oil feed
ratio-12. From the Figures 5 and 8, it is observed that as the molar ratio of methanol is
increased from 3 to 12, the conversion of canola oil changes from 54% to 97%.
3.4 Aspen Plus Simulation of PFR
The performance of a PFR was evaluated using, an AspenPlus simulation under the
following operating conditions:
Reactor feed: 298K (25°C) and 4.050 x 105Pa (4atms).
Reaction temperature as 323K (50°C) and pressure 4.050 x 105Pa (4atms)
Reactor Tube (inside diameter): 0.038m (1.5 inch) OD and 0.051m (2.0 inch) OD.
Reactor length: 2580m and 1140m.
Sensitivity Blockl: vary methanol to canola oil mole ration as 3, 6, 9, 10 and 12.
Sensitivity Block2: vary reactor inside tube diameter as 0.03 and 0.045 meters.
34
A schematic of the process under consideration is shown in Figure 11. The details of
Aspen input file are given in Appendix (B).
Figure 11: ASPEN simulation of PFR diameter-0.038m (1.5 inch) and 0.051m (2.0 inch) OD.
35
3.4.1 Plug Flow Reactor Size for Diameter 0.038m (1.5 inch OD) and 0.051m
(2.0 inch OD) Tubes
A summary of Aspen simulation results for reactor tube OD of 0.038m (1.5 inch) and
0.05 lm (2.0 inch) is given in Table 11:
Table 11: Summary of ASPEN results of PFR tube OD 0.038m (1.5 inch).
Lengt Ratio CONV T BD DG MG ME GY TG h OUT OUT OUT OUT OUT
Graphical representations of results are shown in Figures 12 through 17.
36
Residence time vs Conversion
—MeOH/Oi l -3 MeOH/Oil-12
• MeOH/Oil-6 • Length of reactor
MeOH/Oil-9
s 'vi > s o U • mm o « "3 c « U
100
90 80
g > 0 f e o 0> 3 50
a 4 0
30 20 10
0
s-o u « e*
61) C -J
10 15 20
Residence time (minutes)
Figure 12: Plug flow reactor residence time versus conversion of oil to ester for diameter 0.038m (1.5 inch OD) at 323K (50°C) and 4.053 x 105Pa (4atms).
37
Length of reactor vs Component Flow Rates —•—BDOUT - " - D G O U T MGOUT -*~GYOUT MEOUT —•—TGOUT
O O &
cs pfj £ o
80.000
70.000
60.000
50.000
40.000
30.000
20.000
10.000
0.000 500
•O 0 O O o o o — o o o o o o o o
-X—*- -K X K X X X X X X *
1500 — • — * — • —
2500 Length (m)
— m —
3500 — » —
4500
Figure 13: Length of plug flow reactor versus component flow rate for diameter 0.038m(1.5 inch OD) at 323K (50°C) and 4.053 x 105 Pa (4atms).
38
Figure 14 : Plug flow reactor residence time versus conversion of oil to ester for diameter 0.38m (1.5 inch OD), 323K (50°C) and 4.053 x 105 Pa (4atms).
39
Residence time vs Conversion • MeOH/Oil-3 MeOH/Oil-12
MeOH/Oil-6 Length of reactor
MeOH/Oil-9
10 20 30 40 Residence time (minutes)
m
5000 4500 4000 3500 3000 2500 2000 1500 1000
500
9 -w w «
o xi % c V J
50
Figure 15: Plug flow reactor residence time versus conversion of canola oil to biodiesel for diameter 0.051m (2.0 inch OD) at 323K (50°C) and 4.053 x 105 Pa
(4atms).
40
Length of reactor vs Component Flow Rates —•—BDOUT — D G O U T MGOUT —*— GYOUT -e -MEOUT - • - T G O U T
80.000
70.000
Ja 60.000
50.000
,40.000
o o s a
30.000
20.000
0> -M «
£ E 10.000
0.000 ^ 500
n o—&- o o o o o o—o o o o o p
-X— -X X X X X X X X * »—«.... » — m—e~e
1500 2500 3500 4500 Length (m)
Figure 16: Length of plug flow reactor versus component flow rate for diameter 0.038m (2.0 inch OD) at 323K (50°C) and 4.053 x 105Pa (4atms).
Figure 17: Plug flow reactor residence time versus conversion of canola oil to biodiesel for diameter 0.051m (2.0 inch OD), at 323K (50°C) and 4.053 x 10s Pa
(4atms).
42
3.4.2 Effect of Methanol to Canola Oil Molar Feed Ratio on Conversion
Figures 13 and 16 are the flow rate profiles of TG, DG, MG, ME, BD, GY varying
along the length of the reactor for the two reactors with dimensions of 0.038 m OD, 2580
m long and 0.051 m OD, 1140 m long. In reactor flow rates of compounds vary along the
length of reactor as the transesterification reactions progresses with time. As expected
reactants (TG, ME, DG and MG) flow rates decrease along the length of reactor, whereas
products (BD and GY) flow rates increase along the length of reactor. Residence time
varies with conversion of oil to biodiesel for different methanol to oil molar feed ratios
are shown in Figures 12 and 15. From Figures 12 and 15 it can be concluded that for a
methanol to oil feed ratio of 12, 90.83% overall conversion is achieved under the
residence time of 9.3 minutes.
3.4.3 Plug Flow Reactor Length Required for 1.5 inch OD and 2.0 inch OD Tubes
Two different AspenPlus simulations were carried out. In the first case PFR input
parameters were: 0.38 m (1.5 inch) OD tube diameter, 2580 meters length and methanol
to canola oil molar feed ratio ofl2. In the second case input parameters are: 0.51 m (2.0
inch) OD tube diameter and 1140 meters long and methanol to canola oil feed ratio of 12.
Under these conditions, an overall conversion of canola oil was 90.88% in both the cases.
Then calculations for shell diameter:
N t u h e s = ^ - (3.31)
L = 4m
N t u b e S = 645 for 0.038m (1.5 inch) OD tube
N t u b e s = 285 for 0.051m (2.0 inch) OD tube
43
Dshe„ = 1.25 * Tubepitch * ^Nluhes
DSheii =3.225 m (10.6ft) for 0.038 m (1.5 inch) OD tube
Dsheii =2.144 m (7.0ft) for 0.051 m (2.0 inch) OD tube
Detailed calculations can be found in Appendix C.
3.4.4 Pressure Drop Across Plug Flow Reactor
For a 90.8% conversion, reactor length was found to be 2580 m for a 0.038 m (1.5
inch OD) tube and at the same value of conversion; a 2.0 inch OD reactor should be 1140
meters long. Assuming a shell-tube configuration, for a shell length of 4-meters, an S-
type tube length configuration would require Nbends of 180 degree tube bends as shown
below:
Figure 18: 180 degree tube bends of a PFR.
Pressure drop can be calculated from the following equations (Peters and Timmerhaus,
2003):
44
Pressure drop across straight length tube
2 a jjj r j
AP = — - ^SL.) (3.33) 2 D
Pressure drop across 180 degree bend
2
AP = p*^-*(Kf) (3.34)
Total pressure drop per tube length with one 180 degree bend 2 A jjs /* jJ> J
AP = — J - H2«2!- + K f ) (3.35) 2 D
This pressure drop is for one 4m long one-bend tube
NB e n d s=N l u b e s- l (3.36)
APlnlal=AP*Nliemh (3.37)
From the equation 3.37, NreactorCan be assumed and the total reactor pressure drop will
be:
^ a c l nr=^P l„ l a l*N r e a c , o r (3.38)
Where:
Lreactor ~ Total length of PFR (m)
p - Density (kg/m )
D - Diameter of tube (m)
/ - Friction factor
ve - Velocity in a tube (m/s)
K f - Frictional loss in a 180 degree bend
AP - Pressure drop (Pa)
45
Neends - Total number of 180 degree bends in a PFR
Ntubes - Total number of tubes in a PFR
Nreactor- Total number of reactors required
A P r e a c t o r - Pressure drop in a PFR (Pa)
A P t o t a i - Pressure drop in all tube of a PFR (Pa)
Detailed pressure drop calculations are summarized in Appendix D. For case 1 (0.038
m (1.5 inch OD) and L = 2580 m) the total pressure drop is found to be 3.55 x 106 Pa (35
atms). In case 2 (0.051 m (2.0 inch OD) and L= 1140), the total pressure reduces to 1.32
x 106 Pa (13 atms). It should be noted that the pressure drop can be further reduced by
increasing the tube diameter. However, tube diameters larger than 0.051 meters (2
inches) may not provide the uniform concentration in axial and radial direction of the
reactor.
The cost for the year 2008 of the designed PFR is estimated from the following
correlation (Seider et al., 2004):
3.4.5 Plug Flow Reactor Cost
Cost = Exp(11.0545 - 0.9228 * In (Area) + 0.0979(ln (Area)2)) * Fh M (3.39)
(3.40) 4
(3.41)
The results are shown in Table 13
46
Table 13: Cost of Plug flow reactor for different diameters.
Case Tube OD Reactor Gear Pump Reactor Total Length cost cost cost
1 0.038 m (1.5 inch) 2580 meters $22000 $97,000 $166,700
2 0.051 m (2.0 inch) 1140 meters $9,400 $169,000 $250,000
As the diameter is increased from 0.038 m (1.5 inch) OD to 0.051 m (2.0 inch) OD the
cost of a PFR is increased by 66.7%. Details of plug flow reactor cost calculations are
summarized in Appendix E.
47
CHAPTER 4: CONTINUOUS STIRRED TANK REACTOR (CSTR)
The continuous stirred tank reactor (CSTR) is also known as an ideal reactor or an
agitated tank reactor. This reactor can be used for liquids and slurries as reactants.
4.1 Performance Equations Model for a Continuous Stirred Tank Reactor (CSTR)
A general material balance relationship for liquid phase chemical reactions a CSTR is
Graphical representations of results are shown in Figures 21 to 23 and analysis of
these results is presented below in 4.3.1 and 4.3.2.
Details of all simulation results are given in Appendix G.
Figure 21: Continuous stirred tank reactor residence time versus conversion of canola oil to biodiesel for methanol to oil ratio 3, 6, 9,12 varying along the volume of
Figure 22: Volume of Continuous stirred tank reactor versus component flow rate for methanol to oil ratio 12 at 323K (50°C) and 4.053 x 105Pa (4tams).
54
Figure 23: Continuous stirred tank reactor residence time versus conversion of canola oil to biodiesel for methanol to oil ratio 12 varying along the volume of
reactor at 323K (50°C) and 4.053 x 10sPa (4tams).
4.3.1 Effect of Residence Time on Conversion
These results show that if the reactions are carried out in a single reactor, conversion
depends upon the methanol to canola oil molar feed ratio under the same residence time
in the range of 40 to 45 minutes. As it can be seen from Figure 21, at a given reactant
feed ratio, increasing residence time beyond 40 minutes, there is no significant increase
in the overall conversion. Over 90% conversion can be achieved under 40 minutes of
residence time, if the methanol to oil feed ratio is increased to 12.
55
4.3.2 Effect of Reactor Volume on Conversion
At reactant feed ratio of 12, a single reactor with 7 m3 volume would be required for
90% conversion. Increasing the reactor volume from 4 to 10 m3, there is no significant
increase the production of bio diesel as shown in Figure 22. The reactor performance
with respect to residence time, volume and conversion is shown in Figure 23.
4.4 Design Dimensions of the CSTR at Methanol/Oil Molar Feed Ratio of 12
Input= Fresh feed
Volumetric flow rate = (Fresh feed)/ (pavg)
Assuming
Design Residence time =1.5* Reactor residence time
Design Volume of reactor = Volumetric flow rate* Residence time
Area of reactor = 3.14* (diameter)2 / 4
Height of reactor = Design Volume of reactor / Area
Actual feed volume = Volumetric flow rate*Reactor residence time
Reactor Design Volume = Area of reactor * Height of reactor
Percent over design = Reactor Design Volume/ Actual feed volume
Percent over reactor volume (based on rules of thumb). Calculations of CSTR sizing
can be found in Appendix design for CSTR of volume 7 m3 is 1.5. The author
recommends a 50% over design in H.
56
4.5 Continuous Stirred Tank Reactor Cost
Using M and S index (Chemical Engineering Magazine, 2009) and Figure H, cost of
single CSTR with material of construction as stainless steel is $77,880.
Details of utility cost of CSTR can be found in Appendix I. This represents over
324% lower reactor cost when compared with the PFR cost of $250,000.
4.6 Continuous Stirred Tank Reactor in Series
In a CSTR reactor design, it is important that reactor contents are thoroughly mixed
since perfect mixing is the fundamental basis of reactor performance for desired
conversion. Therefore reactor residence time should be greater than the mechanical
mixing time. Typically, the reactor residence time should be at least 5 times the
mechanical mixing time. It is economically beneficial to operate several CSTR units in
series. The multiple reactors in series require a smaller reactor volume and offer higher
overall conversion. In the production of biodiesel, reactants, catalyst and products are all
in liquid phase, a simple pump around configuration ensures vigorous mixing. This
approach for mixing offers better economics since pumping energy required is much less
than the energy requirements for impeller type mechanical mixing.
4.6.1 A System of Three CSTR Biodiesel Reactors in Series
Figure 24 shows an Aspen simulation diagram of three reactors in series. The material
balance results are presented in Table 16. Each reactor volume is 3 m3. The feed to the
first reactor consists of canola oil, methanol at 323 K (50 °C) and 4 x 105 Pa (4atms) (abs)
pressure. The effluent stream from each reactor is subjected to a pump around system
57
where 90 % of the reactor effluent stream is recycled back to the reactor. The recycle
stream is cooled to 313 K (40 °C) before entering the first reactor. As a result of inter
cooling; feed to the second reactor is at 322 K (49 °C). Under adiabatic conditions, the
temperature in the second reactor increases from 322 K to 343 K (49 °C to 70 °C). In the
third reactor, temperature rises from 343 K to 382 K (70 °C to 79 °C). It should be
pointed out that the reaction rate constants used in this study were determined
experimentally at 333 K (60 °C). If these rate constants are assumed to be valid up to
353 K (80 °C), then no inter cooling for the second and the third reactor will is required.
The Aspen simulation of methanol recovery from reactor effluent stream using direct
distillation reveals that sodium methoxide catalyst will boil-off with methanol in the
overhead stream to be recycled. This is mainly due to lack of liquid phase dissociation of
methanol- sodium methylate system. Thus reactor effluent may be cooled and separated
into two liquid phases. The recovery of various compounds for each liquid phase can be
accomplished by approaches similar to those found in literature (Apostolakou et al., 2009
and Haas et al., 2006).
58
Figure 24: ASPEN simulation of three CSTR's in series.
59
Table 16: Material balance data for three CSTR's in series to produce biodiesel from canola oil (Plant capacity: 10 million gallons/year of biodiesel product).
Stream Name TG Methanol Rl-IN Rl-OUT Rl-RCY R2-IN R2-OUT
Phase Liquid Liquid Liquid Liquid Liquid Liquid Liquid Mass Flow
Condenser water cost $6.12 $10.08 $16.20 $23.04 $29.88
Power cost for Pump -1 $1.44 $1.44 $1.44 $1.44 $1.80 Power cost for Pump -1 $1.44 $1.44 $1.44 $1.44 $1.80 Power cost for Pump -1 $1.44 $1.44 $1.44 $1.44 $1.80
In order to carryout an economic analysis of incremental change in methanol feed
ratio to incremental increase in canola oil conversion, the equipment cost of reactors,
pumps and heat exchanger will not have any impact on the outcome of results since these
costs will be the same for all cases. Therefore the total cost can be expressed as the sum
of the Methanol distillation column cost (Fixed cost) plus the utility cost (variable cost).
The results of total cost form the data given in Table 18 and 19 are presented in Figure 25
as shown below:
63
$86.3
Cost ($/hr) v's Biodiesel Product Purity as a function of Methanol to Oil mole Ratio
Utility -Conversion —A— eqpcost Total-Cost
$69.0
t" $51.8
U) o $34.5
$17.3
$0.0
$17.1 $T2.6
• $19.8
- Jr$30.3 • $37.2
* 62.4%, -A-5t9
• 86.7%r nA-7.0 • 95.5%r • 98.1%, 9 12
Methanol to Oil mole Ratio
12.3
99.0% 15
Figure 25: Methanol to oil mole ratio versus total cost.
64
4.6.3 Design of Methanol Recovery Column
Methanol recovery column was designed using following rigorous correlations (Wankat, 2007):
Figure 26: Schematic diagram of distillation column.
4.6.3.1 Column Diameter: Fair Flow parameter (Csb) is used determine column
diameter at top and bottom column conditions. The calculated results are
Djop = 1.361 m
DBottom= 1.614 m
Use 1.614 m diameter column
4.6.3.2 Tray Hydraulics: To carry out tray hydraulic calculations, the following tray
layout was assumed:
Tray Thickness = T t r a y = 0.0019 m (0.078 inch)
Hole diameter = 0.0047 m (3/16 inch)
Column Area = AT
65
Down comer Area = Ad = 12% of AT
Net Area = An = 88% of AT (Vapor Flow Area)
Active Area = Aa = (AT - 2Ad)
Hole Area = Ah = 10% AT
Weir Length = lw = 77% of Diameter (Top or Bottom)
Using the above tray specifications, the following results are obtained:
i) Bottom column area = 2.055 m ,
ii) Net tray area = 1.808 m2,
iii) Down comer flow area = 0.247 m2,
iv) Active tray area = 1.561 m2,
v) Tray hole area = 0.156 m2,
vi)Weir Length = 1.243 m
Pressure drop across the tray consists of pressure drop caused by the various
hydrodynamic heads which include dry pressure drop (hdC) plus static head by the weir
( h w e i r ) plus head caused by the liquid crest (hcrest) plus static head of the liquid under the
down comer (hdU).
hdc W Dry + Kveir+ ^crest + hdu
Calculation results of each of these terms are presented below. Detailed calculations are
presented in Appendix K.
4.6.3.3 Dry Pressure Drop: It should be noted that all pressure drop results in inches (m)
of water.
66
Calculate orifice coefficient (C0) for a tray hole using the following Kesseler and
Wankat correlation:
C 0 := 0 . 8 5 0 3 2 - 0 . 0 4 2 3 1 -dhole N
V tray ) 0.0017954-
r dhole V
v W ) C0 = 0.759
hAP_Dry := 0.003-' f t V
I V s -
P VBottom jb
ft3
P water
v PLBottom .
W ) " •in
^APDry = 0.025 m
hcrest := 0.092-F wear
C r A "13 Ijrr gal
V min J w Bottom
•in
Merest = 0.022 m
Assume negligible gradient pressure drop,
hgrad : = 0
hdu := 0.56-
gal V min
449 Adu
ft2
•in
- 3 hdu = 7.641 x 10 m
67
hweir = 0.051 m
hdc : = hAP Dry + hweir + hcrest + Hju
hdc = 0.106m
hdc •\ic_aerated
<Pdc
hdc_aerated = 0.211m
This value is less than 0.61 m (24 inch) tray spacing. There should be no problem for
vapor to flow to the above tray.
Where:
Lg = Liquid Flow Rate (L/min)
Fweir = Factor accounting for a curvature of the column wall in the down comer
hgrad= The liquid gradient across the tray (m)
hdc= Total Head of clear liquid in the down comer (m)
hdc_areated = Head of clear liquid in the down comer due to aeration (m)
hdu = Frictional loss due to flow in the down comer and under the down comer onto the
tray (m)
hweir — static head by the weir (m)
hcrest= static head by the liquid crest (m)
(3 = Area of hole to active area (unit less)
Adu = Flow area under the down comer (m2)
68
4.6.3.4 Check for Down Comer Residence Time: According to the rule of thumbs, a
minimum down comer residence time should be at least 3 seconds. Based on foaming
tendency of the liquid, this residence time ranges from 3 seconds (low foaming) to 7
seconds (very high foaming) liquids (Kister, 1992).
(Ad Bottom "PL Bottom •hdc)
^Bottom + e e tdc : =
t<ic = 3.695 s
Where
ee = entrainment
tdc= Down comer residence time
This value is greater than the minimum residence time of 3 seconds, there should be
no problem to maintain the proper liquid flow.
4.6.3.5 Weeping Check: Weeping relates to the direct flow of the liquid through sieves
without maintaining a proper vapor liquid contact. Excessive weeping is estimated using
Fair's (Wankat, 2007) correlation for a surface tension head(ha). f
ho := 0.04in-
dyne
V cm J
P L B o t t o m
lb
\ dhole
V in ;
- 3 ha = 1.299 X 10 m
69
Excessive weeping is defined by equality that the term (h^p £)ry+h a) should be
greater than the value of the term (.10392+0.25119x-0.021675x 2)-
h A P Qry+h a > 0.10392+0.25119x-0.021675x 2
where x is deined as: h Weir+hcrest+hgrad t* ien
h A P _ D r y + h a = i - 0 4 7
0.10392+0.25119x-0.021675x 2 = 0.645
These calculations show that the inequality is satisfied. Therefore, Weeping should
not be a problem. All of the design calculations are presented in Appendix K.
70
CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH
5.1 Conclusions
In this study, reactions of canola oil with methanol to produce biodiesel were
examined in a PFR as well as in a CSTR reactors. Based on kinetic rate expression
available from the literature, a Plug Flow Reactor (PFR) was designed to determine the
reactor length as a function of conversion of canola oil. Two different reactor tube
diameters were investigated in the plug flow reactor system.
For a Plug Flow Reactor with 0.038m (1.5 inch) OD tube, nearly 91% conversion was
achieved in a 2580m long reactor at methanol to oil molar feed ratio of 12 ( as compared
to the stoichiometric ratio of 3) with 9.3 minutes of reactor residence. The total pressure
drop for this reactor was found to be around 35.463 x 105 Pa(35atms). The requirement
of a long tubular reactor is mainly due to the desired conversion of canola oil beyond
90%.Gear pumps must be used to provide the pressure of the feed system to overcome
the prevailing pressure drop.
According to the information on gear pumps from reference (Seider et al., 2004),
"Although gear pumps can be designed to operate over a wide range of flow rates and
discharge pressure, typical ranges are 10 - 1500 gpm, and up to 200 psi for high viscosity
fluids.", for a 0.038 m (1.5 inch) OD diameter PFR, the pressure drop requirements can
71
not be met with a single off-the-shelf gear pump. If two plug flow reactors of 0.038 m
(1.5 inch) OD tube are used, the developed pressure drop of 17.97 x 105 Pa (17.732atms)
per reactor would still be higher and exceeds the pressure developed by the standard size
gear pump, hence more than one gear pumps will be needed. The total estimated cost for
this PFR (diameter of 0.038 m (1.5 inch) OD tube and 2580 m length) using two gear
pumps will be $166,700 (2008). For a PFR with 0.051m (2.0 inch) OD tube, the reactor
tube length decreases to 1140 meters under the same conditions as in for PFR of diameter
0.038 m (1.5 inch) OD tube. The resulting pressure of 13.17 x 105Pa (13atms) can be
attained with a single gear pump unit. The total estimated cost of 1140m long and 2.0
inch OD PFR will be $250,000 (2008). Therefore it is concluded to use a 0.051 m (2
inch) OD tube.
In the case of a CSTR, reactor volume was determined as a function of conversion.
Under each type of reactor design study, methanol to canola oil molar feed ratio was used
as an independent parameter. The reactant flow rates were based on producing 10
million gallons of biodiesel per year. For a 7m3 volume CSTR and methanol to oil molar
feed ratio of 12 yields 90.3% conversion with 41 minutes of reactor residence time.
The estimated cost of a 316-s.s CSTR with 50% capacity over design is found to be
$77,880. The capacity over design provides 30 minutes of hold up volume to
accommodate startup and process upsets.
The CSTR reactor design also included the evaluation of three CSTR reactors in series
with a pump around loop to provide vigorous mixing. The reactor effluents were
separated by distilling the un-reacted methanol for recycle. When the value of 'R'
(defined as mole feed ratio of methanol to canola oil) is between 3 (stoichiometric feed
72
ratio) and 15, an additional distillation column will be required for product purification in
order to meet ASTM- D6751 (biodiesel product standard) which requires > 99 wt% of
biodiesel product purity. The boiling point of biodiesel is approximately 598K (325 °C).
Biodiesel is thermally decomposed when temperature exceeds 523K (250 °C). The
thermal decomposition temperature for glycerol is about 423K (150 °C). In order to keep
low distillation temperature, purification column must be operated under vacuum. Based
on Aspen simulation, 10 milli-bar of vacuum would be required to operate the distillation
column below the decomposition temperatures of these products. If the value of 'R' is
>15, product will meet the biodiesel quality as set forth under the ASTM-D6751 standard
without additional purification. The purification cost via vacuum distillation would be
$40.40/hr. To increase product purity from 87 wt% to 99 wt%, the methanol recovery
and recycle cost increases by $40/hr which comes out to be nearly the same when
compared with vacuum distillation of biodiesel product. However, the vacuum
distillation process step will be eliminated.
Because of a very large difference in the boiling points of methanol (33 8K or 65°C)
and biodiesel (598K or 325°C), separation of excess methanol can be easily accomplished
by simple distillation with nearly 100% methanol recovery for recycle.
A detailed design of a sieve tray distillation column using rigorous correlations
suggests a single column of 1.6 meters diameter and a total height of 10 m (33 ft) would
be required. This column will have 12 stages with 24 inches of tray spacing. The feed
will enter at stage 6. The column is designed to operate at 75% flooding. The column
operation for weeping and entrainment was checked, using tray hydraulic correlations.
73
5.2 Recommendations for Future Research
The results of this thesis provide the necessary guidelines for the production of 10
million gallons per year of biodiesel from canola oil. The following recommendations
for future research are made:
A reactor consisting of a reactive distillation column should be investigated as it will
serve as a reactor as well as a separator for the recycle of methanol.
It is recommended that total cost of a 'Reactive Distillation' column should be compared
with the CSTR reactor system.
If commercial scale biodiesel plants are built, there will a glut of glycerol co-product
(1 kg of glycerol is produced for every 10 kg of biodiesel manufacture). It is
recommended that some new process chemistry to convert glycerol into more useful and
economically attractive products be explored. For example, pyrolysis of glycerol may
yield synthesis gas as envisioned by the following reaction (C3H8O3 ->3CO + 4H2).
74
REFERENCES
REFERENCES
Apostolakou, A.A., Kookos, I.K., Marazoiti, C., and Angelopoulos, K.C. (2009). Techno-economic analysis of a biodiesel production process from vegetable oils. Fuel Processing Technology, 90, 1023-1031.
Byrd, A.J., Pant, K.K., and Gupta, R.B. (2008). Hydrogen production from glycerol by reforming in supercritical water over Ru/A1203 catalyst. American Chemical Society, Fuel 87, 2956-2960.
BP Statistical Review of world energy 2008. Available from BP statistical review 2008 internet pages <http://www.investis.com/bp_acc_ia/stat_review_2008/htdocs /reports/reportl 9.html>.
Chavanne, C.G. (1938). Belgian Patent 422, 8 77, Aug. 31, 1937; Chem. Abs.52:4313.
Chemical Engineering Magazine. (2009). M & S index, Pg 64, September edition.
Fogler, H.S. (2006). Elements of Chemical Reaction Engineering. 4th ed. Prentice Hall, NJ.
Gabbard, A. (1993). Coal Combustion Nuclear Resource or Danger. Oakridge National Laboratory Review (ORNL), 26, 25-33.
Haas, M.J., McAloon, A.J., Yee, W.C. and Foglia, T.A. (2006). A process model to estimate biodiesel production costs. Biosource Technology, 97, 671-678.
Kister, H.Z. (1992). Distillation Design. McGraw Hill, NY.
Klepacova, K., Mravec, D., Kaszonyi, A. and Bajus, M. (2007). Etherification of glycerol and ethylene glycol by isobutylene. Applied catalysis, 328, 1-13.
Knothe, G. (2001). Historical Perspectives on Vegetable Oil- Based Diesel Fuel. Inform 12(11), 1103-1107, Retrieved 2009-06-24.
Komers, K., Skopal, F., Stloukal, R., and Machek, J. (2002). Kinetics and mechanism of the KOH - catalyzed methanolysis of rapeseed oil for biodiesel production. European Journal of Lipid Science and Technology, 104(11), 728-737
Leevijit, T., Wistmethangoon, W., Prateepchaikul, G., Tongurai, G., and Allen, M. (2004). A second order kinetics of palm oil transesterification. Joint International Conference on Sustainable Energy and Environment (SEE), 3, 277-281.
Marchetti, J.M., Miguel, V.U., and Errazu, A.F. (2007). Possible methods for biodiesel Production. Renewable and Sustainable Energy Reviews, 11, 1300-1311.
McCance, R. A., Widdowson, E.M., Holland, B., and Paul, A. A. (1991). The Composition Of Food. 5th ed. Ministry of Agriculture, Fisheries and Food., Royal Society of Chemistry, Cambridge.
National Biodiesel Board, cited in Pearl, G.G. (2001, August). Biodiesel Production in the U.S. Render Magazine.
Noureddini, H., and Zhu, D. (1997). Kinetics of Transesterification of Soybean oil. JAOCS, 74, 1457-1463.
Peters, M.S., and Timmerhaus, K.D. (2003). Plant Design and Economics for Chemical Engineers. 5th ed. McGraw-Hill Companies, Inc., NY.
Schuchard, U., Sercheli, R., and Vargas, R.M. (1998). Transesterification of Vegetable Oils: A Review. J. Braz. Chem. Soc., 9, 199-210.
Seider, W.D., Sieder, J.D., and Lewin, D.R. (2004). Product and Process Design Principles:Synthesis, Analysis and Evaluation. 2nd Ed. John Wiley and Sons, Inc.
Sharma, Y.C., Singh, B., and Upadhyay, S.N. (2008). Advancements in development and characterization of biodiesel: A review, fuel, 87, 2355-2373.
Appendix K Design of Methanol Recovery Column Calculations
Top and Bottom Property Data from AspenPlus Simulation
kg P L_Top := 7 3 9 . 8 2 ^ pL_Bottom := 8 8 4 . 1 3
kg 3
m kg
m p is density o f the material
N PV_Top := 1-2639 p v _ B o t t o m : = 2 .6121 ^ °Bottom == 0 - 0 1 8 6 9 9
m
N
m m
a T o p := 0 . 0 3 7 7 9 3 — L T o p := 5 3 6 4 . 2 ^ ~ hr m
V T o p := 8045.8 kg
hr Vsottom := 14649 ^
hr
a is surface tension
b o t t o m := 2 1 4 4 5 ^ hr
m C sb Top := 0 . 0 5 9 — c s b Bottom := 0 . 0 5 6 — C s b = Fair's F lood Factor
Column Diameter
f T FLV_Top :
^Top
Lt, op
V VTop J
= 0 . 6 6 7
P V Top
P L T o p
0.5
vTop
FLV_Top = 0 .028
FLV_Bottom •'=
^Bottom
LBottom
VBottom
P VBottom
V P L Bottom
v 0 .5
^Bottom = 1.464
F L V Bottom = 0 .08
127
^nf Top := Csb_Top f \0.2 aTop
20 dyne
,PL_Top ~ P V Top j
PVTop
0.5
V cm y ft
VnfTop = 5.314 —
vnf_Top = 1.62 — s
V„f Bottom Csb Bottom r \0.2 CTBottom
20 dyne
, P L Bottom ~ PVBottom)
PV Bottom
0.5
V cm y
Vnf Bottom = 3.33 — s
VnfBottom = 1.015 — S
Assuming Flooding = 75%
Assume downcomer occupies 15% of cross-sectional area
vn_Top 0.75-Vnf Top
Vn_Top = 3.986-s
Vn Top = 1.215-s
Vn Bottom : = 0.75-Vnf Bottom
VnBottom = 0.761 — S
V_dotTop := VT, op
PVTop
ft V_dotTop = 62.447 —
V_dotTop = 1.768 — s
128
^Bottom V_dotBottom :=
P V Bottom
ft3
V_dotBottom = 55.014 — s
m3
V_dotBottom = 1-558 — s
V_dotTop AreaTop := —
V n T o p
AreaTop = 15.667 ft2
2 Areaxop = 1.456 m
Areaxop D T o p : = / 4 -
71
DTop = 1.361 m
V_dotBottom AreaBottom : -
'n Bottom
2 Area B o t t 0 m = 22.027 ft
2 Area B o t t 0 m = 2.046 m
AreaB o ttom D B o t t o m := / 4
71
^Bot tom = 5.296 ft
DBottom = 1 . 6 1 4 m
Tray Hydraulics
Rules of thumb for Tray specidicatic
Tray th ickness = t t r a y = 14 g a u g e ( 0 . 0 7 8 i n c h / 0 . 0 0 1 9 m )
H o l e diameter = 0 . 0 0 4 7 m ( 3 / 1 6 inch)
C o l u m n Area = A x
D o w n c o m e r Area = A d = 1 2 % o f A x
N e t Area = A n = 8 8 % o f A T ( V a p o u r F l o w Area)
A c t i v e A r e a = A a = ( A T - 2 A d )
H o l e Area = A h = 10% A T
Weir Length = 1 w = 7 7 % o f Diameter ( T o p or B o t t o m )
2
ATBottom := 2 2 . 1 1 5 f t
2
AT_Bottom = 2 . 0 5 5 m
An Bottom : = 0 . 8 8 - A x Bottom
A n B o t t o m = 19 .461 ft2
2 An Bottom = 1-808 m
A d B o t t o m := 0 . 1 2 - A x Bottom
2 Ad Bottom = 2 . 6 5 4 ft Ad Bottom = 0 - 2 4 7 m 2
Aa_Bottom : = ( A T_Bot tom " 2-A(j_Bottom )
2
Aa_Bottom = 16 .807 ft
2 Aa_Bottom = 1.561 m
Ah Bottom •'= 0-1 Aa_Bottom
2 AhBottom = L681 ft 2
AhBottom = 0 . 1 5 6 m
130
IwBottom 0.77-Deottom
IwBot tom = 4.078 ft
lw Bottom = 1 . 2 4 3 m
^Bot tom Vr, :=
P v Bottom' A h Bottom
v0 = 32.732-ft s
v 0 = 9 . 9 7 7 -m s
Dry Pressure Drop
dhole := — in 16
- 3 dhole = 4 . 7 6 2 x 10 m
ttray := 0.078in
t t r a y = 1 .981 x 10 m
Ahole := ™ (dhole2)
Ahole = 1-781 X 10 5 m 2
C0 := 0.85032-0.04231-f dhole
V W ) + 0 . 0 0 1 7 9 5 4 -
r dhole V
^ ttray j
131
C0 = 0.759
Pwate r := 6 1 . 0 3 -lb ft?
Pwate r = 9 7 7 . 6 0 7 kg
m
Ah Bottom Aa Bottom
|3 = 0.1
v0 = 32.732
v0 = 9 .977-
lb PL_Bottom= 55.194 — ft
PL_Bottom= 8 8 4 . 1 3 ^ m
lb Pv_Bottom =0.163 — ft
hAP_Dty := 0.003-' f tV
PV_Bottom lb
f P water 1 in
— ^PL_Bottom J
hiP_Dry = 0.996 in
132
hAP Dry = 0.025 m
For 75% flooding
Flv = 0.05, y = 0.045(relative entrainment)
v|/ := 0.045
L Bottom
LBottom
L Bottom ee := u/
3 lb ee = 2.228 x 10 —
hr
ee = 0.281 — s
Entrainment := LBottom + e e
4 lb Entrainment = 4.951 x 10 —
hr
Entrainment = 6.238 — s
Fw=Weir Correction Factor due to wall curvature. For large Diameter Column Fw approaches 1.0 otherwise use figure by Bolles
Lg = Liquid Flow rate including entrainmen (L+ee) in the units of gallons/min
(ee + LBottom) Lg :=
P l Bottom
4 l b = 4.728 X 10 —
hr
= 5.957 kg
133
L„ = 111.826 gal min
LP = 7.055 x 10 3 —
Bottom = 4.078 ft
l w Bottom = 1 . 2 4 3 m
Abscissa := \2-5
V Bottom
ft
Abscissa = 3.33
Parameter w Bottom
^ B o t t o m
Parameter = 0.77
F\vear 1 - 0 3 5 (fromgraph)
hcrest := 0.092-Fv
f La ^ gal
V min y
'w_Bottom
ft
•in
hcrest = 0.866 in
hcrest = 0.022 m
134
Assume negligible h gradient. See A P Figure
hgrad := Oin
Assume Downcomer gap= 1 inch
Gap := lin
Gap = 0.025 m
A d u := Iw Bottom - G a p
Ad u = 0.34 ft2
Ad u = 0.032 m2
hdu := 0.56 •
hdu = 0.301 in
hdu = 7.641 x 10" 3 m
hdc = total Head of clear liquid in the Downcomer
Assume Weir Height = 2.0 inches
hweir := 2in
hweir = 0 . 0 5 1 m
hdc : = h A P_Dry + h w e i r + hcrest + hdu
hdc = 4.163 in
gal
V min y
Ad u 449
ft2
•in
135
hdc = 0.106m
Note : In the operating column, the liquid in the Downcomer is Aerated. The densi aerate liquid will be less than that of clear liquid. Thus the height of liquid in Downcomer will be greater thar\j|a
For normal operatiorfdc=0.5
(j)dc := 0.5
hdc hdcae ra t ed
<Pdc
hdcae ra ted = 8.325 in
hdcae ra t ed = 0 . 2 1 1 m
This value is less than 24 inch tray spacing.
There should be no problem for vapour flow to the tray above.
Check for Downcomer time tdc
hdc = 4.163 in
hdc
tdc :
tdc = 3.695 s
The value is greater the minimum residence time of 3 seconds.
There should be no problem.
- 0.106m
( A d . Bottom "PL Bottom "hdc
^Bot tom + e e
136
Weeping Check
Caluclate surface tension head h a
Excessive weeping based on Fair correlation
if X = IVEIR+HCREST+HGRAD then
hAP_Dry+h CT >0.10392+0.25119x-0.021675x2
a := 13.2 dyne cm
dhole := 0.187in
- 3 dhole = 4 . 7 5 x 10 m
ho := 0.04in-
< a ^ dyne
V cm J
P L B o t t o m
j b
ft3
\ f dhole ^
V in J
ho = 0.051 in
- 3 dhole = 4 . 7 5 x 10 m
ho := 0.04in-
' a ^ dyne
V cm J
P L B o t t o m
lb
ftJ
\ ^ dhole
V in J
ho = 0.051 in
137
h a = 1 . 2 9 9 X 1 0 3 m
hwei r = 2 i n
hwei r = 0 . 0 5 1 m
hgrad = 0 i n
hcrest = 0.866 in
hcrest = 0 . 0 2 2 m
x := (2 + 0.865 + 0)
x = 2.865
H A P _ D R Y + h a >0.10392+0.25119x-0.021675x 2
RHS := 0.10392 + 0.2511 - x - 0.021675 -x2
RHS = 0.645
hAP Dry - 0.996 in
hAP Dry = 0.025 m
ha = 0.051 in
ha = 1.299 x 10~3m
hAP Dry + ha = 0.027 m
hAP Dry + h a = 1.047 in
LHS := 1.047
LHS (1.047) > RHS (0.645)
The inequality is satisfied. Therefore, Weeping should not be a problem.
138
BIOGRAPHICAL SKETCH
BIOGRAPHICAL SKETCH
Name of the Author: Shali Vemparala
Place of Birth: Hyderabad, India
Date of Birth: July 14, 1985
Graduate and Undergraduate Schools Attended:
University of South Alabama, Mobile, Alabama Bhoj Reddy Engineering College for Women, Hyderabad, India
Degrees Awarded:
Masters of Science in Chemical Engineering, 2010, Mobile, Alabama Bachelors of Science in Chemical Engineering, 2007, Hyderabad, India